Zulkifly 2019 PH D

ANTICOAGULATION CONTROL IN PATIENTS WITH ATRIAL FIBRILLATION
By
HANIS HANUM BINTI ZULKIFLY
A thesis submitted to the University of Birmingham for the degree of
DOCTOR OF PHILOSOPHY
UNIVERSITY OF BIRMINGHAM
Institute of Cardiovascular Sciences

College of Medical and Dental Sciences
University of Birmingham
Edgbaston
United Kingdom
March 2019
University of Birmingham Research Archive
e-theses repository
This unpublished thesis/dissertation is copyright of the author and/or third parties. The
intellectual property rights of the author or third parties in respect of this work are as defined
by The Copyright Designs and Patents Act 1988 or as modified by any successor legislation.
Any use made of information contained in this thesis/dissertation must be in accordance with
that legislation and must be properly acknowledged. Further distribution or reprodcution in any
format is prohibited without the permisssion of the copyright holder.
Abstract
The studies contained in this thesis highlight key issues relating to knowledge, psychological
aspects and quality of life among newly anticoagulated atrial fibrillation (AF) patients. In
addition, it also included objective measures of the quality of vitamin K antagonist (VKA)
therapy among atrial fibrillation and operated valvular heart disease (VHD) patients on long
term VKA therapy. Three studies were conducted in separate cohorts to achieve these
objectives.
Study 1 (TREAT-2) examined self-report assessment of depression, anxiety, beliefs about

medication and knowledge of AF, and quality of life among 139 newly anticoagulated AF
patients at baseline and 105 patients at six months follow up. Findings suggest that patients
appear to have low levels of depression and anxiety and had positive beliefs about medication.
However, AF knowledge and quality of life was poor. These findings remained unchanged at
6 months follow up. Nevertheless, more patients were aware of AF consequences and had
improvements in AF symptoms.
Study 2 (TTR in relation to ethnicity) investigated objective measures and predictors of quality
of anticoagulation control (time in therapeutic range, TTR) and adverse clinical outcomes in
991 AF patients on VKA therapy over 5.2 years. TTR was compared among patients in
different ethnic groups, elderly (≥80 vs. <80 years), and patients with different categories of
kidney disease (eGFR≥90 vs. 60-89 vs. ≤59ml/min/1.73m2). TTR was significantly lower in
South-Asians [60.5% (95% CI 58.0-63.0)] and Afro Caribbeans [61.3% (95% CI 58.2-64.4)]
compared to Whites [67.9% (95% CI 67.1-68.8); p<0.001] despite similar INR monitoring
intensity. No significant difference in TTR was seen among the elderly or patients with different
categories of kidney disease within this cohort. Non-white ethnicity was the strongest
independent predictor of poor TTR after adjusting for demographics and clinical variables [OR
2.62 (95% CI 1.67-4.10); p<0.001].
Study 3 (TTR in operated VHD) examined TTR, predictors and adverse clinical outcomes
among 456 operated VHD patients with and without AF over 6.2 years. Results showed that
TTR was significantly poorer in operated VHD patients with AF (TTR 55.7%) compared to
those without AF (TTR 60.1%; p=0.002). Independent predictors of poor TTR included female
sex, AF and anaemia/bleeding history.
Findings from Study 1 suggest that among newly anticoagulated AF patients, improvements
are needed in AF knowledge. Although their quality of life is reduced, they were not anxious
or depressed and they hold positive beliefs about their medication. Interventions (educational
1
and motivational) are required so that enhancements in knowledge and quality of life can be
achieved among newly anticoagulated AF patients. Studies 2 and 3 indicate that good
anticoagulation control is more difficult to achieve in non-white AF patients and operated VHD
patients with AF, respectively. This suggests that more frequent INR checks and closer
examination of the individual reasons for poor anticoagulation among these patients is
required to improve quality of anticoagulation control so that adverse events can be prevented.
The findings and conclusions provide a platform for future research to develop interventions
to improve quality of life among anticoagulated AF patients and to improve INR control,
particularly among ethnic minority patients with AF and those with VHD.
2
Publications arising from thesis
Papers
1. Zulkifly H, Lip GY, Lane DA. Bleeding Risk Scores in Atrial Fibrillation and Venous
Thromboembolism. Am J Cardiol 2017; 120 (7): 1139-1145
2. Zulkifly H, Lip GY, Lane DA. Use of the SAMe-TT2R2 score to predict anticoagulation
control in atrial fibrillation and venous thromboembolism patients treated with vitamin
K antagonists: A review. Heart Rhythm. 2018; 15 (4): 615-623
3. Zulkifly H, Lip GY, Lane DA. Epidemiology of AF. Int J Clin Pract. 2018; 73 (3):
e13070.
Abstracts
1. Zulkifly H, Cheli P, Lutchman I, Bai Y, Lip GYH, Lane DA. Quality of and predictors of
anticoagulation control among multi ethnic cohorts receiving VKA therapy for stroke
prevention in atrial fibrillation: Heart Rhythm Congress, Birmingham, UK, 8 October
2018 [Poster]
2. Zulkifly H, Cheli P, Lutchman I, Bai Y, Lip GYH, Lane DA. Anticoagulation control in
elderly AF patients receiving vitamin K antagonists for stroke prevention in atrial
fibrillation: the West Birmingham AF Project: European Society of Cardiology, Munich,
Germany, 27 August 2018 [Poster]
different ethnic-minority patients receiving vitamin K antagonists for stroke prevention
in atrial fibrillation: the West Birmingham AF Project: Festival of Graduate Research,
University of Birmingham, UK, 10 April 2018 [Poster]
in atrial fibrillation: the West Birmingham AF Project: Midlands Cardiovascular
Scientific Meeting, University of Leicester, 23 November 2017 [Poster]
in atrial fibrillation: the West Birmingham AF Project. European Society of Cardiology,
Barcelona, Spain 28 August 2017 [Poster]
6. Zulkifly H. Use of the SAMe-TT2R2 score to predict anticoagulation control in atrial
fibrillation and venous thromboembolism patients treated with vitamin K antagonists:
A review. MySecon, London, 14th May 2017 [Poster]
3
Dedication
I dedicate this thesis to my husband, Naza and my two daughters, Raina and Raisha.
4
Acknowledgements
I would like to say a very-very big thank you (‘terima kasih’) to both of my supervisors, Dr
Deirdre Lane and Professor Gregory Lip for their guidance, patience and support over the past
three years of my PhD journey. This would not have been possible without you.
I am grateful to the staff of the Anticoagulant Clinic, Sandwell and West Birmingham Hospitals
and University Hospital Birmingham for being very helpful and cooperative whilst I was
recruiting patients from their clinics. Also, for giving me access to their anticoagulation
management software to identify patients needed for all my studies. Not to forget to all the
patients who participated in the TREAT-2 study, staff, doctors, nurses and others who helped
me at the AF clinic. A big thank-you to you all.
Special thanks to my friends whom I’ve met along this journey: Paola, Marco, Daniele, Ivana,
Maria, Josemy, Reem, Mazaya, Aeron, Kazuo, and Amy. You have been incredible in giving
me support, strength and motivation to finish this work.
A very big thank you to my family, mum, dad, Hanani, Faiz, Faizal and Naufal for coming all
the way from Malaysia for emotional support and inspiring me to persist in difficult times. To
mom and dad, a very big thank you for keeping me in your prayers and always believing in
me. To dad, special thanks for inspiring me to write in a better way. You have always been my
strength. To Aunt Misbah, a very special thank you for always being there for me, right from
the beginning until the end. You have supported me all the way and encouraged me to
persevere throughout my ups and downs.
Not to forget, a very big thank you to my sponsors, Universiti Teknologi MARA and
Kementerian Pendidikan Tinggi for funding my PhD study as well as my family here in the UK.
Last but not least, to my lovely husband, Naza and my two daughters, Raina and Raisha. A
very-very big thank you to you all for being with me throughout this journey. You have been
very patient with me, my shoulder to cry on and my pillar of strength! Thank you for your
sacrifices, your prayers, your support and for keeping me sane! I could have not made it
without you.
5
List of tables
Table 1.1: Incidence and prevalence of AF and AF- associated mortality rate with 95%
uncertainty intervals (UI) (per 100,000) for males and females (data extracted from Chugh
2014) (1) ............................................................................................................................. 19
Table 1.2: Worldwide prevalence of AF by continent........................................................... 22
Table 1.3: Prevalence of AF by ethnicity and race .............................................................. 26
Table 1.4: Patterns of atrial fibrillation (taken directly from ESC guideline 2016) ................. 34
Table 1.5: Modified European Heart Rhythm Association symptom scale (38).................... 35
Table 1.6: Comorbid conditions/risk factors associated with AF (38)................................... 35
Table 1.7: Adjusted-dose warfarin versus placebo or no treatment, taken directly from Hart
2007 (117) .......................................................................................................................... 46
Table 1.8: Pharmacokinetics of warfarin versus NOACs and baseline characteristics for four
randomised controlled trial cohorts comparing warfarin versus NOACs .............................. 48
Table 1.9: Baseline characteristics of the four randomised controlled trial cohorts comparing
warfarin versus NOACs ...................................................................................................... 48
Table 1.10: Efficacy outcomes in the four major randomised controlled trials comparing
warfarin versus NOACs in AF populations .......................................................................... 49
Table 1.11: Safety outcomes in the four major randomised controlled trials comparing warfarin
versus NOACs in AF populations........................................................................................ 50
Table 1.12: Meta-analysis of real word studies comparing dabigatran, rivaroxaban and
apixaban to warfarin for efficacy and safety outcomes [from Ntaios et al(136)] ................... 52
Table 1.13: Risk factors in stroke risk stratification schemes, updated from Lip et al 2015 (141)
........................................................................................................................................... 56
Table 1.14: CHA2DS2-VASc score components .................................................................. 57
Table 1.15: CHA2DS2-VASc score and thromboembolic risk taken directly from Lip et al from
the European Heart survey (146) and SPORTIF III and IV trial (147) ................................. 57
Table 1.16: Guideline recommendations for stroke prevention in AF .................................. 60
Table 1.17: Risk factors for bleeding included in each bleeding risk score .......................... 66
Table 1.18: Baseline patient characteristics of the derivation cohorts for each bleeding risk
score................................................................................................................................... 68
Table 1.19: Characteristics of the Derivation and Validation cohorts for each of the bleeding
risk scores and composition of each score.......................................................................... 70
Table 1.20: Risk factors, risk categories and bleeding events in the validation cohorts ....... 72
Table 1.21: Demographic and clinical factors affecting anticoagulation control ................... 78
Table 1.22: Non-clinical factors affecting anticoagulation control ........................................ 79
6
Table 1.23: Major studies reporting anticoagulation control in different ethnic groups ......... 85
Table 1.24: Scores to predict anticoagulation control .......................................................... 88
Table 1.25: Mean TTR% in studies validating the SAMe-TT2R2 score................................. 95
Table 1.26: Models of care and anticoagulation control and/or clinical events .................. 100
Table 1.27: Studies assessing the SAMe-TT2R2 score in atrial fibrillation cohorts ............. 103
Table 1.28: Baseline characteristics of all studies assessing the SAMe-TT2R2 score in AF
population ......................................................................................................................... 106
Table 1.29: Predictive ability (C-statistics) of SAMe-TT2R2 for anticoagulation control and
clinical events ................................................................................................................... 112
Table 1.30: Classification of AF patients with valvular heart disease [taken directly from (356,
358)] ................................................................................................................................. 113
Table 1.31:Target INR values for VKA among patients with prosthetic valves [taken directly
from (358, 368)] ................................................................................................................ 116
Table 2.1: Components of the TREAT intervention ........................................................... 127
Table 2.2: Baseline demographics and clinical characteristics of newly anticoagulated AF
patients stratified by the SAMe-TT2R2 score ..................................................................... 143
Table 2.3: Baseline and 6 months follow up psychological measures of AF patients overall and
according to OAC groups (warfarin vs. NOACs) ............................................................... 147
Table 2.4: Changes in psychological measures between baseline and 6 months follow up
among overall AF patients (N=105) .................................................................................. 148
Table 2.5: Baseline and 6 months follow up knowledge scale of AF patients overall and
Table 2.6: Change in knowledge scale between baseline and 6 months follow up among
overall AF patients (N=105) .............................................................................................. 153
Table 2.7: Specific questions in the AF knowledge scale and percentages of patients with
correct response at baseline and 6 months follow up (N=105) .......................................... 154
Table 2.8: Baseline and 6 months follow up scores on beliefs about medication of AF patients
overall and according to OAC groups (warfarin vs. NOACs) ............................................. 157
Table 2.9: Change in score on beliefs about medication between baseline and 6 months follow
up among overall AF patients (N=105).............................................................................. 158
Table 2.10: Baseline and 6 months follow up quality of life scores of AF patients overall and
Table 2.11: Change in quality of life scores between baseline and 6 months follow up among
overall AF patients (N=105) .............................................................................................. 161
Table 3.1: CHA2DS2-VASc, HAS-BLED, and SAMe-TT2R2 scores .................................... 177
Table 3.2: Categories of chronic kidney disease from the NICE guidelines (424).............. 180
7
Table 3.3: Baseline characteristics of the study population overall and stratified by ethnicity
and age (≥80 vs. <80 years) ............................................................................................. 186
Table 3.4: Distribution of patients in the current cohort according to the categories of kidney
disease, N=974................................................................................................................. 190
Table 3.5: Baseline characteristics of overall population with eGFR results and according to
three categories of kidney disease .................................................................................... 192
Table 3.6: Measures of anticoagulation control overall and by ethnic group...................... 197
Table 3.7: Measures of anticoagulation control among overall population and in patients aged
≥80 and <80 years ............................................................................................................ 200
Table 3.8: Measures of anticoagulation control among different categories of kidney disease,
N=974 ............................................................................................................................... 203
Table 3.9: Unadjusted demographic and clinical predictors of time in therapeutic range (TTR)
by the Rosendaal method (linear regression).................................................................... 206
Table 3.10: Adjusted demographic and clinical predictors of time in therapeutic range (TTR)
by the Rosendaal method (linear regression).................................................................... 207
Table 3.11: Unadjusted demographic and clinical predictors of percentage of INRs in range
using the PINRR method (linear regression) ..................................................................... 208
Table 3.12: Adjusted demographic and clinical predictors of percentage of INRs in range using
the PINRR method (linear regression) .............................................................................. 209
Table 3.13: Logistic regression for significant predictors of TTR<70% and PINRR <70% (using
Rosendaal and PINRR methods) ...................................................................................... 210
Table 3.14: Major adverse clinical outcomes among patients receiving warfarin for stroke
prevention for AF overall and by ethnic group ................................................................... 212
Table 3.15: Patients experiencing a major adverse clinical event stratified by TTR (<70% vs.
≥70% and <65% and ≥65%) ............................................................................................. 213
Table 3.16: Patients experiencing a major adverse clinical event stratified by PINRR (<70%
vs. ≥70% and <65% and ≥65%) ........................................................................................ 214
Table 3.17: Cox proportional hazard regression analysis for thromboembolic, bleeding events,
CV hospitalisations and composite outcomes of thromboembolic events, major bleed and
clinically relevant non-major bleeding, cardiovascular hospitalisation and all-cause mortality
(≥1 MACE) ........................................................................................................................ 217
prevention in AF overall and in patients ≥80 and <80 years .............................................. 220
Table 3.19: Cox proportional hazard regression analysis for the impact of age (≥80 years) on
all bleeding events, including major bleeding and clinically relevant non-major bleeding and
≥1 MACE .......................................................................................................................... 221
8
prevention in AF overall and by different categories of kidney disease ............................. 222
Table 4.1: Baseline characteristics of patients with operated valvular heart disease, with and
without AF......................................................................................................................... 248
Table 4.2: Measures of anticoagulation control of patients with operated valvular heart
disease, with and without AF ............................................................................................ 252
Table 4.3: Measures of anticoagulation control of patients according to different target INRs
......................................................................................................................................... 255
Table 4.4: Demographics and clinical characteristics associated with predictors of poor TTR
(<70%), in univariate analysis among patients with operated valvular heart disease, with and
without AF......................................................................................................................... 257
Table 4.5: Models of predictors of poor TTR (<70%) in the overall cohort of patients with
operated valvular heart disease ........................................................................................ 259
Table 4.6: Adverse clinical outcome among patients with operated valvular heart disease, with
and without AF.................................................................................................................. 261
Table 4.7: Adverse clinical outcome vs. TTR among patients with operated valvular heart
disease, with and without AF ............................................................................................ 263
9
List of figures
Figure 1.1: AF associated mortality stratified by sex and region (developed and developing
countries) ............................................................................................................................ 20
Figure 1.2: Prevalence of AF by country ............................................................................. 24
Figure 1.3: ABC pathway [taken directly from Lip 2016 (90)]............................................... 39
Figure 1.4: NICE-AF care management pathway taken from NICE-AF guideline 2014 (91, 92)
........................................................................................................................................... 40
Figure 1.5: European Society of Cardiology integrated pathway adapted from the ESC AF
Guidelines 2016 (3) ............................................................................................................ 41
Figure 1.6: Mean TTR% among studies that validated the SAMe-TT2R2 score.................... 97
Figure 1.7: Mean TTR vs. SAMe-TT2R2 categories in validation studies ........................... 108
Figure 1.8: Predictive ability (C-statistics and 95% confidence intervals) of SAMe-TT2R2
towards anticoagulation control in validation studies ......................................................... 111
Figure 2.1: Study design and patient selection flow chart.................................................. 133
Figure 2.2: Flowchart of patient’s inclusion and follow-up in the study .............................. 142
Figure 2.3: Mean/median score at baseline and 6 months follow up for depression, anxiety,
knowledge of AF, beliefs about medication and quality of life among overall AF patients who
completed the questionnaire at both time points (N=105) ................................................. 149
Figure 2.4: GAD-7 scores in categories among overall AF patients at baseline and 6 months
follow up (N=105).............................................................................................................. 150
Figure 2.5: PHQ-9 scores in categories among overall AF patients at baseline and 6 months
follow up (N=105).............................................................................................................. 150
Figure 2.6: Major depression (PHQ-9 ≥15) and generalised anxiety disorder (GAD-7≥10)
among overall AF patients at baseline and 6 months follow up (N=105) ........................... 151
Figure 2.7: Proportion of AF patients with correct answers in each specific question at baseline
and 6 months follow up (N=105) ....................................................................................... 155
Figure 2.8: Quality of life domain scores assessed by the AFEQT questionnaire in AF patients
overall at baseline and 6 months follow up (N=105) .......................................................... 162
Figure 3.2: Measures of anticoagulation control (including TTR by Rosendaal and PINRR
method) by ethnicity.......................................................................................................... 198
Figure 3.3: Percentage of patients by ethnic group with a therapeutic (TTR≥70%) TTR by the
Rosendaal and PINRR method ......................................................................................... 198
Figure 3.4: Percentage of INRs within therapeutic range (PINRR), and below (INR<2.0) and
above................................................................................................................................ 199
Figure 3.5: Percentage of patients with an INR >5.0 and >8.0 .......................................... 199
10
Figure 3.6: Mean percentage TTR and PINRR and number of visits among patients aged ≥ 80
and <80 years ................................................................................................................... 201
Figure 3.7: Measures of anticoagulation control in different categories of kidney disease . 204
Figure 3.8: The proportion of INRs within, below and above therapeutic range in different
categories of kidney disease ............................................................................................. 204
Figure 3.9: Impact of TTR on cardiovascular hospitalization ............................................. 218
Figure 3.10: Impact of TTR on composite endpoints of thromboembolic events, major and
clinically relevant non-major bleeding events, cardiovascular hospitalisation and all-cause
mortality (≥1 MACE).......................................................................................................... 218
Figure 3.11: Kaplan-Meier curve of bleeding events among patients age ≥80 and <80 years
......................................................................................................................................... 221
Figure 4.2: Percentage of patients with optimal TTR/PINRR among operated valvular heart
disease patients with and without AF ................................................................................ 253
Figure 4.3: Percentage of INRs within range, below the range and above the range among
operated valvular heart disease, with and without AF ....................................................... 254
Figure 4.4: Percentage of patients with INRs>5.0 and INRs >8.0 among operated valvular
heart disease, with and without AF ................................................................................... 254
Figure 4.5: Kaplan-Meier curves among operated VHD patients stratified by the presence of
AF for all-cause mortality .................................................................................................. 262
Figure 4.6: Kaplan-Meier curves among operated VHD patients stratified by categories of TTR
(TTR <70% vs. TTR≥70%) for all-cause mortality ............................................................. 264
Figure 4.7: Kaplan-Meier curves among operated VHD patients stratified by categories of TTR
(TTR <70% vs. TTR≥70%) for composites of thromboembolic, bleeding event, cardiovascular
hospitalisation and all-cause mortality (≥1 MACE) ............................................................ 264
11
List of abbreviations
AC Anticoagulant clinic
ACCP American College of Chest Physicians
ACV Anticoagulant variability
AF Atrial fibrillation
AFEQT Atrial Fibrillation Effect on Quality of Life
ALP Alkaline phosphatase
ALT Alanine transferase
AV Atrioventricular
AVR Aortic valve replacement
BDI Black Depression Inventory
BMQ Beliefs about medication
CABG Coronary Artery Bypass Graft
CAVD Calcific aortic valve disease
CCS Canadian Society of Cardiology
CDA Clinical data archive
CI Confidence interval
CKD Chronic kidney disease
CrCl Creatinine clearance
CRNMB Clinically relevant non-major bleed
cTnT-hs High sensitive cardiac troponin
CV Cardiovascular
CYP2C9 Cytochrome P450 2C9
DASS Depression Anxiety Stress Scale
DSM-IV Diagnostic and Statistical Manual of Mental Disorders, 4th Edition
ECG Electrocardiogram
EF Ejection fraction
eGFR Estimated glomerular filtration rate
EHRA type 1 VHD Evaluated Heart valves, Rheumatic or Artificial type 1 valvular
heart disease
EHRA type 2 VHD Evaluated Heart valves, Rheumatic or Artificial type 2 valvular
heart disease
ESC European Society of Cardiology
ESRD End stage renal disease
GAD-7 Generalised Anxiety Disorder-7
GDF-15 Growth differentiation factor 15
GI Gastrointestinal
HADS Hospital Anxiety and Depression Scale
Hct Haematocrit
HR Hazard ratio
HRA Health Research Authority
HRQoL Health related quality of life
ICD International Classification of Diseases
ICM Implantable cardiac monitoring
INR International normalised ratio
IQR Interquartile range
ISTH International Society on Thrombosis and Haemostasis
KDIGO Kidney Disease Improving Global Outcomes
MACE Major adverse clinical events
MD Mean difference
mEHRA Modified European Heart Rhythm Association
MI Myocardial infarction
12
MTAC Medication Therapy Adherence Clinic
MVR Mitral valve replacement
NICE The National Institute for Health and Care Excellence
NNT Number needed to treat
NNH Number needed to harm
NOAC Non-vitamin K antagonist oral anticoagulants
NVAF Non-valvular atrial fibrillation
NYHA New York Heart Association
OAC Oral anticoagulant
OR Odds ratio
PAF Paroxysmal atrial fibrillation
PCI Percutaneous coronary intervention
PE Pulmonary embolism
PHQ-9 Patient Health Questionnaire-9
PINRR Percentage of INRs in range
PMAS Pharmacist managed anticoagulant services
POC Point of care
PSM Patient self-monitoring
PST Patient self-testing
PTCA Percutaneous transluminal coronary angioplasty
QoL Quality of life
R&D Research and Development
RCT Randomised controlled trial
REC Research Ethics Committee
RHD Rheumatic heart disease
RR Relative risk
SA Sinoatrial
SD Standard deviation
SE Systemic embolism
SF-36 Short Form Health Survey
STAI-S State-Trait Anxiety Inventory
SWBH Sandwell and West Birmingham Hospitals
TE Thromboembolism
TIA Transient ischemic attack
TTR Time in therapeutic range
UC Usual care
UHB University Hospitals Birmingham
UI Uncertainty intervals
USA United States of America
VHD Valvular heart disease
VKA Vitamin K antagonist
VKORC1 Vitamin K Epoxide Reductase Complex Subunit1
VTE Venous thromboembolism
13
Table of Contents
Abstract _______________________________________________________________________ 1
Publications arising from thesis ____________________________________________________ 3
Dedication _____________________________________________________________________ 4
Acknowledgements _____________________________________________________________ 5
List of tables ___________________________________________________________________ 6
List of figures __________________________________________________________________ 10
List of abbreviations ____________________________________________________________ 12
Table of Contents ______________________________________________________________ 14
Chapter 1. Literature review _______________________________________________ 18
1.1 Epidemiology of atrial fibrillation ___________________________________________ 18

1.1.1 Incidence and prevalence of atrial fibrillation worldwide _______________________________ 18
1.1.2 Prevalence of atrial fibrillation by continents ________________________________________ 21
1.1.3 Ethnicity, age and prevalence of atrial fibrillation _____________________________________ 25
1.1.4 Incidence and prevalence of atrial fibrillation in stroke patients _________________________ 28
1.2 Overview of atrial fibrillation _______________________________________________ 30

1.2.1 Pathophysiology of atrial fibrillation _______________________________________________ 30
1.2.2 Pathophysiology atrial fibrillation related thromboembolism ___________________________ 31
1.2.3 Diagnosis and detection of atrial fibrillation _________________________________________ 31
1.2.4 Patterns of atrial fibrillation ______________________________________________________ 32
1.2.5 Symptoms of atrial fibrillation ____________________________________________________ 34
1.2.6 Risk of developing atrial fibrillation ________________________________________________ 35
1.2.7 Treatment of atrial fibrillation ____________________________________________________ 36
1.2.8 Pathways to management of atrial fibrillation _______________________________________ 38
1.3 Psychological aspects in patients with atrial fibrillation and the impact on quality of life
42
1.3.1 Depression and anxiety _________________________________________________________ 42
1.3.2 Anticoagulation and quality of life _________________________________________________ 43
1.4 Stroke prevention in atrial fibrillation ________________________________________ 45

1.4.1 Antithrombotic therapy in stroke prevention in atrial fibrillation _________________________ 45
1.4.2 Assessing risk factors for stroke___________________________________________________ 53
14
1.4.3 Predicting stroke risk with clinical risk scores ________________________________________ 54
1.4.4 Guidelines and recommendations for stroke risk stratification and antithrombotic therapy____ 58
1.4.5 Assessing bleeding risk in atrial fibrillation __________________________________________ 65
1.5 Anticoagulation control in stroke prevention __________________________________ 78

1.5.1 Factors affecting anticoagulation control ___________________________________________ 78
1.6 Use of the SAMe-TT2R2 score to predict anticoagulation control in atrial fibrillation
treated with vitamin K antagonists _______________________________________________ 101
1.6.1 SAMe-TT2R2 score validation studies ______________________________________________ 102
1.6.2 Importance of good anticoagulation control ________________________________________ 109
1.6.3 Impact of different methods of calculating TTR _____________________________________ 110
1.7 Antithrombotic therapy in atrial fibrillation associated with valvular heart disease __ 113
1.7.1 Epidemiology of valvular heart disease with atrial fibrillation __________________________ 114
1.7.2 Anticoagulation therapy in AF patients with valvular heart disease ______________________ 114
1.8 Aims and objectives _____________________________________________________ 121
Chapter 2. A prospective study examining non-vitamin K antagonist oral anticoagulants

(NOACs) versus warfarin based on the SAMe-TT2R2 score strata in anticoagulant-naïve
patients with atrial fibrillation: the TREAT-2 study ______________________________ 122
2.1 Abstract _______________________________________________________________ 122
2.2 Background and rationale ________________________________________________ 125

2.2.1 Study objective_______________________________________________________________ 127
2.3 Methods ______________________________________________________________ 128

2.3.1 Procedure___________________________________________________________________ 129
2.3.2 Assessment of medication adherence _____________________________________________ 131
2.3.3 Study outcomes ______________________________________________________________ 134
2.3.4 Hypothesis __________________________________________________________________ 138
2.3.5 Statistical analysis ____________________________________________________________ 139
2.4 Results ________________________________________________________________ 141

2.4.1 Baseline demographics and clinical characteristics of AF patients _______________________ 141
2.4.2 Psychological measures, knowledge and beliefs about medication of AF patients overall and
according to OAC groups (warfarin vs. NOACs) ____________________________________________ 146
2.4.3 AF knowledge________________________________________________________________ 152
2.4.4 Beliefs about medication _______________________________________________________ 156
2.4.5 Quality-of-life ________________________________________________________________ 159
15
2.4.6 Time in therapeutic range, medication adherence and adverse clinical outcome at 6 months
follow up __________________________________________________________________________ 163
2.5 Discussion _____________________________________________________________ 164

2.5.1 Limitations __________________________________________________________________ 168
2.5.2 Clinical implications and future research ___________________________________________ 169
2.6 Conclusion _____________________________________________________________ 170
Chapter 3. Anticoagulation control in different ethnic groups receiving vitamin K

antagonist therapy for stroke prevention in atrial fibrillation: the West Birmingham
AF Project 171
3.1 Abstract _______________________________________________________________ 171
3.2 Introduction ___________________________________________________________ 173

3.2.1 Study objectives ______________________________________________________________ 174
3.3 Methods ______________________________________________________________ 174

3.3.1 Study design _________________________________________________________________ 174
3.3.2 Patient selection _____________________________________________________________ 175
3.3.3 Procedure___________________________________________________________________ 177
3.3.4 Variables and definitions _______________________________________________________ 178
3.4 Results ________________________________________________________________ 184

3.4.1 Baseline characteristics ________________________________________________________ 184
3.4.2 Measures of anticoagulation control ______________________________________________ 196
3.4.3 Predictors of time in therapeutic range (TTR) by the Rosendaal and PINRR methods ________ 205
3.4.4 Major adverse clinical outcomes _________________________________________________ 210
3.5 Discussion _____________________________________________________________ 223

3.5.1 Anticoagulation control in the overall cohort _________________________________________ 223
3.5.2 Predictors of anticoagulation control in the whole cohort _____________________________ 230
3.5.3 Adverse clinical outcomes in the whole cohort ______________________________________ 231
3.5.4 Strengths and limitations _______________________________________________________ 234
3.5.5 Clinical implications ___________________________________________________________ 235
3.5.6 Future research ______________________________________________________________ 236
3.6 Conclusions ____________________________________________________________ 236
Chapter 4. Anticoagulation control in operated valvular heart disease patients with and
without atrial fibrillation receiving vitamin K antagonist _________________________ 238
16
4.1 Abstract _______________________________________________________________ 238
4.2 Introduction ___________________________________________________________ 240

4.2.1 Study objectives ______________________________________________________________ 240
4.3 Methods ______________________________________________________________ 241

4.3.1 Study design _________________________________________________________________ 241
4.4 Results ________________________________________________________________ 247

4.4.1 Baseline characteristics ________________________________________________________ 247
4.4.2 Quality of anticoagulation control of patients with operated valvular heart disease, with and
without AF_________________________________________________________________________ 251
4.4.3 Predictors of poor anticoagulation control, TTR <70% ________________________________ 256
4.4.4 Adverse clinical outcome _______________________________________________________ 260
4.5 Discussion _____________________________________________________________ 265

4.5.1 Strengths and Limitations ______________________________________________________ 270
4.5.2 Clinical implications ___________________________________________________________ 270
4.5.3 Future work _________________________________________________________________ 271
4.6 Conclusion _____________________________________________________________ 272
Chapter 5. General discussion and conclusions ________________________________ 273
Appendix 1 __________________________________________________________________ 280
Appendix 2 __________________________________________________________________ 282
Appendix 3 __________________________________________________________________ 295
Appendix 4 __________________________________________________________________ 299
Appendix 5 __________________________________________________________________ 311
References___________________________________________________________________ 330
17
Chapter 1. Literature review
1.1 Epidemiology of atrial fibrillation
This section summarises the literature on the epidemiology of atrial fibrillation (AF) worldwide
according to continents, age and ethnicity/race, and also includes the prevalence of AF in
stroke patients. This has been published in the International Journal of Clinical Practice (1).
Atrial fibrillation (AF), the most common form of arrhythmia with clinical significance, is a major
global health burden worldwide (2). In the United States (US) and European countries, one in
every four middle-aged adults will develop AF. Most importantly, AF is associated with a five-
fold greater risk of stroke, increased risk of death and development of heart failure, and a
greater risk of hospital admission, with 10-40% of AF patients hospitalised annually.
Additionally, irrespective of other cardiovascular-related conditions, AF patients have poorer
quality of life and unfortunately, despite anticoagulation for stroke prevention, they can still
develop vascular dementia and a decline in cognitive function (3).
1.1.1 Incidence and prevalence of atrial fibrillation worldwide

According to the Global Burden of AF(4), worldwide, the projected number of people with AF
in 2010 was 33.5 million, consisting of 20.9 million males (UI, 19.5-22.2 million) and 12.6
million females (UI, 12.0-13.7 million), with higher incidence and prevalence rates in
developed countries (Table 1.1). Mortality associated with AF globally is higher in females,
primarily driven by higher mortality among females in developing countries (Figure 1.1).
18
Table 1.1: Incidence and prevalence of AF and AF- associated mortality rate with 95% uncertainty intervals (UI) (per 100,000) for
males and females (data extracted from Chugh 2014) (4)
1990 2010 1990 2010

Male Female
Incidence of AF
Globally, all ages 60.7 (49.2-78.5) 77.5 (65.2-95.4) 43.8 (35.9-55.0) 59.5 (49.9-74.9)
Developed Countries 78.4 (67.5-91.9) 123.4 (107.6-141.5) 52.8 (45.0-62.9) 90.4 (77.8-104.5)
Developing Countries 50.0 (33.8-76.8) 53.8 (38.7-79.8) 36.0 (24.5-54.7) 40.0 (27.2-62.6)
Prevalence of AF
Developing Countries 546.6 (503.0-599.6) 656.7 (522.9-617.6) 358.2 (329.8-393.0) 366.1 (337.4-400.8)
Mortality rate
Developing
Countries 0.4 (0.2-0.8) 0.7 (0.4-1.3) 0.7 (0.4-1.4) 1.0 (0.6-1.7)
AF=atrial fibrillation
19
3
2.7
2.5 2.4
2
Mortality per 100,000
1.5
1.3 Male
1.1 Female
1
1
0.7 0.7
0.5 0.4
0
1990 2010 1990 2010
Developed Countries Developing Countries
Figure 1.1: AF associated mortality stratified by sex and region (developed and
developing countries)
20
1.1.2 Prevalence of atrial fibrillation by continents
In Europe, AF currently affects eight million people and is expected to rise drastically, 2.3-fold
by 2060 (5, 6). In the United Kingdom, projections from the Clinical Practice Research
Database suggest that AF will affect between 1.3 and 1.8 million people by 2060 (7). In the
United States, about 3-5 million people are currently affected by AF and by 2050 this figure is
expected to be greater than 8 million people (8). In Australia, Europe and the USA, the current
estimated prevalence of AF is between 1-4% (2, 5). Table 1.2 and Figure 1.2 illustrate the
prevalence of AF in nine countries, stratified by continents worldwide. Australia has the highest
prevalence of AF i.e., 5.4% followed by Africa 4.6%, (although the prevalence was lower
(0.7%) in another African study) (9) then Iceland (2.4%) and lowest in Asian countries (0.49%-
1.9%).
A recent review (10) on AF epidemiology of 58 studies from five Asian (China, Japan, South
Korea, India, Malaysia) and eight Middle Eastern countries (Turkey, Bahrain, Qatar, Kuwait,
Saudi Arabia, Oman, United Arab Emirates and Yemen)(10) reported the annual incidence of
AF to be 5.38 per 1000 person-years. These are mainly from Chinese, Japanese and Korean
studies (10 studies in total) (11-20) conducted from 1991-2012, with study populations ranging
from 1485 (15) to 471,446 (13). Prevalence of AF varies between hospital-based and
community-based studies; being higher in the latter (0.37%-3.56% vs. 2.8%-15.8%) (10).
21
Table 1.2: Worldwide prevalence of AF by continent
-Years data Study population

Prevalence (%) total,
Country Study obtained Sample size Age (SD) years
Data source (men and women)
-Study design [Men, women (%)]
Africa
Hospital-based,
Sliwa et al -2006-2008 5328 cardiac 59 (18)
Africa single centre, urban 4.6 (†)
2010 (21) -Prospective cases [39, 61% AF]
population
One hospital
-2008-2010 ≥18
Kenya Shavadia et al 44, 144 admission in 0.7 (†)
-Retrospective [56, 44% AF]
2013 (9) Nairobi
Asia
18 urban, 22 rural
Malaysia Lim et al -2007-2014 10,805 52.6 (11.6) 0.49 (†)
communities across
2016 (22) -Prospective
Malaysia
Singapore Yap et al † 1,839 Community-based ≥55 1.4 (†)

2008(23) -Prospective study
† Cross section of
Thailand Phrommintikul et al 1,277 ≥65 1.9 (†)
-Prospective Maerim District,
2016 (24) [45.8, 54.2% AF]
Chiang Mai
Australia
Ball et al 7 international
Australia -June 2014 6,140,651 ≥55 5.4 (5.97, 4.79)
2015(25) epidemiology study
Sturn et al
50 consecutive
Australia 2002 (26) -2000 14, 194 ≥30 4.0 (6.0, 4.0)
patients at 321
-Prospective
general practices
22
Table 1.2 continued
Study population
-Years data
Age (SD) Prevalence (%) total,
Country Study obtained Sample size
Data source [Men, women (men and women)
-Study design
(%)]
Europe
Health & Social
National primary
Care Information -2014-2015
England 56,939,507 care practice † 1.6 (†)
Centre 2014-2015 -Retrospective
database
(27)
UK Clinical
United Lane et al -1998-2010 Practice ≥18 1.26 (1.33, 1.18) per 1000 pt-
57, 818
Kingdom 2017 (7) -Retrospective Research [51.7, 48.3% AF] yrs [age-adjusted incidence]
Datalink (CPRD)
The National
-1987-31 December 4905-AF 2.4-age and sex
Iceland Stefansdottir et al University 20–99
2008 cases standardised
2011(6) Hospital of
-Retrospective
Iceland
North America
National
databases of
Naccarelli et al -2004-2005
USA 242, 903 employer-funded ≥20 1.1 (†)
2009(28) -Retrospective
insurance and
Medicare
South America
Marcolino et al -Jan-December 658
50.3 (19.3)
Brazil 2015(29) 2011 262 685 municipalities, 1.8 (2.4, 1.3)
[40.4, 59.6]
-Retrospective primary Care
† Not reported
23
Figure 1.2: Prevalence of AF by country
24
Differences in the incidence and prevalence rates between studies is likely dependent on the
time the study was conducted, the design (nationwide studies, medical insurance databases
etc., retrospective, prospective, cross-sectional) and the study population (age of patients,
urban compared to remote areas as their risk factors may vary, for example the prevalence of
rheumatic heart disease is higher in rural populations in India)(30) which in turn affect the
quality of data obtained (5).
1.1.3 Ethnicity, age and prevalence of atrial fibrillation
The prevalence of AF across different ethnic groups differs, although most of the studies
investigating these differences have been conducted in the United States. For the purpose of
this thesis, ethnicity is classified as Whites, (Europeans, Americans) Afro-Caribbeans,
(Blacks, Black British, African-American) East Asians, (Chinese, Japanese, Malaysian and
other Asians) South Asians, (Indian, Pakistani, Bangladeshi) Hispanics, (Hispanic or Latino)
and others, as reported in individual studies (31, 32).
Table 1.3 shows the prevalence of AF by race and ethnicity according to 11 studies,(33-43)
conducted in the United States and one (33) multicentre study conducted in North America,
Europe and Asia. In all studies, the prevalence of AF was highest among the Whites compared
to Afro-Caribbeans, East Asians and Hispanics, ranging from 42% (42) to 2.5% (41) among
the Whites, and 21%(41) to 1.7% (40) among Afro-Caribbeans. Only three studies (33, 35,
43) reported AF prevalence among East Asians which ranged from 3.9% (35) to 10.1% (33),
while only one study reported AF prevalence among Hispanics (3.9%) (35). A meta-analysis
of 10 studies examining the prevalence of AF among African-Americans compared to Whites
in the United States, concluded that being African-American was associated with a ‘protective
effect’ from AF [OR 0.51 (95% CI 0.44-0.59); p<0.001]. Despite the lower prevalence of AF,
African-Americans in the US have twice the risk of first ever stroke compared to Whites and
this might be due to higher risk factor burden of stroke, for example, hypertension (44).
25
Table 1.3: Prevalence of AF by ethnicity and race
Country/State Study a) Study design Mean (SD) age Prevalence of AF (%)

b) Follow up study population Whites Afro- East Hispanics
c) Sample size Caribbeans Asians
North America, Lau et al a) Prospective Europeans: 76.2 (6.6) 18 8.3 10.1 N/A
Europe, Asia 2013(33) b) 2.5 years Black African: 75.2 (6.2) Chinese,
c) 2580 Chinese: 76.2 (6.7) 9.5
Japanese: 78.4 (7.0) Japanese
Michigan, Lahiri et al a) Retrospective African American: 33.0% 29 19 N/A N/A

USA 2011(34) b) N/A ≥70
c) 1001 European American:
35.4% ≥70
USA Winkelmayer et al a) Cross sectional 70.9 (11.8) 14 6.5 9.0* N/A

2011(43) b) 15 years
c) 2,483,199
California, Shen et al a) Cross-sectional White: 70 (64-77) 8 3.8 3.9 3.6

USA 2010(35) b) N/A Black: 68 (64-74)
c) 430, 317 Asian: 67 (63-73)
Hispanic: 67 (61-71)
15 U.S states, Marcus et al a) Combination of CHS CHS 23 15 N/A N/A

Washington, 2010(36) and ARIC study African American: 73 (6)
DC b) – Whites: 73 (6)
c) 19, 784 ARIC
African Americans: 53
(6)
Whites: 54 (6)
26
Table 1.3 continued
Country/State Study a. Study design Mean (SD) age Prevalence of AF (%)

b. Follow up study population Whites Afro- East Hispanics
c. Sample size Caribbeans Asians
Maryland, Alonso et al a) Prospective Whites: 54.4 (5.7) 7.9 4.8 N/A N/A
Minnesota, 2009(37) b) 228,976 person-years African Americans: 53.6
Mississippi, N. c) 15, 407 (5.8)
Carolina, USA
Ohio, Smith et al a) Prospective N/A 24 17 N/A N/A
USA 2006 (38) b) -
c) 9671
California, Ruo et al a) Retrospective and 73 overall 38 20 N/A N/A

USA 2004 (39) prospective
b) –
c) 1373
Georgia, Upshaw et al a) Retrospective 14% age 70-79 7.8 2.5 N/A N/A
USA 2002(40) b) –
c) 2123
California, Go et al a) Cross sectional 71.2 (12.2) whole cohort 2.5 1.7 N/A N/A
USA 2001(41) b) N/A
c) 17, 974
a) Prospective Blacks: 63.8 (13.7) 42 21 N/A N/A
Michigan, Afzal et al b) 6 months Whites: 70.8 (13.1)
USA 1999(42) c) 163
Pt-yrs = patient years; *Asian Americans; N/A = not applicable
27
In addition to ethnic differences, age distribution of AF diagnosis may also differ between
regions. More than 70% of AF patients in Western Europe, Australia and North America were
aged >65 years (2). A different pattern in the average age of AF diagnosis is evident from
other regions. AF patients are younger from the Arabic (45), Ethiopian (46), South Korean (2)
and South African (21) studies with mean age ranging from 41-65 years. Results from the RE-
LY AF registry which enrolled AF patients from the 164 emergency departments worldwide to
evaluate the differences in the presentation and management of AF, also shows some regional
variation in terms of age at AF diagnosis; patients from America and Europe countries were
on average 10-12 years older than those from Africa, India and the Middle East (47).
1.1.4 Incidence and prevalence of atrial fibrillation in stroke patients
AF increases the risk of stroke approximately 5 times compared to those without AF (3). The
presence of AF was 24.6% in patients [mean (SD) age 78.8 (13.3) years] with ischemic stroke
in one Italian population-based study. In this study, AF was more frequent in women, elderly
patients (>80 years), those with coronary heart disease and peripheral arterial disease. AF
was also an independent predictor of 30-day and 1-year mortality in Cox regression analysis
(48).
Another prospective study in Germany reported an overall prevalence of AF slightly higher
than the Italian study, i.e., 28.6% in patients (N=692) with ischemic stroke or transient ischemic
attack, with prevalence increasing with age (49). The prevalence of AF in ischaemic stroke
patients in the USA (i.e., 23% in acute ischaemic stroke patients from eight states) (50), two
European studies (24.6 and 28.6% respectively) (29, 30) and one Australian study (25%; study
population 26,960) (51) was similar to that reported by the Italian study. However, the
prevalence of AF was reportedly lower in some Asian countries, approximately 10% in China
(52), 5.8-6% in India (53, 54), but higher in Japan (32%) (55).
28
Sposato et al conducted a meta-analysis of 50 studies in 2015 (56) to estimate the proportion
of newly diagnosed AF patients experiencing stroke or transient ischemic attack (TIA) after
undergoing four sequential phases of cardiac monitoring; phase 1: electrocardiogram (ECG)
at admission, phase 2: continuous inpatient ECG, phase 3: Holter monitoring and phase 4:
mobile cardiac outpatient telemetry. In this study, they reported an overall presence of AF in
23.7% (95% CI 17.2-31.0) of their post-stroke patients and an estimated prevalence of AF in
post-stroke patients with either known or newly diagnosed AF of about 39.0%, higher than
previously reported studies (56).
The prevalence of AF worldwide is increasing steadily although large variation can be seen
between studies and countries. A larger proportion of ischemic stroke patients are also found
to have AF either during admission or upon investigation post-discharge that becomes a major
concern as AF related to stroke has poorer outcomes and prognosis (3, 5). This increase in
the prevalence of AF may be explained by the fact that better detection methods have been
used to detect AF (2, 3, 57) and also greater awareness among physicians and other
healthcare providers who are able to detect patients with AF during routine check-ups, flu
injections and also during hospital admissions.
Further epidemiological studies should be undertaken globally, especially in Asian and African
countries, in urban and rural areas, so that a more accurate picture of the incidence and
prevalence of AF can be captured, thereby allowing appropriate implementation of stroke
prevention strategies to reduce stroke risk and burden.
The next section summarises the pathophysiology, diagnosis, pattern and symptoms of risk
and general treatment of AF.
29
1.2 Overview of atrial fibrillation
Atrial fibrillation is a condition where there is an ineffective atrial contraction that results from
an uncoordinated atrial activation. The sino-atrial (SA) node which sends electrical impulses
to the atria for atrial contraction (thus forcing the blood to enter the ventricles) and is also
responsible for controlling and coordinating the heart rate, is no longer functioning in an
organised manner, thereby causing an irregular and rapid heart rate (58).
1.2.1 Pathophysiology of atrial fibrillation
Atrial fibrillation occurs when the electrical signalling pathway is abnormal (3). The signals are
generated from all over the atria causing a fibrillating or quivering atrial activity. The signals
are no longer systematically triggered via the SA node (58). Specifically, the pulmonary veins
located in the left atrium generate multiple impulses in majority of AF cases (3). The generated
impulses can be fired at a very rapid rate of about 300-600 beats per minute, however, not all
of the impulses can be filtered by the atrioventricular (AV) node (59). The signals originating
from multiple areas within the atria are chaotic, fast and irregular, leading to inadequate atrial
emptying. Excessive signals passing through the AV node causes the ventricular rate to be
increased (59). The emptying of the ventricles is affected by the increase in ventricular activity
(60-130 beats per minute) and if this continues, it leads to reduced general circulation of the
blood causing symptoms of fatigue, light-headedness, breathlessness and chest pain (59).
Hypertension, diabetes, heart failure, coronary artery disease, and ageing are among the
factors that can lead to changes to the pathophysiology of the atria, including hypo-
contractility, inflammation, vascular remodelling, inflammation, fatty infiltration and ion channel
dysfunction (3). These changes can cause conduction disturbances and thus lead to the
development of AF. Some of these changes are also involved in the manifestation of a
hypercoagulable state in AF.
30
1.2.2 Pathophysiology atrial fibrillation related thromboembolism
Tissue factor exposure in the blood stream as a result of hypocontractility and ischemia
induces inflammation and adds to the thrombogenic environment in the atria of AF patients,
thus leading them to have an increased risk of thromboembolic events such as stroke or
transient ischemic attack (TIA) (3). Along with the structural remodelling, the rhythm of AF
itself predisposes the atrial myocardium to a prothrombotic state (60). Additionally, the
myocardial damage within the atrium that develops within short periods of AF stimulates the
release of prothrombotic factors onto the endothelial surface causing platelet aggregation.
This could partly elucidate the reason of long-term stroke risk even in short episodes of AF
(61, 62).
1.2.3 Diagnosis and detection of atrial fibrillation
AF diagnosis is made using the ECG that shows a typical pattern of AF involving irregular RR
intervals and no distinct P waves. Any episode of AF lasting at least 30 seconds is considered
as a diagnosis of AF (3). Patients with AF can be symptomatic or asymptomatic (‘silent AF’).
Silent AF can have severe consequences such as stroke and death just as with symptomatic
AF (63-65). Such events can be avoided or reduced with early detection and OAC initiation.
Suggestions have been made to screen AF in a more widespread manner including within
community healthcare practices (66) and using sophisticated diagnostic tools which could
detect short and long episodes of AF. One systematic review (67) of eight trials (N=18,189)
conducted in the GP/outpatient clinic and community setting identified a 1.4% incidence of
undiagnosed AF using a single time point screening AF method (pulse palpation or ECG) of
patients aged ≥65 years (67). The pulse palpation method was reported to have a sensitivity
of 94% and a specificity of 72% in detecting AF (68) while the handheld single lead ECGs
have higher sensitivity and specificity ranging from 94%-98% and 76%-97% respectively (69).
Evidence has shown that detection of undiagnosed AF can be obtained by prolonged ECG
monitoring (70, 71). In 2014, two trials (70, 72) were conducted to investigate whether long
31
term ECG monitoring is superior to conventional 24-hour monitoring in detecting AF in patients
with cryptogenic stroke. The EMBRACE-AF trial (72) included 572 patients with cryptogenic
stroke within the previous 6 months and investigated the benefits of longer monitoring periods
(30 days) versus conventional 24-hour ECG monitoring. In the study, at least 16.2% of patients
had AF for 30 seconds over 90 days of monitoring compared to just 3.2% for those undergoing
24-hour monitoring (an absolute difference of 12.9 percentage points [95% CI (8.0 to 17.6);
p<0.001])(72).
The CRYSTAL AF trial (70), with a slightly lower number of participants (N=447), examined if
‘conventional follow-up’ was better than continuous cardiac monitoring via implantable cardiac
monitor (ICM) in detecting AF in patients with cryptogenic stroke after 3 months of the event.
At 6 months, it was similarly shown that the rate of AF detection is higher in patients receiving
ICM compared to the conventional group [8.9% vs. 1.4%, (HR 6.4; 95% CI 1.9 - 21.7;
p<0.001)] (70). Essentially, the benefit of continuous monitoring was also seen at 12 months
with a 12.4% AF detected in patients with ICM compared to 2.0% in the conventional group
(70). Despite the available evidence to date, the best method for detecting and screening AF
is still unclear.
Evidence demonstrates the benefits of continuous ECG monitoring, the current guideline (3)
recommends ECG rhythm strip and pulse palpation for primary prevention and at least 72
hours of monitoring in post stroke/TIA patients.
1.2.4 Patterns of atrial fibrillation
In most cases, AF progresses from infrequent, short episodes to longer and more frequent
episodes, and a sustained form of AF can develop eventually over time. Five types of AF are
classified based on the presentation, duration, and spontaneous termination of AF episodes
(Table 1.4). Currently, guidelines do not differentiate between types of AF and stroke
prophylaxis as observational studies suggest that stroke risk is dependent on concomitant
stroke risk factors regardless of AF type (73, 74). However, a recent meta-analysis (75)
32
suggests that the risk of stroke may differ between patients with paroxysmal AF and non-
paroxysmal AF. Twelve studies (10 RCTs and 2 prospective studies) involving almost 100,
000 patients evaluated the impact of AF type (PAF vs. non-PAF) on thromboembolic (TE)
events, bleeding and death. After adjustment for stroke risk factors (hypertension, heart failure,
age, gender, previous thromboembolism and diabetes) the hazard ratio for TE events was
1.38 (95% CI 1.19–1.61; p<0.001); bleeding events was 1.03 (95% CI 0.90–1.17; p=0.715)
and all-cause mortality was 1.22 (95% CI 1.09–1.37; p<0.001) respectively in non-PAF vs.
PAF patients (75). Results from this meta-analysis showed that thromboembolism and
mortality were significantly higher in non-PAF compared to PAF patients (75).
33
Table 1.4: Patterns of atrial fibrillation (taken directly from ESC guideline 2016)
AF pattern Definition
First diagnosed AF Undiagnosed AF before, irrespective of duration of
arrhythmia or presence and severity of AF-related
symptoms
Paroxysmal AF AF that self-terminates usually within 48 hours. Some
paroxysms can occur for up to 7 days OR AF episodes that
are cardioverted within 7 days
Persistent AF AF episodes lasting longer than 7 days, including episodes
terminated by cardioversion (pharmacological or direct
current cardioversion), after 7 days or more
Long-standing persistent AF Prolonged AF lasting for ≥1 year when a rhythm control
strategy is adopted
Permanent AF AF accepted by patients and physicians and rhythm control
strategy is not pursued. If rhythm control therapy is to be
adopted, AF should be classified as long-standing
persistent AF
1.2.5 Symptoms of atrial fibrillation

Some AF patients (25-40%) report no symptoms or very minimal symptoms while others (15-
30%) report severe disabling symptoms. Symptoms include palpitations, lethargy, shortness
of breath, chest tightness, sleeping difficulties and psychological distress. Symptomatic AF
patients tend to report poorer quality of life than asymptomatic patients (3). The modified
European Heart Rhythm Association (EHRA) symptom scale (Table 1.5) can be used to
categorise the severity of AF symptoms.
34
Table 1.5: Modified European Heart Rhythm Association symptom scale (3)
Modified
Symptoms Description
EHRA score
1 None No symptoms
2a Mild Symptoms related to AF are not affecting normal daily
activity
2b Moderate Patients are troubled by symptoms of AF but normal daily
activity is not affected
3 Severe Symptoms related to AF are affecting normal daily
activity
4 Disabling Discontinuation of normal daily activity
1.2.6 Risk of developing atrial fibrillation

Many concomitant conditions and cardiovascular diseases increase the risk of AF
development. Table 1.6 lists the most common concomitant disease associated with AF.
Identifying these risk factors and managing them is crucial to prevent AF and its burden of the
disease (3).
Table 1.6: Comorbid conditions/risk factors associated with AF (3)
Comorbidities/risk factors
• Genetic predisposition • Obstructive sleep apnoea
• Older age • Chronic kidney disease
• Hypertension • Smoking
• Heart failure • Alcohol consumption
• Valvular heart disease • Vigorous exercise
• Myocardial infarction
• Thyroid disorder
• Obesity
• Diabetes mellitus
• Chronic obstructive pulmonary
disease
35
The original Framingham Heart Study (76) (38 years follow up) investigated the predictors of
AF development has shown that men had 1.5 times higher risk of developing AF. Other risk
factors identified from the study include: hypertension (OR 1.5 men and OR 1.4 women),
congestive heart failure (OR 4.5 men and 5.9 women), myocardial infarction (OR 1.4 men),
valvular heart disease (OR 1.8 men and OR 3.4 women), aging and diabetes (OR 1.4 men
and 1.6 women) were all independent predictors of AF development (76). Other studies also
showed that obesity (HR 1.37) (77), thyroid dysfunction (78), COPD (FEV1<60%: RR 2.53)
(79), chronic kidney disease (stage 4/5: OR 3.52) (80), current smoker (RR 2.05) (81) and >21
drinks of alcohol per week (RR 1.39)(82) were also associated with AF development.
1.2.7 Treatment of atrial fibrillation
Essentially, there are five targets of treatment in AF patients: 1) acute rate and rhythm control
to achieve hemodynamic stability, 2) managing precipitating factors which involves lifestyle
changes and treating underlying conditions for the purpose or cardiovascular risk reduction,
3) assessing stroke risk and offering oral anticoagulants (OAC) therapy in patients with stroke
risk factors (this will be discussed in more detail in section 1.4), 4) assessing heart rate with
rate control therapy and lastly symptoms assessment with anti-arrhythmic drugs,
cardioversion or catheter ablation. These targets are set for the benefits of patients to improve
life expectancy, quality of life autonomy and social functioning (3).
1.2.7.1 Rate and rhythm control therapy
Essentially, rate control therapy is often offered to patients to improve AF-related symptoms.
Very little evidence exist that showed the best type and intensity of rate control therapy
compared to stroke prevention. Most data were derived from observational studies and short-
term cross over trials (83-86). Medical treatment options for rate control therapy for acute or
long-term rate control are beta blockers (bisoprolol, carvedilol), non- dihydropyridine calcium
36
channel blockers (diltiazem and verapamil), digoxin and some rhythm control agents which
have rate control properties, for example amiodarone, dronedarone and sotalol (3).
Another part of AF management involves restoring and maintaining sinus rhythm either with
pharmacological agents or with catheter ablation or in combination. At the moment, rhythm
control therapy is indicated in patients who are still symptomatic despite on adequate rate
control therapy (3). When anti-arrhythmic drugs are ineffective, electrical cardioversion is
usually offered in symptomatic AF patients. Evidences comparing rate and rhythm control
therapy versus rate control therapy alone have shown neutral outcomes (87-89).
Pharmacological rhythm control therapy options include amiodarone, flecanide and
propafenone. However, long term use of rhythm control therapy should be considered based
on the safety of each agent. For example, amiodarone could cause QT prolongation so it
should be avoided (if possible) with drugs causing the same effect. In addition, dronedarone
is contraindicated in patients with decompensated heart failure or patients with NYHA class
IV heart failure as these patients are prone to negative inotropic action and ventricular
proarrhythmic effects of anti-arrhythmic drugs (3).
37
1.2.8 Pathways to management of atrial fibrillation
1.2.8.1 The ABC pathway
Recently, ‘The Atrial Fibrillation Better Care (ABC)’ pathway was introduced for the
management of AF patients in an integrated manner (90) (Figure 1.3). The ‘A’ in the ABC
pathway stands for ‘Avoid’ stroke, which can be achieved by implementing the ‘Birmingham 3
step’ approach (Figure 1.3). The ‘B’ acronym stands for ‘Better management of symptoms’
including offering rate or rhythm control therapy and lastly ‘C’ stands for managing
cardiovascular and other Comorbidities. Simply, the ‘Birmingham 3 step approach’ involves:
1) Identifying low risk patients
2) Offering appropriate stroke prevention (OAC) to patients not in the ‘low risk group and
assessing bleeding risk (using the HAS-BLED score)
3) Deciding on OAC therapy either with a NOAC or VKA, emphasising anticoagulation
control (TTR ³70%) in those receiving VKA
1.2.8.2 National Institute for Health and Care Excellence (NICE) AF care
pathway and European Society of Cardiology (ESC) integrated pathway
Similarly, the NICE AF care pathway (91, 92) (Figure 1.4) and the ESC integrated pathway
(3) (Figure 1.5) suggest offering stroke prevention strategy to all AF patients at risk of stroke
and better management of AF symptoms with either rate or rhythm control strategy. The ABC
pathway (90) and the ESC integrated pathway (3) adds the ability to manage cardiovascular
risk factors and other comorbidities. The ESC integrated pathway necessitates changes in
lifestyle that will result in a better quality of life and improved life expectancy. All three
pathways were created to guide physicians to manage AF patients in a more integrated and
efficient manner.
38
Figure 1.3: ABC pathway [taken directly from Lip 2016 (90)]
NOAC: Non-vitamin K antagonist oral anticoagulant; OAC: oral anticoagulation; TTR: time in therapeutic range;
VKA: vitamin K antagonist
39
Assess stroke risk stratification Assess bleeding risk stratification
using CHA2DS2-VASc using HAS-BLED
Discuss risk and benefits of anticoagulation

Individual who decline treatment
Identify low risk patients i.e., CHA2DS2-VASc =0 (men) or 1 (women) Low risk No anticoagulation therapy
Anticoagulation contraindicated
CHA2DS2-VASc =1 (men) CHA2DS2-VASc ≥2 offer
consider anticoagulation anticoagulation
Discuss the options for anticoagulation with the person and base the choice
on their clinical features and preferences
Poor
control
VKA (assess anticoagulation
NOACs
control, TTR≥70%) NOACs
contraindicated
or not tolerated
Symptomatic Left atrial appendage occlusion
Symptomatic
Rate and rhythm control strategies Left atrial ablation strategies
Annual review for all patients
AF: atrial fibrillation; NOACs: non-Vitamin K antagonist oral anticoagulants; TTR: time in therapeutic range; VKA: vitamin K antagonist
Figure 1.4: NICE-AF care management pathway taken from NICE-AF guideline 2014 (91, 92)
40
Treatment Desired outcome Patient benefit
Acute rate and Hemodynamic stability

rhythm control
•Lifestyle changes,
Manage treatment of
precipitating CV risk reduction
underlying CV
factors conditions Improved life expectancy
•Oral anticoagulation in
Assess
patients at risk of Stroke prevention
stroke risk stroke Improved quality of life,
autonomy, social
Symptom improvement, functioning
Assess •Rate control preservation of LV
heart rate therapy function
Antiarrhythmic
Assess drugs, cardioversion, Symptom improvement
symptoms catheter ablation, AF
surgery
Figure 1.5: European Society of Cardiology integrated pathway adapted from the ESC AF Guidelines 2016 (3)
41
1.3 Psychological aspects in patients with atrial fibrillation and the
impact on quality of life
1.3.1 Depression and anxiety

Studies have reported psychological distress among patients with AF (93). Five studies (94-
98) have reported the prevalence of depression in AF cohorts which ranged from 10%-38%
measured using validated questionnaires [Beck Depression Inventory (BDI) (3-5) and Hospital
Anxiety and Depression Scale (HADS)] (94, 98). Meanwhile, the prevalence of state and trait
anxiety is reported as between 28-35% and 35%-38%, respectively, among AF patients, in
other studies, using the State-Trait Anxiety- Inventory (STAI-S) questionnaire (96, 99).
Recently, one Greek study of 170 permanent AF patients (94) found that depression and
anxiety were associated with patients’ age and number of years since diagnosis; those aged
>70 years who had the disease for 6-10 years were more anxious and depressed than those
aged <70 years who had AF for less than 6 years. Elderly AF patients are likely to be more
anxious and depressed perhaps due to various factors like cognitive impairment, physical
impairment, and poor treatment adherence (100, 101). In this study, poor/no knowledge of the
state of their health and having moderate/poor relationship with medical staff were also
associated with depression. Patients may be more depressed if they are unsure about the
disease, their prognosis and management. However, with adequate knowledge, patients tend
to have better control of their disease, present with fewer symptoms and are emotionally less
distressed (102, 103). In this study, females were also found to be more anxious compared to
males. This could be due to the differences in the type of AF and severity of symptoms and
outcomes of cardiac disease, as well as daily living or health behaviours (94, 104). For
example, females may be more prone to distress after facing an emotional trauma (example
divorce or death of family members) and perhaps this added stress would further cause
negative impact to their current comorbid condition.
42
Research has shown that quality of life (QoL) is poorer among AF patients with anxiety and
depression. Thrall et al (96) compared 101 AF and 97 hypertensive patients (age and sex
matched) and reported that AF patients were significantly more anxious (mean trait anxiety
score 37.4 vs. 33.3; p=0.02 in AF vs. hypertensive patients) than hypertensive patients at
baseline and this persisted 6 months later (mean trait anxiety score 36.9 vs. 32.6; p=0.03 in
AF vs. hypertensive patients). Symptoms of depression were also present in 38% of AF
patients at baseline and this also persisted at 6 months. Furthermore, symptoms of depression
and anxiety were independently related to health related quality of life (HRQoL) at baseline
but at 6 months, only symptoms of depression predicted HRQoL (96). In this study, AF patients
were more anxious and depressed than hypertensive patients probably because AF is an
illness associated with symptoms and an increased risk of stroke whereas hypertension is
without any specific symptoms and patients may have the perception of hypertension being a
benign condition. Studies (105-107) have shown that AF patients with greater severity of
symptoms (106, 107) and with recurrence of AF (108) have higher rates of depression and/or
anxiety.
Similarly in 2009, Lane et al (99) reported that lone AF patients presented with more anxiety
symptoms (38% state and 41% for trait anxiety) and had poor QoL, however, depression level
was low [median score 2.0 (0-3.0)] at baseline. In contrast to Thrall et al, (who did not
investigate the beliefs about medication)(96), Lane et al (99) reported that beliefs about
medication and number of AF symptoms predicted their physical quality of life.
1.3.2 Anticoagulation and quality of life
Despite the marked improvement in the prevention of stroke with the use of oral
anticoagulation therapy, studies (109, 110) have shown that the QoL among patients with VKA
therapy is affected as it requires behaviour and lifestyle modification.
Several studies (109, 111-113) from Brazil, Argentina and Spain among anticoagulated
patients (N=72 to 905) mainly for AF and mechanical heart valves, have assessed QoL using
43
the SF-36 tool. They reported overall SF-36 scores ranging from 54-62 indicating poor QoL.
These studies (109, 111-113) reported that patients have more limitations in daily activity
compared to emotional health evident by lower scores in physical domain compared to
emotional domains. This indicates that anticoagulated patients appear to have greater
impairment in their physical health rather than emotional health.
However, other studies (109, 111, 114, 115) from Brazil, Turkey, Malaysia and Spain (N=72
to 339) among mostly anticoagulated AF patients using VKA showed a relatively better QoL
with scores ranging from 67-86 with the Duke Anticoagulation Satisfaction Scale (DASS) (109,
111-113). Three studies (109, 111, 114) showed that patients were more limited and had
higher burden in their physical activity, however, only one study (115) showed patients were
more psychologically affected by their anticoagulation therapy. Although the studies differed
methodologically (small sample size, cross-sectional study), they suggest that anticoagulated
patients are more likely to be affected by treatment inconvenience rather than by their
emotional health.
Another Turkish study (116) compared HRQoL using the SF-36 tool among 182 NOAC and
warfarin patients (91 patients in each group). They showed that even after adjustment for age,
gender, adherence, and duration of OAC therapy, warfarin treated patients had significantly
lower HRQoL scores in all domains (p<0.05); self-reported symptoms of anxiety (HADS-A: 6.2
vs. 4.6; p<0.001) and depression (HADS-D: 4.9 vs. 3.6; p<0.001) were significantly higher
among warfarin treated patients compared to NOAC patients respectively. (116) This may be
expected given that NOACs have fewer drug interactions, no known food interactions and do
not require frequent INR monitoring, which have all been shown to impact patient’s QoL (109,
111, 114, 115).
The next section discusses stroke prevention in AF including identifying stroke risk factors,
different treatment strategies (VKA vs. NOACs) and guideline recommendations.
44
1.4 Stroke prevention in atrial fibrillation
1.4.1 Antithrombotic therapy in stroke prevention in atrial fibrillation
Currently, there are five types of OAC therapy available for stroke prevention in AF patients,
including Vitamin K antagonists (VKA e.g., warfarin) and non-VKA oral anticoagulants
(NOACs e.g., dabigatran, rivaroxaban, apixaban and edoxaban) (3).
1.4.1.1 Vitamin K antagonist (VKAs)
The principal priority in managing patients with AF is stroke prevention. Warfarin and other
types of VKAs (for example acenocoumarol, phenprocoumon) were among the first
anticoagulants used in AF patients. Clinical trials (117) have shown that compared to placebo,
dose-adjusted VKA reduces the risk of stroke and systemic embolism by 64% (95% CI 49-74)
(Table 1.7) and all-cause mortality by 26% (95% CI 3-43) (117). Compared to antiplatelet
therapy, dose-adjusted VKA was also more efficacious in a reduction of TE complication
[relative risk reduction 39% (95% CI 22-52) from 12 trials, 12, 963 participants] (117).
However, VKAs have a narrow therapeutic index requiring dose adjustments and frequent
monitoring of the international normalised ratio (INR) to achieve maximum therapeutic effect
and minimise harm. Furthermore, individual response to VKA can be influenced by many
factors (see section 1.5, pages 78-100 for more detail) (3). Nonetheless, VKA is still the
recommended OAC of choice in AF patients with severe renal impairment (CrCl<15ml/min)
and patients with valvular heart disease requiring mechanical valve prosthesis (3). When VKA
is used, attention needs to be given on the quality of anticoagulation reflected by the time in
therapeutic range (TTR) of the INR (3). To maximise effectiveness and safety of VKA, the
European guidelines (3) have recommended a TTR of ≥70% while the NICE guideline (118)
recommended TTR of ≥65%. Several studies (119-122) have also shown that TTR
independently predict thromboembolic (120, 122), bleeding events (120, 121) and improves
survival (119, 122) in AF patients prescribed with VKA therapy.
45
Table 1.7: Adjusted-dose warfarin versus placebo or no treatment, taken directly from
Hart 2007 (117)
Study, year Patients, Target INR Strokes/patients, n/n Relative risk

N a. warfarin reduction
b. placebo or control (95% CI), %
AFASAK I, 671 2.8-4.2 a. 9/335 54 *

1989(123), b. 19/336
1990(124)
BAATAF, 1990(125) 420 1.5-2.7 a. 3/212 78*
b. 13/208
SPAF 1, 1991(126) 421 2.0-4.5 a. 8/210 60*
b. 19/211
CAFA, 1991(127) 378 2.0-3.0 a. 6/187 33*
b. 9/191
SPINAF, 1992(128) 571 1.4-2.8 a. 7/281 70*
b. 23/290
EAFT, 1993(129) 439 2.5-4.0 a. 20/225 68*
b. 50/214
6 trials 2900 - a. 53/1450 64 (49-74)

b. 133/1450
INR: international normalized ratio; * 95% CI not stated in the Hart 2007 meta-analysis (117)
1.4.1.2 Non-vitamin K antagonist oral anticoagulants (NOACs)
With the emergence of non-VKA oral anticoagulants (NOACs), the practice of antithrombotic
management has shifted toward prescribing NOACs for stroke prevention in patients with AF
given the relative efficacy, safety and convenience of NOACs compared to VKAs (26). Four
NOACs including the direct thrombin inhibitor, dabigatran and factor Xa inhibitors rivaroxaban,
apixaban and edoxaban are suitable alternatives to VKAs for stroke prevention in AF. Table
1.8 list the different characteristics of warfarin and the four NOACs and Table 1.9 summaries
the baseline characteristics of the four major NOAC trials. NOACs have a more targeted mode
of action, shorter time to reach the maximum anticoagulant effect, shorter half-life, less drug-
drug interactions and no food restrictions compared to warfarin. Tables 1.10 and 1.11 present
the efficacy and safety results, respectively, from the four NOAC trials. Compared to warfarin,
46
dabigatran 150mg twice daily [RR 0.66 (95% CI 0.53-0.82)] and apixaban 5mg twice daily [RR
0.79 (95% CI 0.66-0.95)] are associated with a lower risk of ischemic stroke and systemic
embolism. All four NOACs significantly reduced the risk of haemorrhagic stroke [dabigatran
150 mg RR 0.26 (95% CI 0.14-0.49); rivaroxaban 20 mg RR 0.59 (95% CI 0.37-0.93);
apixaban 5 mg RR 0.51 (95% CI 0.35-0.75); edoxaban 60 mg RR 0.54 (95% CI 0.38-0.77)]
and intracranial haemorrhage [dabigatran 150 mg RR 0.40 (95% CI 0.27-0.60); rivaroxaban
20 mg RR 0.67 (95% CI 0.47-0.93); apixaban 5 mg RR 0.42 (95% CI 0.30-0.58); edoxaban
60 mg RR 0.47 (95% CI 0.34-0.63)] compared to warfarin. A meta-analysis (3) of the four
major Phase 3 NOAC trials have also concluded that compared to warfarin, NOACs
significantly reduced the risk of stroke and systemic embolism by 19% [RR 0.81 (95% CI 0.73-
0.91; p<0.0001)], and this is largely due to a significant reduction in haemorrhagic stroke [RR
0.49 (0.38-0.64; p<0.0001)] (130). NOACs also significantly reduce intracranial haemorrhage
[RR 0.48 (0.39-0.59; p<0.0001)] and all-cause death [RR 0.90 (0.85-0.950; p=0.0003)] but
were also associated with increased risk of GI bleeding [RR 1.25 (1.01-1.55); p=0.04)] when
compared to warfarin (3).
47
Table 1.8: Pharmacokinetics of warfarin versus NOACs and baseline characteristics for four randomised controlled trial cohorts comparing
warfarin versus NOACs
Warfarin (131) Dabigatran (132) Rivaroxaban (133) Apixaban (134) Edoxaban (135)
Mechanism of Interfere with synthesis Direct thrombin inhibitor Activated factor Xa inhibitor Activated factor Xa Activated factor Xa
action of vitamin K dependent inhibitor inhibitor
clotting factors by
inhibiting VKORC1
Bioavailability, % 98 3-7 66 without food 50 62
100% with food
Time to reach peak 1.5-3 days 3 2-4 3 1-2
level, hours
Half-life, hours 20-60 12-17 5-13 9-14 10-14
Clearance Renal 80% renal 66% liver, 33% renal 27% renal 50%
Dosing in AF Variable dose, once 150mg or 110mg twice 20mg or 15mg once daily 5mg or 2.5mg twice daily 60mg or 30mg or 15mg
daily daily once daily
Not recommended if - CrCl<30ml/min CrCl<15ml/min CrCl<15ml/min CrCl<15ml/min
Antidote Vitamin K Idarucizumab Not available Not available Not available
Table 1.9: Baseline characteristics of the four randomised controlled trial cohorts comparing warfarin versus NOACs
Dabigatran (132) Rivaroxaban (133) Apixaban (134) Edoxaban (135)

Patients, N 18,113 14,264 18,201 21,105
Mean/median age 71.5 (8.7) 73 (65-78) 70 (63-76) 72 (64-78)
Male, % 63.6 60.3 64.5 61.9
Follow up, years 2 1.9 1.8 2.8
CHADS2 score (mean) 2.1 3.5 2.1 2.8
Median individualised TTR 67 (54-78) 58 (43-71) 66 (52-77) 68 (57-77)
AF: atrial fibrillation; CrCl: creatinine clearance; CHADS score: 1 point each for congestive heart failure, hypertension, age >75 years, diabetes mellitus, and 2 points for stroke/TIA; TTR: time in therapeutic range;
VKORC1: vitamin K epoxide reductase
48
Table 1.10: Efficacy outcomes in the four major randomised controlled trials comparing warfarin versus NOACs in AF populations
Dabigatran (132) Rivaroxaban (133) Apixaban (134) Edoxaban (135)

(RE-LY) (ROCKET-AF) (ARISTOTLE) (ENGAGE AF-TIMI 48)
Efficacy Outcomes
Dabigatran Warfarin, Rivaroxaban Warfarin Apixaban Warfarin, Edoxaban Warfarin,
150mg 20mg 5mg 60mg,
N=6076 N=6022 N=9120† N=9081† N=7035† N=7036†
Stroke/ systemic 134 (1.11) 199 (1.69) 269/7081† 306/7090† 212 (1.27) 265 (1.60) 296 (1.57) 337 (1.80)
embolism, N (%/yr) (2.1) (2.4)
Relative risk (95% CI) 0.66 (0.53-0.82); p<0.001 0.88 (0.75-1.03); p=0.12* 0.79 (0.66-0.95); p=0.01* 0.87 (0.73-1.04)§; p=0.08
Ischemic stroke, N 111 (0.92) 142 (1.20) 149/7061‡ 161/7082‡ 162 (0.97) 175 (1.05) 236 (1.25) 235 (1.25)
(%/yr) (1.34) (1.42)
Relative risk (95% CI) 0.76 (0.60-0.98); p=0.03* 0.94 (0.75-1.17); p=0.581* 0.92 (0.74-1.13); p=0.42 1.00 (0.83-1.19); p=0.97
‡ ‡
Haemorrhagic stroke, N 12 (0.10) 45 (0.38) 29/7061 50/7082 40 (0.24) 78 (0.47) 49 (0.26) 90 (0.47)
(%/yr) (0.26) (0.44)
Relative risk (95% CI) 0.26 (0.14-0.49); p<0.001* 0.59 (0.37-0.93); p=0.0248 0.51 (0.35-0.75); p=<0.001* 0.54 (0.38-0.77); p<0.001
All-cause mortality, N 438 (3.64) 487 (4.13) 208/7061‡ 250/7082‡ 603 (3.52) 669 (3.94) 773 (3.99) 839 (4.35)
(%/yr) (1.87) (3.53)
Relative risk (95% CI) 0.88 (0.77-1.00); p=0.051* 0.85 (0.70-1.02); p=0.073* 0.89 (0.80-0.998); p=0.047* 0.92 (0.83-1.01); p=0.08
* p for superiority; † based on intention to treat population; ‡based on safety on-treatment population; §97.5%CI was used
49
Table 1.11: Safety outcomes in the four major randomised controlled trials comparing warfarin versus NOACs in AF populations
Dabigatran Rivaroxaban Apixaban Edoxaban (135)

(RE-LY) (132) (ROCKET-AF) (133) (ARISTOTLE) (134) (ENGAGE AF-TIMI 48)
Safety outcomes
Dabigatran Warfarin, Rivaroxaban Warfarin, Apixaban Warfarin, Edoxaban 60, Warfarin,
150, 20, 5,
N=6076 N=6022 N=7111 N=7125 N=9088 N=9052 N=7012 N=7012
Major bleeding N (%/yr) 375 (3.11) 397 (3.36) 395 (3.6) 386 (3.4) 327 (2.13) 462 (3.09) 418 (2.75) 524 (3.43)
Relative risk (95% CI) 0.93 (0.81-1.07); p=0.31 1.04 (0.90-1.20); p=0.58 0.69 (0.60-0.80); p<0.001 0.80 (0.71-0.91); p<0.001
Intracranial 36 (0.30) 87 (0.74) 55 (0.5) 84 (0.7) 52 (0.33) 122 (0.80) 61 (0.39) 132 (0.85)
haemorrhage, N (%/yr)
Relative risk (95% CI) 0.40 (0.27-0.60); p<0.001 0.67 (0.47-0.93); p=0.02 0.42 (0.30-0.58); p<0.001 0.47 (0.34-0.63); p<0.001
GI bleed, N (%/yr) 182 (1.51) 120 (1.02) 224 (3.15%) ¶ 154 (2.16%)¶ 105 (0.76) 119 (0.86) 232 (1.51) 190 (1.23)
Relative risk (95% CI) 1.50 (1.19-1.89); p<0.001 - 0.89 (0.70-1.15); p=0.37 1.23 (1.02-1.50); p=0.03
Net clinical outcome, N 832 (6.91)† 901 (7.64)† - - 1009 (6.13)‡ 1168 (7.20)‡ 1323 (7.26)§ 1462(8.11)§
(%/yr)
Relative risk (95% CI) 0.91 (0.82-1.00); p=0.04 - 0.85 (0.78-0.92); p<0.001 0.89 (0.83-0.96); p=0.003
†composite of stroke, systemic embolism, pulmonary embolism, myocardial infarction, death or major bleeding, ‡ composite of stroke, systemic embolism, major bleeding or all-cause death; §composite of stroke, systemic
embolic event, major bleed, all-cause death;¶ percentage not %/yr and p<0.001
50
Another meta-analysis of ‘real-world’ high quality studies (28 studies included only from
nationwide or health insurance claims database with adjusted or matched findings) of
dabigatran, rivaroxaban and apixaban compared to warfarin for stroke prevention in AF has
produced findings consistent with the RCTs (Table 1.12) (136). Dabigatran, rivaroxaban and
apixaban all reduced the risk of intracranial bleed compared to warfarin, with similar risk of
ischemic stroke and ischemic stroke/systemic embolism. Mortality risk was lower with
dabigatran and apixaban. There was less gastrointestinal (GI) bleeding and major bleeds with
apixaban but higher risk of GI bleed with dabigatran and rivaroxaban, and similar risk of
myocardial infarction with dabigatran and rivaroxaban compared to warfarin (Table 1.12).
(136) Edoxaban was not included in this meta-analysis as no real world studies assessing
edoxaban was published at the time of the searches.
51
Table 1.12: Meta-analysis of real word studies comparing dabigatran, rivaroxaban and apixaban to warfarin for efficacy and safety outcomes [from
Ntaios et al(136)]
Efficacy outcomes Dabigatran vs. warfarin Rivaroxaban vs. warfarin Apixaban vs. warfarin
Ischemic stroke (HR, 0.96; 95% CI, 0.80–1.16) (HR, 0.89; 95% CI, 0.76–1.04) (HR, 0.95; 95% CI, 0.75–1.19)
Number of No of patients 9 476, 924 5 108, 810 3 48, 549
studies
Ischemic stroke/systemic (HR, 1.17; 95% CI, 0.92–1.50) (HR, 0.73; 95% CI, 0.52–1.04) (HR, 1.07; 95% CI, 0.87–1.31)
embolism
studies
Any stroke/systemic embolism (HR, 0.93; 95% CI, 0.77–1.14) HR, 0.87; 95% CI, 0.71–1.07 (HR, 0.67; 95% CI, 0.46–0.98)
studies
Myocardial infarction (HR, 0.96; 95% CI, 0.77–1.21) (HR, 1.02; 95% CI, 0.54–1.89) -
Number of No of patients 5 316, 180 2 24, 621 - -
studies
Safety outcomes
Intracranial haemorrhage (HR, 0.42; 95% CI, 0.37–0.49) (HR, 0.64; 95% CI, 0.47–0.86) (HR, 0.45; 95% CI, 0.31–0.63)
studies
Gastrointestinal bleed (HR, 1.20; 95% CI, 1.06–1.36) (HR, 1.24; 95% CI, 1.08–1.41) (HR, 0.63; 95% CI, 0.42–0.95)
studies
Major bleed (HR, 0.83; 95% CI, 0.65–1.05) (HR, 1.00; 95% CI, 0.92–1.08) (HR, 0.55; 95% CI,0.48–0.63)
studies
Death (HR, 0.63; 95% CI, 0.52–0.76) (HR, 0.67; 95% CI, 0.35–1.30) (HR, 0.65; 95% CI, 0.56–0.75)
studies
52
1.4.2 Assessing risk factors for stroke
Assessing risk factors for stroke in AF patients is essential before starting anticoagulation
treatment. Stroke risk factors related to AF have been reported in various systematic reviews
(137, 138). History of stroke/TIA [RR 2.5 (95%CI 1.8-3.5)], increasing age [RR 1.5 per decade;
(95% CI 1.3-1.7)], history of hypertension [RR 2.0 (95% CI 1.6-2.5)] and diabetes mellitus [RR
1.7 (95%CI 1.4-2.0)] were the most consistent risk factors for stroke as reported by the Stroke
in AF Working group study (137). Female sex, vascular disease and heart failure have been
less consistently associated with stroke risk in AF. One systematic review and meta-analysis
reported an increased risk of strokes among females with AF [RR 1.31 (95%CI 1.18-1.46)]
(139), while in another non-anticoagulated AF population, only females ≥75 years were at
increased the risk of stroke [HR 1.20 (95%CI 1.12-1.28)] (140). Stroke risk in women is age-
dependent where women age <65 with no other stroke risk factors are classified as being at
low risk of stroke (141). Recently, another nationwide registry showed that female and male
(with no additional risk factor) AF patients have similar thromboembolic risk. In addition, female
gender acts as a ‘risk modifier’ rather than a risk factor whereby risk of TE is only higher in
females with ≥2 non-gender stroke related risk factors with the acronym CHA2DS2-VA (heart
failure, hypertension, aged 75 years, diabetes, stroke, vascular disease and age 65-74).
Furthermore, in this registry, the risk of stroke differs by different score categories; the risk
increases significantly with a score ≥2 (but not 3) thus modifying the stroke risk. [CHA2DS2-
VA score 2: HR 1.21 (1.08-1.34); score 3: HR 1.03 (0.93-1.13); score 4: HR 1.25 (1.14-1.36);
score 5: HR 1.41 (1.27-1.56)] (142).
In terms of vascular disease, the OPTIMAL trial showed increased risk of stroke [adjusted HR
of 14.6 (95% CI 5.87-36.3)] among new-onset AF patients 30 days after acute myocardial
infarction (143), while another Danish study also reported increased risk of stroke among
patients with peripheral arterial disease [HR 1.93 (95% CI 1.70-2.19)] (144). Left ventricular
dysfunction (moderate to severe) was also seen as risk factor for stroke (RR 2.5 95% CI 1.5-
4.4) in one study (92) while another study showed no difference in the risk of stroke or
53
thromboembolism between different categories of ejection fraction (EF) [HR 0.75 (95% CI
0.44-1.30) for EF <35% and HR 1.25 (95% CI 0.83-1.93) for EF 35-49% (145).
1.4.3 Predicting stroke risk with clinical risk scores
Various clinical risk scores to predict stroke, TIA or thromboembolism (TE) in AF patients have
been developed since the 1990’s (Table 1.13). The risk scores were developed based on
common demographic and clinical factors found in AF patients (141). Table 1.13 presents
eleven risk scores to predict stroke in AF population. The most common risk factors present
in these risk scores are prior thromboembolic events and diabetes mellitus (present in all
eleven risk scores). Increasing age and hypertension were present in ten of the risk scores
followed by female sex and heart failure (in 6 scores), and vascular disease (5 scores).
1.4.3.1 CHA2DS2-VASc score
Among all the risk scores mentioned in Table 1.13, only the CHA2DS2-VASc score will be
discussed in more detail as it is the most widely validated risk score and is recommended by
most major AF clinical guidelines (see next section). The CHA2DS2-VASc score (Table 1.14)
was developed in 2010 (146), almost a decade after the development of CHADS2 score.
Compared to the CHADS2 score, it included three additional stroke risk factors: female sex,
age 65 to 74 and vascular disease. In addition, the CHA2DS2-VASc score more clearly defined
the ‘congestive heart failure’ factor to include those with moderate to severe LV dysfunction
(ejection fraction<40%), recent decompensation of heart failure either with preserved or
reduced ejection fraction. Furthermore, 2 points was awarded for those aged ≥75 years (146)
as increasing age is a strong predictor of stroke (137).
The CHA2DS2-VASc score was first validated in the Euro Heart Survey cohort of 1084 non-
valvular, non-anticoagulated AF patients (146). In this study, increasing rates of TE can be
seen as the CHA2DS2-VASc score increases (Table 1.15). The same trend was also seen
when it was validated in the SPORTIF III and IV cohort (147) but lower TE rates from the
54
anticoagulated cohort were evident compared to the non-anticoagulated cohort. Slight
improvement in the ability to predict stroke with the CHA2DS2-VASc score compared to the
CHADS2 score were shown in both studies [C index 0.60 vs. 0.56 in Euro Heart Survey cohort
(146) and C index 0.65 vs. 0.64 for in the SPORTIF III and IV cohort (147) for CHA2DS2-VASc
and CHADS2 respectively].
The CHA2DS2-VASc score was further validated in several other cohorts (118, 144, 148, 149)
including in non-Western cohorts (150, 151). Compared to the CHADS2 score, the CHA2DS2-
VASc score was able to predict stroke events in two Chinese studies [C index: 0.53 (150) and
0.72, respectively] (151).
55
Table 1.13: Risk factors in stroke risk stratification schemes, updated from Lip et al 2015 (141)
Risk factors Other factors
Scores Age, y Female sex Prior TE Hypertension Heart failure Diabetes Vascular
event mellitus disease
AFI, 1994(152) 65-75, >75 ü ü ü
SPAF, 1999(153) >75† ü† ü ü ü
CHADS2, 2001(154) ≥75 ü ü ü ü
Framingham, 2003 (155) ü ü ü ü ü
vanWalraven, 2003(156) ü ü ü ü
Rietbrock, 2008 (157) ü ü ü ü
CHA2DS2-VASc, 65-74, ≥75 ü ü ü ü ü ü
2010(146)
QStroke, 2013(158) ü ü ü ü ü ü ü Many other‡
ATRIA, 2013(159) ü ü ü ü ü ü ü Proteinuria, eGFR
CHADS2, 2013, ≥65 ü ü ü ü ü Cardiomyopathy
Japan(160)
CHADS65, 2014 (161) ≥65 ü ü ü ü
eGFR: estimated glomerular filtration rate; TE: thromboembolism; vascular disease: prior myocardial infarction, aortic plaque or peripheral arterial disease
†age and female sex is combined as a single factor; ‡ ethnicity (self-assigned), smoking status, ratio of total serum cholesterol to high density lipoprotein, cholesterol concentrations, body mass index,
family history of coronary heart disease in first degree relative under 60 years, Townsend deprivation score, treated hypertension, diabetes type I and II, renal disease, rheumatoid arthritis, valvular
heart disease and atrial fibrillation
56
Table 1.14: CHA2DS2-VASc score components
CHA2DS2-VASc components
CHF or LVEF ≤40% 1
Hypertension 1
Age ≥75 years 2
Diabetes 1
Stroke/TIA/ thromboembolism 2
Vascular Disease* 1
Age 65-74 1
Female sex 1
Total score: 9
Low risk: 0 male
1 female
High risk: ≥ 1 male
≥2 female
LVEF: left ventricular ejection fraction*prior myocardial infarction, peripheral artery disease or aortic plaque
Table 1.15: CHA2DS2-VASc score and thromboembolic risk taken directly from Lip et
al from the European Heart survey (146) and SPORTIF III and IV trial (147)
European Heart survey cohort (146) SPORTIF III and IV cohort (147)
CHA2DS2-
VASc score TE rate at 1 year Adjusted 1 year TE rate at 1 year Adjusted 1 year
(95% CI) TE rate, %# (95% CI) TE rate, %†
0 0 0 0 0
1 0.6 (0.0-3.4) 0.7 0.46 (0.10-1.34) 1.3
2 1.6 (0.3-4.7) 1.9 0.78 (0.44-1.29) 2.2
3 3.9 (1.7-7.6) 4.7 1.16 (0.79-1.64) 3.2
4 1.9 (0.5-4.9) 2.3 1.43 (1.01-1.95) 4.0
5 3.2 (0.7-9.0) 3.9 2.42 (1.75-3.26) 6.7
6 3.6 (0.4-12.3) 4.5 3.54 (2.49-4.87) 9.8
7 8.0 (1.0-26.0) 10.1 3.44 (1.94-5.62) 9.6
8 11.1 (0.3-48.3) 14.2 2.41 (0.53-6.88) 6.7
9 100 (2.5-100) 100 5.47 (0.91-27.0) 15.2
TE: thromboembolic events # Theoretical TE rates without therapy: corrected for the % of patients receiving aspirin within each
group, assuming that aspirin provides a 22% reduction in TE risk, based on Hart et al. (1) †Theoretical TE rates without therapy:
assuming that warfarin provides a 64% reduction in TE risk, based on Hart et al. (1); CI indicates confidence interval.
57
All of the ‘clinical factor’ based risk scores have modest predictive ability (C-index
approximately 0.6) for identifying ‘high risk’ groups thus addition of biomarkers (such as D-
dimer, natriuretic peptides, von Willebrand factor) has been shown to improve the predictive
ability of identifying the high-risk group (141, 162-164). However, despite addition of several
biomarkers, only slight improvement in the predictive abilities of the scores can be seen
changing the C-index to 0.65-0.70 (162-164). Measurement of biomarkers results in additional
cost and loss of simplicity in risk score calculation making such scores less easy to use in
everyday clinical practice (141).
1.4.4 Guidelines and recommendations for stroke risk stratification and
antithrombotic therapy
Previously, the focus of the older risk scores was to divide patients into low risk, moderate risk
and high risk of stroke. However, evidence has shown that identifying ‘high risk’ patients leads
to under treatment with OAC in these groups (165, 166). Thus, the focus now has shifted
towards identifying ‘low risk’ patients (CHA2DS2-VASc =0 in males or 1 in females). Many
studies (167-169) have demonstrated that the CHA2DS2-VASc score is best at identifying the
‘truly low risk’ patients for whom the risk of stroke or systemic embolism is <1% per year. Due
to this, the latest European Society of Cardiology (ESC)(3), the National Institute for Health
and Care Excellence (NICE) (118), American (170), Australian (150) and Asia Pacific (171)
guidelines (Table 1.16) have recommended using the CHA2DS2-VASc score to stratify stroke
risk. The Canadian Cardiovascular Society (CCS) 2016 (172) recommends the modified
CHADS65 score, although it acknowledges that other risk factors (also present in CHA2DS2-
VASc score) such as age >65, prior myocardial infarction, aortic plaque and peripheral arterial
disease.
All of the latest AF guidelines recommend OAC, with either a NOAC or well-controlled VKA
(TTR ≥70%), for all AF patients who are not deemed ‘low risk’, with preference for a NOAC in
most guidelines (1, 28-30). The recommendation for no antithrombotic therapy for low risk
58
patients and use of OAC in high-risk patients (those with ≥2 stroke risk factors) is consistent
in all guidelines (except in the Australian guideline (150), where the recommendation is no
antithrombotic or aspirin in the low risk group). However, AF guidelines (3, 150, 170, 173)
have conflicting recommendations for patients in the intermediate risk groups with a single
stroke risk factor where some (3, 173) considered OAC based on patient preferences and
others (150, 170) with the recommendation of ‘No antithrombotic therapy or treatment with
OAC or aspirin may be considered’.
59
Table 1.16: Guideline recommendations for stroke prevention in AF
Risk score Risk categories (scores) Recommendations

APHRS 2017(171) CHA2DS2-VASc Asia pacific CHA2DS2-VASc =0 or 1 in female No antithrombotic therapy
CHA2DS2-VASc=1 in male NOAC preferred (D, R, A, E) or well
controlled VKA
CHA2DS2-VASc ≥2 NOAC preferred (D, R, A, E)
ESC 2016 (3) CHA2DS2-VASc European CHA2DS2-VASc =0 or 1 in female No antithrombotic therapy

CHA2DS2-VASc = 1 in male and 2 in OAC should be considered depending on
female individual characteristics or patient
preference
CHA2DS2-VASc ≥2 in male ≥3 in female NOAC (preferred) or well controlled VKA
CCS 2016 (172) CHADS65 Canada Age <65 without CHADS2 risk factors, No antithrombotic therapy
CAD/ coronary, aortic or peripheral
vascular disease
Age <65, no CHADS2 risk factor but have OAC- NOAC preferred over to VKA
CAD/coronary, aortic or peripheral
vascular disease
Age <65 and one of CHADS2 risk factors OAC- NOAC preferred over to VKA
-Heart failure
-Hypertension
-Diabetes Mellitus
-Stroke/TIA/peripheral embolism
Age ≥65 OAC- NOAC preferred over to VKA
60
Table 1.16 continued
NICE 2014 (118) CHA2DS2-VASc United Low risk No antithrombotic therapy

Kingdom (0 male)
(1 female)
High risk Well controlled VKA or NOAC
(1 male)
(≥2 female)
2014 AHA/ACC/HRS CHA2DS2-VASc America CHA2DS2-VASc =0 No antithrombotic therapy

Guideline
CHA2DS2-VASc score =1 No antithrombotic therapy or treatment with
for the Management
oral anticoagulant or aspirin may be
of Patients
considered
With Atrial Fibrillation
With prior stroke, TIA, or CHA2DS2-VASc VKA, D, R, A
(170)
score ≥2
CHA2DS2-VASc score ≥2 and end stage VKA

CKD (CrCL<15ml/min) or on
haemodialysis
CHA2DS2-VASc score ≥2 moderate to Reduced doses of direct thrombin or factor

severe CKD Xa inhibitors may be considered
CCS 2014 (161) CHADS65 Canada Age <65 without CHADS2 risk factors, No antithrombotic therapy
CAD/ coronary, aortic or peripheral
vascular disease
Age <65, no CHADS2 risk factor but have ASA

CAD/coronary, aortic or peripheral
vascular disease
61
Age <65 and one of CHADS2 risk factors OAC- NOAC preferred over to VKA
-Heart failure
-Hypertension
-Diabetes Mellitus
-Stroke/TIA/peripheral embolism
Age ≥65 OAC- NOAC preferred over to VKA
Atrial Fibrillation CHA2DS2-VASc Australia Age <65 without CHADS2 risk factors, No antithrombotic therapy or aspirin only
Information for the CAD/ coronary, aortic or peripheral
Health Practitioner vascular disease
(2014)(150) CHA2DS2-VASc=1 Evidence of treatment is limited but options
1 clinically relevant non-major risk factor include:
-No antithrombotic treatment, aspirin 75-
300mg daily or OAC. Aspirin and OAC is
unlikely to have a net clinical benefit unless
HAS-BLED score is low
CHA2DS2-VASc ≥2 -New OAC is preferred to warfarin (target
1 ‘major’ risk factor or ≥ 2 clinically INR 2.5)
relevant non-major risk factor
APHRS 2013 (174) CHA2DS2-VASc Asia pacific CHA2DS2-VASc =0 No antithrombotic therapy

CHA2DS2-VASc 1 NOAC (D/A)
W/R (alternative)
CHA2DS2-VASc ≥2 OAC (D/R/A/W)
ESC 2012 (173) CHA2DS2-VASc European CHA2DS2-VASc =0 No antithrombotic therapy
62
CHA2DS2-VASc= 1 Well controlled VKA or NOAC
OAC should be considered, based upon an
assessment of the risk of bleeding
complications and patient preferences.
CHA2DS2-VASc ≥2 Well controlled VKA or NOAC
When patient refuse any form of OAC,
antiplatelet therapy with combination of ASA
and clopidogrel or less effectively ASA
ESC 2010(173) CHADS2 score European Low risk ASA or no antithrombotic
-No risk factor (<65 with lone AF)
Intermediate risk Adjusted dose VKA (target INR 2.5) OR ASA

1 ‘clinically relevant non-major’ 75-325 mg daily
-Heart failure (moderate to severe LV
systolic dysfunction, LVEF <40%)
-Hypertension and/or diabetes mellitus
-Female sex and/or age 65-74 years
-Vascular disease
High risk (CHADS2≥2) OR 1 ‘major’ or VKA with target INR 2.5

≥2 ‘clinically relevant non-major’
Major risk factor

-Previous stroke, TIA, systemic embolism
-Age ≥75
63
NICE 2006 (175) NICE United Low risk ASA 75-300 mg daily
Kingdom Age <65 with no moderate or high-risk
factors
Moderate risk Consider anticoagulation or ASA 75-300mg
-Age ≥65 with no high-risk factors daily
-Age <75 with hypertension, diabetes,
vascular disease
High risk VKA with target INR 2.5
-Previous thromboembolic event
-Age ≥75 with hypertension, diabetes,
vascular disease
-Clinical evidence of valve disease, heart
failure or impaired LV function
ACC/AHA/ESC 2006 America No risk factors ASA 81 to 325 mg daily
Guidelines for the One moderate-risk factor ASA 81 to 325 mg daily, or warfarin (INR 2.0
Management of -Age ≥75 to 3.0, target 2.5)
Patients With Atrial -hypertension
Fibrillation(58) -heart failure (EF ≥35%)
-Diabetes mellitus
Any high-risk factor or more than 1 VKA (target INR 2.5)
moderate-risk factor
High risk factors
Previous stroke. TIA, embolism
Prosthetic heart valves
AF: atrial fibrillation; ASA: Aspirin; A: apixaban; D: dabigatran; E: edoxaban; R: rivaroxaban; CAD: coronary artery disease; CHADS score: 1 point each for congestive heart failure, hypertension, age
>75 years, diabetes mellitus, and 2 points for stroke/TIA; CHA2DS2-VASc score: 1 point each congestive heart failure/left ventricular ejection fraction ≤40%, hypertension, age 65-74, diabetes mellitus,
vascular disease and 2 points each for age>75 years and prior stroke/TIA; CrCl: creatinine clearance; EF: ejection fraction; INR: international normalised ratio; HAS-BLED: 1 point each for uncontrolled
hypertension, abnormal renal or liver function, stroke, prior bleeding labile INR, age >65, interacting drugs and alcohol excess; LV: left ventricular; NOAC: non-vitamin K antagonist; TIA: transient
ischemic attack; TTR: time in therapeutic range; VKA: vitamin K antagonist
64
1.4.5 Assessing bleeding risk in atrial fibrillation
This section will summarise bleeding risk and the various bleeding risk scores developed for
AF. The review has been published in the American Journal of Cardiology (176).
Assessing bleeding risk is also important before prescribing OAC therapy to AF patients. Over
the last decade, several risk scores have been proposed to predict bleeding events in AF
patients on anticoagulant therapy. These scores have been tested and validated worldwide in
many cohorts of AF (177-185) to support physicians in assessing bleeding risks (186).
Recently, the ESC Guidelines on the management of AF 2016 has summarised bleeding risks
(into modifiable and non-modifiable) and encouraged prompt attention to common modifiable
bleeding risk. This includes hypertension (especially when systolic blood pressure is >160
mmHg), labile INR or TTR <60% (in patients on VKA), medications predisposing to bleeding
such as antiplatelet and non-steroidal anti-inflammatory drugs and lastly excessive alcohol
consumption (≥8 drinks/week) (3, 141).
1.4.5.1 Risk factors of bleeding
Many risk factors of bleeding are included in the various bleeding scores shown in Table 1.17.
The number of risk factors included in the bleeding risk scores ranges from three (177) to 12
(181). All bleeding scores (177-182, 187) include common clinical factors that influence the
risk of bleeding for example age, utilising different age ranges and cut-offs (ranging from above
50 years old to above 85 years old) to indicate greater risk of bleeding; three scores include
age ≥75 (178, 179, 181). After age, the most common bleeding risk factors included in the
scores are as follows: (i) previous/remote bleeding (reported in 7 scores) (177-182, 187), (ii)
renal disease (included in 5 scores) (178-181, 187), and (iii) anaemia (in 5 scores) (178, 179,
181, 182, 187), hypertension (179-181), stroke (180, 181, 187), combined antiplatelet therapy
(178, 180, 182) and alcohol excess (180-182) (all included in 3 scores). Two scores included
diabetes (182, 187) and liver disease (180, 181) and one score included malignancy (181),
reduced platelet count (181) and female sex (182).
65
Table 1.17: Risk factors for bleeding included in each bleeding risk score
Risk factor ABC (177) ORBIT(178) ATRIA (179) HAS-BLED (180) HEMORR2HAGES(181) Shireman (182) OBRI (187) Total
Age≥75 ✓ ✓ ✓ 3
Age≥70 ✓ 1
Age≥65 ✓ ✓ 2
Age≥50 ✓ 1
Biomarkers ✓ 1
Previous/remote bleed ✓ ✓ ✓ ✓ ✓ ✓ ✓ 7
Recent bleed ✓ 1
Anaemia ✓ ✓ ✓ ✓ ✓ 5
Renal disease ✓ ✓ ✓ ✓ ✓ 5
Liver disease ✓ ✓ 2
Hypertension ✓ ✓ ✓ 3
Myocardial infarction ✓ 1
Diabetes ✓ ✓ 2
Malignancy ✓ 1
Stroke ✓ ✓ ✓ 3
Combined antiplatelet therapy ✓ ✓ ✓ 3
Labile INR ✓ 1
Alcohol excess ✓ ✓ ✓ 3
Excessive fall risk ✓ 1
Genetic factors ✓ 1
Reduced platelet count ✓ 1
Female sex ✓ 1
Total risk factors 3 5 5 9 12 8 7
ICU/CCU: intensive coronary care unit/ coronary care unit; INR: international normalised ratio; PE: pulmonary embolism
66
Two bleeding risk scores, HEMORR2HAGES (181) and the ABC bleeding score, (177)
included factors that are not routinely available in daily clinical practice. HEMORR2HAGES
included genetic testing, although this was not available in their cohort, and the ABC score
included 3 biomarkers, GDF-15, cTnT-hs and haemoglobin.
1.4.5.2 Derivation and validation studies for the bleeding risk scores
Seven bleeding (177-182, 187) risk scores have been developed and validated between 1989
to 2016. Six were developed in AF patients (177-182), and one in a mixed disease cohort of
patients (valvular heart surgery, mitral valve disease, AF, stroke, TIA, pulmonary embolism
(PE), deep vein thrombosis (DVT), and other thromboembolism) (187).
The mean/median age of study population in the derivation cohort ranged from 61(187) to 80.2
years(181) (Table 1.18). Almost half of the population in the derivation studies were female
and only three studies (178, 179, 187) reported ethnicity, which was predominantly White.
Five out of six (177-180, 182) studies from the AF cohort reported hypertension as the
common co-morbid disease present in their population whereas one study from the mixed
cohort (187) reported kidney disease to be more prevalent in their patient population.
67
Table 1.18: Baseline patient characteristics of the derivation cohorts for each bleeding risk score
Patients, % ABC ORBIT ATRIA HAS-BLED HEMORR2HAGES† Shireman OBRI

Number of patients 14,537 7411 6123 3456 1604 19,875 556
Mean age (SD)/median 70 (19-97) 75(68-82) - 66.8(12.8) 80.2 88% ≥70 years 61 (14)
(IQR)
Sex (female) 36 42.4 41.8 39.3 57 52.5 53
White ethnicity - 89.6 86.6 - - - 93
History/ diagnosis of - 23.9 15.2 - 4.8 2.5 -
cancer
Anaemia/abnormal Hb/Hct - 36.6 12.4 - 8.5 7.5 -
Hypertension 87 84.9 62.0 65.6 0.4 72 -
Diabetes 25 30.6 20.6 18 - 29.6 8
CHF 31 34.7 - 29.5 - 59.8 -
# #
MI 13 15.8 0.4 34.6 - 68.5 4
Prior stroke/TIA 19 9.5 12.6 10.4 37.2 32.1 12
‡
GI bleed - 8.0 7.1 1.8 - 11.9 10
α § ¥
eGFR <60ml/min - 32.1 2.9 5.3 - 0.6 18
Antiplatelet - 37.9 1.0 - - 22.3 -
Warfarin - 93.5 - - 42.3 28.7 -
NOACs - 6.5 - - - - -
CAD: coronary artery disease; CHF: congestive heart failure; GI: gastrointestinal; hb: haemoglobin; hct: haematocrit; MI: myocardial infarction; NOACs: newer oral anticoagulant;
† warfarin users only; ‡ major bleed; # CAD; α<30ml/min; § renal failure; ¥ hepatic/renal failure; (-) not reported
68
Table 1.19 presents the characteristics of derivation and validation studies of bleeding risk
scores for AF. A prospective study design was used in three (177, 178, 180) out of six scores
for AF populations. Two (179, 182) studies used a retrospective analysis and one study (181)
derived their score from the previous bleeding score available in AF, while one retrospective
study design was used in mixed population (187).
Most had follow-up for at least 1 year except by the first score developed by Shiremen et al
(182) which followed their patients for the first 90 days following hospital discharge following
AF diagnosis. All studies derived their risk score using bleeding risk factors from large cohorts
of patients ranging from 3456 (180) to almost 20,000 (182) patients, apart from one study,
Landefield et al which only included 556 patients (187).
All bleeding risk scores stratified patients into three categories of bleeding risk (low,
intermediate and high) except for the HAS-BLED score which initially categorised bleeding
risk as high (score ≥3) and low-moderate risk (0-2) (141). These bleeding risk scores showed
major bleeding rates ranging from 0.6%-3% in the low risk group and 4.9%-30% in the high-
risk group in the validation cohorts (Table 1.20).
69
Table 1.19: Characteristics of the Derivation and Validation cohorts for each of the bleeding risk scores and composition of each
score
Author, year, Derivation cohort Validation cohort Major Bleed definition

country a. Study design a. Study design
b. Sample size b. Sample size
c. Length of follow up c. Length of follow up
d. Indication of d. Indication of
anticoagulation anticoagulation
ABC (177) a. Prospective a. Prospective ISTH criteria: clinically overt bleeding with at least one of (i)
2016 b. 14, 537 b. 8468 decrease haemoglobin concentration 2 g/L or more; (ii) transfusion
Sweden c. 1.7 years (median) c. 1.9 years (median) of 2 or more units of packed RBC; (3) that was fatal or occurred in
d. AF d. AF critical area or organ (intracranial, intraspinal, intraocular,
pericardial, intra-articular, intramuscular with compartment
syndrome or retroperitoneal)
ORBIT (178) a. Prospective a. Prospective (i) fatal bleeding and/or (ii) symptomatic bleeding in a critical area
2015 b. 7411 b. 14264 or organ (intracranial, intraspinal, intraocular, retroperitoneal,
USA c. 2 years c. Median 1.9 year intra-articular or pericardial, or intramuscular with compartment
d. AF d. AF syndrome), and/or (iii) bleeding causing haemoglobin level to fall of
20 g/L (1.24 mmol/L) or more, or leading to transfusion of two or
more units of whole blood or red cells.
ATRIA (179) a. Retrospective a. Retrospective Fatal, requiring transfusion of 2 U packed blood cells, or
2011 b. 6123 b. 3063 haemorrhage into a critical anatomic site (e.g., intracranial,
USA c. 6 years c. 6 years retroperitoneal)
d. AF d. AF
70
Author, year, Derivation cohort Validation cohort Major Bleed definition
country
HAS-BLED a. Prospective a. Prospective Any bleeding requiring hospitalization and/or causing a decrease in
(180) b. 3456 b. 3071 haemoglobin level of >2 g/L and/or requiring blood transfusion that
2010 c. 1 year c. 1 year was not a haemorrhagic stroke
UK a. AF d. AF
HEMORR2HAG a. Score derived from 3 a. Retrospective Any hospitalisation for haemorrhage determined by Medicare
ES (181) previously published b. 3791 claims
2006 bleeding scheme c. 36 months
USA d. AF
Shireman et al a. Retrospective a. Retrospective Hospital admission for either a GI haemorrhage

(182) b. 19,875 b. 6,470 or intracranial haemorrhage according to the DRG and ICD-9 CM
2006 c. 3 months c. 3 months codes
USA d. AF d. AF
OBRI (187) a. Retrospective a. Prospective Overt bleeding that led to the loss of at least 2.0 units in 7 days or
1989 b. 556 b. 264 less, or was otherwise life-threatening
USA c. 48 months c. 48 months
d. Valvular heart surgery, d. VTE, prosthetic heart
mitral valve disease, valve, others
AF, stroke, TIA, PE,
DVT, other
thromboembolism
AF: atrial fibrillation; DRG and ICD-9 CM codes: Diagnosis Related Group and International Statistical Classification of Disease and Related Health Problems; GI: gastrointestinal; ISTH: International
Society on Thrombosis and Haemostasis; PE: pulmonary embolism; RBC: red blood cell; TIA: transient ischemic attack; U: units; VTE: venous thromboembolism
71
Table 1.20: Risk factors, risk categories and bleeding events in the validation cohorts
Bleeding events in validation

Risk factors and points awarded Risk categories
cohort (per 100 patient years)
ABC Age
Biomarkers Low: <1% 0.62
GDF-15 Medium: 1-2% 1.67
cTnT-hs High: >3% 4.87
Haemoglobin
Clinical history (previous bleeding)
ORBIT Older age ≥75 years 1 Low : 0-2 2.4a

Reduced haemoglobin/ hct/anaemia 2 Medium : 3 4.7
Bleeding history 2 High : ≥4 8.1
Insufficient kidney function 1
Treatment with antiplatelet 1
Total 7
ATRIA Anaemia 3 Low : 0-3 0.83

Renal disease 3 Intermediate : 4 2.41
Age ≥75 years 2 High : 5-10 5.32
Prior bleeding 1
Hypertension 1
Total 10
72
Bleeding events in validation
Risk factors and points awarded Risk categories
HAS-BLED Elevated systolic Hypertension 1 Low : <3 0=1.13
Abnormal renal and liver function (1 pt 1/2 High : ≥3 1=1.02
each)
Stroke 1 2=1.88
Bleeding 1 3=3.74
Labile INR 1 4=8.70
Elderly >65 years 1 5=12.50
Drugs or alcohol (1 pt each) 1/2 6=0.0
Total 9 7=-
8=-
9=-
Any score=1.56
HEMORR2HAGES Hepatic or renal disease 1 Low : 0-1 1.9-2.5

Ethanol abuse 1 Intermediate : 2-3 5.3-8.4
Malignancy 1 High : ≥4 10.4-12.3
Older age >75 years 1
Reduced platelet count or function 1
Re-bleeding risk 2
Hypertension (uncontrolled) 1
Anaemia 1
Genetic factors 1
Excessive fall risk 1
Stroke 1
Total 12
73
Risk factors and points awarded Risk categories Bleeding events in validation
Shireman et al Age ≥70 years 0.49 Low: ≤1.07 0.9% Within 90 days
Female 0.31 Moderate : >1.07 but <2.19 2.0% within 90 days
Remote bleeding 0.58 High : ≥2.19 5.4% within 90 days
Recent bleeding 0.62
Alcohol/drug abuse 0.71
Diabetes 0.27
Anaemia 0.86
Antiplatelet 0.32
Total 4.16
OBRI Age ≥65 years 1 Low : 0 3% at 12 month

History of stroke 1 Intermediate: 1-2 8% at 12 months
History of GI bleed 1 High : 3-4 30% at 12 months
Recent MI, anaemia, DM, 1
creatinine>1.5mg/dl
Total 4
cTnT-hs: Troponin T; DM: diabetes mellitus; GDF-15: growth differentiation factor-15; GI: gastrointestinal; hct: haematocrit; INR: international normalised ration; MI: myocardial
infarction; PE: pulmonary embolism; P/Y: person years; pt: point
a
bleeding event in original derivation cohort; b clinically important bleeding: sum of major bleed and clinically relevant non-major; (-) not available
74
The earliest bleeding score developed by Landefeld et al (187) in 1989 derived five predictive
factors of major bleeding in a mixed population. One of the original risk factors was AF but this
was later removed when the score was validated in 1989, as its association with major
bleeding in the derivation cohort was no longer significant in the validation cohort. Diabetes
mellitus was substituted instead of AF as a new predictor of major bleeding.
1.4.5.3 Performances of bleeding risk scores
The ability of the bleeding risk scores to predict bleeding risk has been validated in both similar
cohort where the score was derived (3 studies) (179, 180, 182) and in independent validation
cohort (4 studies) (177, 178, 181, 188). In the validation and comparison study by Hijazi et al
(177), the ABC score statistically outperformed the HAS-BLED and ORBIT scores in
predicting major bleeding in both the derivation cohort [0·68 (95% CI 0·66–0·70) vs. 0·61
(0·59–0·63) vs. 0·65 (0·62–0·67) respectively; ABC-bleeding vs. HAS-BLED p<0·0001 and
ABC-bleeding vs. ORBIT p=0·0008] and the external validation cohort [0·71 (95% CI 0·68–
0·73) vs. 0·62 (0·59–0·64) for HAS-BLED vs. 0·68 (0·65–0·70) for ORBIT; ABC-bleeding vs.
HAS-BLED p<0·0001 and ABC-bleeding vs. ORBIT p=0·0016](177). Although the ABC score
performed better than the HAS-BLED and ORBIT scores in this report, the complexity of the
algorithm and testing for biomarkers which are not routinely performed in daily clinical practice,
may make it difficult and costlier, for physicians to apply routinely.
One recent meta-analysis (189) compared the diagnostic accuracy between HAS-BLED and
HEMORR2HAGES, ATRIA, CHADS2 or CHA2DS2-VASc scores in anticoagulated patients with
AF. The findings revealed that the HAS-BLED score performed better than the
HEMORR2HAGES and ATRIA bleeding scores, as well as being superior to CHADS2 or
CHA2DS2-VASc in predicting bleeding. Despite having better performance when compared
to HEMORR2HAGES, ATRIA and ORBIT, an additional advantage of the HAS-BLED score
over the other five bleeding scores is the inclusion of quality of anticoagulation control (the ‘L’
75
acronym for labile INR or poor TTR<65%). TTR reflects anticoagulation control in patients
taking a VKA; a target TTR of ≥70% is optimal for efficacy and safety (173).
In a post-hoc analysis evaluating the performance of HAS-BLED, ATRIA and ORBIT bleeding
risk scores in the AMADEUS trial (190), TTR was strongly correlated with clinically relevant
bleeding events in patients using the ATRIA and ORBIT score, thus demonstrating that
incorporating TTR in bleeding scores improves their ability to predict future bleeding events.
Another comparison of four bleeding risk scores (HAS-BLED, ORBIT, ATRIA and
HEMORR2HAGES) in the SPORTIF cohort (191) also investigated whether the addition of
‘labile INR’ (TTR<65%) improved bleeding risk prediction (with the exception of the HAS-BLED
score which already contains labile INR). Addition of ’labile INR’ to ORBIT, ATRIA and
HEMORR2HAGES bleeding risk scores, significantly improved the predictive performance of
each score for major bleeding [integrated discriminatory improvement (IDI) 0.0023, p=0.0092
vs. IDI 0.0020, p=0.00014 vs. IDI 0.0015, p=0.0016 respectively](191).
Apostolakis et al (192) compared the predictive performance of HAS-BLED with
HEMORR2HAGES and ATRIA in the AMADEUS trial and demonstrated that the HAS-BLED
score performed better than HEMORR2HAGES and ATRIA score in predicting any clinically
relevant bleeding, with only the HAS-BLED score demonstrating significant improvement for
intracranial haemorrhage(192). In another ancillary analysis of the same trial (193), the HAS-
BLED score performed better than the ORBIT score in predicting any clinically relevant bleed
in a non-oral anticoagulant (idraparinux)(193).
More recently the predictive ability of the HAS-BLED score was also investigated in patients
receiving NOAC therapy, with rivaroxaban, in a small retrospective case-control study(194);
the HAS-BLED score demonstrated some diagnostic ability to predict major bleeding events
although this was not statistically significant (C statistics=0.68; p=0.07) (194). Analyses have
demonstrated that the HAS-BLED score not only performs well in predicting bleeding events
76
in VKA treated patients with AF, it can also be used to predict bleeding events in non-VKA
treated patients which is very useful as more AF patients are being treated with NOACs.
The next section will summarise factors affecting anticoagulation control in stroke prevention
in AF.
77
1.5 Anticoagulation control in stroke prevention
Studies (119, 122, 195) have shown that good anticoagulation control is associated with
reduction of thromboembolic complications while the risk of intracranial haemorrhage is
significantly higher in patients with high INR values (INR>4.0) (196). Therefore, it is important
to identify factors that might affect anticoagulation control in patients taking oral
anticoagulants.
1.5.1 Factors affecting anticoagulation control
These factors can be divided to demographic and clinical factors (Table 1.21) and non-clinical
factors (Table 1.22).
Table 1.21: Demographic and clinical factors affecting anticoagulation control
Demographic factors Lab parameters

Sex (female) (197-203) Albumin, g/dL† (204)
Age † (198-200, 204-207) Neutrophil, %† (204)
Younger age (<50 years) (198-200, 204-207) Red blood cell count, x106/mcL† (204)
Tobacco use (within 2 years) (199, 205, 208) Red blood cell distribution width, %† (204)
Ethnicity (non-white)(195, 198, 209, 210)
Alcohol (204)
Clinical factors
Warfarin naïve (195) Non-standardised target INR (201) (211)
Medical history* (200, 201) (131, 204, 212) Changes in gut flora (131)
Pneumonia (213) Numerous drug interaction (e.g.,
amiodarone, antibiotics, pain medications,
aspirin) (131, 199, 204, 213)
Bleeding history (213) Genetics (214-217)
Hospital stays ≥7days (213) Uncontrolled systolic blood pressure, mm
Hg†
No enhanced anticoagulation care¥ (213) Body mass index kg/m2 † (204)
Paroxysmal AF (195)
†continuous variable; * hypertension, diabetes mellitus, coronary artery disease/myocardial infarction, peripheral arterial disease,
congestive heart failure, previous stroke (ischemic or haemorrhagic), pulmonary disease, and hepatic or renal disease, venous
thromboembolism, thrombocytopenia, anaemia; ¥lack of participation (no access or plan) in a dedicated anticoagulation
management when VKA is initiated
78
Table 1.22: Non-clinical factors affecting anticoagulation control
Patient factors Physician factor Health care system factors

Knowledge and behavioural Bleeding risk(221) Hospital based vs. community-
factors (218-220) based vs. clinical trial (206,
222, 223)
Adherence (224-227) Lower targeted INR range Models of care (228-232)
(1.6-2.5) (221)
Socioeconomic status (200, Different treatment priorities
233) (221)
Dietary and herb interaction
(234, 235)
Fasting (236, 237)
1.5.1.1 Demographic and clinical factors affecting anticoagulation control
1.5.1.1.1 Female sex
Differences in the quality of anticoagulation therapy between males and females have been
observed in some studies while other studies have not confirmed this. Seven studies (197-
203) investigating predictors of TTR have demonstrated that women have poorer
anticoagulation control compared to men although the precise mechanism remains unclear
and should be investigated further (199). Poor anticoagulation control can be translated to
poorer clinical outcomes among women. The SPORTIF trial comparing warfarin vs.
ximelagatran (238) has shown that compared to males, females with AF were older, had more
stroke risk factors and had higher risk of stroke and thromboembolic events [2.08%/year, (95%
CI 1.60–2.56%/year vs. 1.44%/year, 95% CI 1.18–1.71%/year in men; p=0.016)] (238). A
meta-analysis (139) has also shown that female patients have higher residual risk of stroke
and systemic embolism despite the use of warfarin [OR 1.3, (95% CI 1.11 to 1.47); p=0.001]
as compared to the male patients (239). It can be speculated that maybe females have more
interruptions (238) (probably due to menopausal transitions), have more stroke risk factors
79
(238) and have the fear of the bleeding complications while on anticoagulation therapy leading
to poorer anticoagulation control compared to males.
1.5.1.1.2 Younger age (<50-60 years)
Some studies (198-200, 204-207) have shown that younger patients, in particular those <50
years old (206, 207, 240), have poorer anticoagulation control compared to older patients,
although the precise mechanism remains uncertain. It could be speculated that perhaps
younger patients have more active lifestyle, are less motivated to manage their AF due to
competing demands on their time (job, less leisure time) (199, 200) and have medication
adherence issues which might impact their quality of anticoagulation control.
1.5.1.1.3 Older age
Although studies have shown that younger patients have poorer anticoagulation control with
VKA therapy, the quality of anticoagulation control among the very elderly (aged ≥80 years) is
also perceived to be low (241-244). This perception leads to a lower prescription rate among
the very elderly patients (241, 245, 246). The ageing population with AF is increasing; they
too require effective anticoagulation therapy. However, anticoagulation in the elderly is not a
simple matter due to the increased risk of bleeding associated both with age per se and the
greater risk of co-morbidities and polypharmacy (241, 247, 248). Also, VKA therapy may be
more difficult among the elderly due to the frequency of INR monitoring required (which may
be more problematic if it requires travel to an anticoagulant clinic), dietary intake and drug
interactions (249).
Among the studies in the elderly anticoagulated population, few report anticoagulation control
with the exception of the WASPO (250) and BAFTA (251) trials with mean TTR 67 and 69
respectively. Good anticoagulation control (mean TTR 71% in both studies) was seen in
another two Italian studies of their elderly cohort of ≥80 (247) and ≥75 (252) years. However,
a much lower TTR (mean age 77, mean TTR 58%) can be seen in another study (241) of
80
elderly patients in the inception warfarin period. None of these studies (241, 247, 252)
investigated the association of age with TTR but two studies (241, 247) showed that increasing
age (age ≥80 years for both studies) was significantly associated with bleeding events
[adjusted OR 2.0 (1.1-4.0); p=0.05 (247) and unadjusted incidence rate 2.75 (1.27-5.95)
(241)].
1.5.1.1.4 Medical history
Numerous studies have shown that comorbid diseases influence patients’ quality of
anticoagulation with warfarin therapy. Poorer anticoagulation control is associated with heart
failure (197-202, 253-255), diabetes (197-200, 202, 205, 256), kidney disease (198, 199, 201,
213, 257), liver disease (198, 199, 254, 257), lung disease (199, 201, 204, 205), coronary
artery disease (199, 201), peripheral vascular disease (199, 201), stroke (199, 204),
pneumonia (205, 213), cancer (201, 258), major depression (201, 259), venous
thromboembolism (204, 258), previous bleeding (213), thrombocytopenia (204), bipolar
disorder (259), and psychosis (259). The exact mechanism of this relationship is unclear but
perhaps this reflects greater illness burden and complexity, more medications leading to
increased risk of non-adherence, polypharmacy and drug interactions (will be described in
more detail in section 1.5.1.1.9 page 86), poorer quality of life which all might lead to poorer
anticoagulation control (200).
1.5.1.1.5 Chronic kidney disease (CKD)
Among all the other comorbid disease affecting anticoagulation control, the impact of CKD
towards anticoagulation control will be discussed in more detail. This is because studies have
shown an increased risk of bleeding among AF patients with CKD (especially those with
severe renal impairment and on dialysis) while using OACs (260-262). Nonetheless, studies
(263-266) have also shown that good anticoagulation control while on VKA therapy among
CKD patients is associated with reduced risk of stroke [HR 0.60 (95%CI 0.39-0.93)], major
81
bleeding [HR 0.58 (95%CI 0.42-0.80)] and mortality [HR 0.61 (95% CI 0.46-0.82). (264) To
date, eight studies (263-270) have presented information on anticoagulation control among
AF patients with CKD. All studies showed a decrease in TTR as the kidney function worsened.
Indeed, the presence of CKD was negatively associated with achieving good TTR [OR 0.75
(0.67-0.92)] in the SPORTIF III and V trial cohort (264). Similarly, CKD was also as
independent predictor of TTR in the Current Perspective of Anticoagulation in Clinical Practice
in the Primary Care Setting (PAULA) study [unstandardized coefficient -3.4 (95% CI -5.51 to -
1.29); p=0.002] (270).
The exact mechanism of poor TTR among CKD patients in unknown but studies have shown
(270-272) patients with CKD are usually at risk for under- or overanticoagulation; among the
latter this is a result of reduced clearance of S-warfarin in CKD patients (273). Indeed, patients
with end stage renal disease (ESRD) have a 50% increase in plasma warfarin S/R ratio
compared to patients with normal renal function (274). This could perhaps reflect a decrease
in CYP2C9 activity in patients with renal failure, thus necessitating a lower dose in these
patients (274). Following that, Limdi et al (268, 275) showed that after accounting for clinical
and genetic factors, patients with reduced kidney function were able to maintain therapeutic
anticoagulation with lower warfarin dosage (average dose 3.9mg/day in severe group vs.
4.8mg/day in normal group; p=0.0002) (268, 275).
1.5.1.1.6 Treatment with interacting drug (e.g.: amiodarone)
Pharmacological rhythm control strategies in AF, particularly with amiodarone, are known to
have some effect on INR readings. Amiodarone, a potent inhibitor of both the S-enantiomer
and R-enantiomer of warfarin (276) is known to inhibit the metabolism of warfarin thus
potentiating an enhanced anticoagulant effect of warfarin (277). In addition, amiodarone has
a long half-life thus causing this potential drug interaction to occur for several weeks or months
after cessation of amiodarone (277). This might lead to an increase in INR values in patients
taking warfarin together with amiodarone thus translating into poorer TTR among these
82
patients. Apostolakis et al (199), in the original cohort of the SAMe-TT2R2 score, reported
14.3% of patients were prescribed Amiodarone (for rhythm control) and this was associated
with low TTR (ß=-0.03 95%CI (-0.06 to 0.0); p=0.05) (199).
1.5.1.1.7 Smoking history
Smoking is not only a predictor of poor anticoagulation control in three studies (199, 205, 208);
it also has also been shown to predict severe bleeding in patients treated with warfarin therapy
[HR 1.32; (95%CI 1.04-1.67; p=0.02)] in the Loire Valley AF Project (278). Meanwhile, a meta-
analysis of 13 studies (279) assessing the interaction between smoking and warfarin has
shown that warfarin clearance might be enhanced by the effects of smoking, which in return
leads to a reduction in the effects of warfarin. This meta-analysis also found that a significantly
higher dose of warfarin was required in active smokers compared with non-smokers to achieve
a therapeutic INR (279) which might explain the increased risk of bleeding among smokers.
The exact relationship between smoking and anticoagulation control is unclear but it may
reflect less interest in maintaining good health (among smokers) that may translate into poorer
adherence to OACs, thus resulting in poor TTR.
1.5.1.1.8 Ethnicity
Studies have shown that ethnic minority groups (Blacks, Hispanics and Asians) have poorer
anticoagulation control compared to Whites. Three studies (198, 209, 210) conducted in the
United States and two trials (280, 281) comparing TTR among the Blacks and Whites have
shown that mean TTR among Blacks was lower compared to Whites (Table 1.23). Similarly,
in the SPORTIF III and IV trials (195), the proportion of patients from Black/African Americans
was greater in the poor anticoagulation control group compared to moderate and good
anticoagulation control group (Table 1.23).
These observations may be due to various reasons for example differences comorbid disease,
socioeconomic status, poor understanding of therapy, adherence issue and genetic
83
background. Ethnic differences in anticoagulation control were evident in a cohort of 98,053
patients (210) receiving warfarin therapy for various indications (AF, VTE and other mixed
conditions), with lower mean TTR among the Blacks compared to Whites. Blacks were
younger and lived in areas of highest quartile of poverty, had higher illness burden including
more comorbid disease, requiring more medications and hospitalisations to manage those
conditions compared to White patients (210). After accounting for all these factors, which are
mostly non-modifiable, Black patients still had a recorded TTR 2.3% lower than White patients
(210) (Table 1.23).
In terms of pharmacogenetics, warfarin metabolism and dose requirements might differ
between ethnic groups. Studies have shown that warfarin dosage requirements are higher in
Blacks compared to Whites partly due to racial differences in genotype frequencies (216).
Blacks have been found to have additional CYP2C9 alleles which are associated with reduced
function of the CYP2C9 activity and thus might contribute to dose variability (216). In addition,
issues like health literacy, adherence to medication might also contribute to the differences in
quality of anticoagulation therapy among different ethnic groups (210).
84
Table 1.23: Major studies reporting anticoagulation control in different ethnic groups
Good
Poor INR Moderate
Study Follow Mean TTR, INR
Patients Ethnicity control INR control
design up % control
(TTR≤60) (TTR 60-75)
(TTR≥75)
SPORTIF III and IV trial, 3587 AF RCT 16.6 (6.3) White - 87.3 93.7 96.1
USA (195) months Asian - 9.6 4.5 3
Black/African
- 2.5 1.6 0.8
American
IMPACT trial, North 2718 but 229 RCT 2 years White 55 - - -
America, Europe, with INR results Black 44 - - -
Australia (280, 281) Asian 68 - - -
Non-Hispanic 54 - - -
Hispanic 48 - - -
VARIA study, 98,053 AF, VTE, Retrospective 2 years White 62.3 - - -
USA (198) others Black 55.8 - - -
TREAT-AF study, 184, 161 AF Retrospective 90 days White 59 (18) - - -
USA (209) Black 52 (20) - - -
Asian 59 (18) - - -
ORBIT-AF, 10, 132 AF Prospective 2.1 years White 68 (53-80) - - -
USA (210) Black 59 (41-75) - - -
Hispanic 62 (46-78) - - -
AF: atrial fibrillation; INR: International normalised ratio; RCT: randomised controlled trial; TTR: Time in therapeutic range; VTE: venous thromboembolism
85
1.5.1.1.9 Other clinical factors affecting anticoagulation control
As seen in Table 1.21 there are other clinical factors that may affect anticoagulation control.
Seven studies (199, 200, 204, 205, 208, 213, 254) have demonstrated that numerous drug
interactions with warfarin such as amiodarone, aspirin, pain medications, and antibiotics might
impact the quality of anticoagulation therapy. This can be explained by the fact that these
medications are inhibitor (amiodarone, analgesics, antibiotics, ex macrolides, quinolones and
azoles groups) or inducer of CYP2C9 (carbamezipine) enzyme that is involved in warfarin
metabolism, thus concomitant use of these drugs might cause potentiation or inhibition of the
warfarin effect (282). The concomitant use of these medications with warfarin should be
avoided but if essential, they should be used with caution (dose adjustment of warfarin
required) and with careful, regular monitoring of INR (277, 282).
Warfarin is mainly metabolised in the liver by the enzyme cytochrome-P450 2C9 (CYP2C9)
and it exerts its anticoagulant effect by inhibiting the protein VKORC1. Anticoagulant effects
of warfarin have been found to be influenced by the effect of three single nucleotide
polymorphism (SNPs); two in the CYP2C9 gene (CYP2C9*2 and CYP2C9*3) and one in the
VKORC1 gene. Patients with CYP2C9*2 (more commonly in Whites), CYP2C9*3 and
VKORC1 variants will metabolise warfarin less efficiently thus require lower doses of warfarin
(216). Although patients with CYP2C9*2, CYP2C9*3 and VKORC1 variants are at risk of over
anticoagulation during the initiation phase, (214, 215) the impact on TTR among the
maintenance phase remains debatable (217).
Combining common clinical and demographic predictors of INR control together, Apostolakis
et al (199) developed a scoring system, known as the SAMe-TT2R2 score. In 2017, Lin et al
(213) and Williams et al (204) published two novel scoring systems; the former with seven
factors known as the PROSPER score (213), the latter (204) with 15 and the SAMe-TT2R2
score (199) with six predictors of anticoagulation control (Table 1.24).
86
All three scores (199, 204, 213) included concomitant medical history and potential drug
interaction as predictors of anticoagulation control. Both SAMe-TT2R2 score (199) and
Williams et al (204) included demographic factors of age, whereas SAMe-TT2R2 score
included additional demographic factors such as tobacco use and ethnicity while Williams et
al (204) included body mass index and alcohol. PROSPER score and Williams et al included
more clinical factors where there might be an overlap between pneumonia and prescription of
antibiotics in the PROSPER score (213) and the latter included more laboratory variables
(204) as compared to the SAMe-TT2R2 score (Table 1.24).
In the SAMe-TT2R2 (199) and PROSPER (213) score, if patients had a score of >2, they are
predicted to have poor TTR with the latter (213) being emphasized for the geriatric population.
Whereas in the model by Williams et al (204), if patients have ≥4 and ≥7 poor TTR factors,
their estimated TTR will be <60% and <50% respectively (Table 1.24). NOAC is the preferred
anticoagulant of choice rather than VKAs if patients had high scores in all three scores (199,
204, 213). These scores were developed to identify patients at risk of having good/poor
anticoagulation control with warfarin and thus can aid physicians to choose the appropriate
anticoagulant therapy for stroke prevention in AF. Although the two new scores (204, 213)
seem very comprehensive, they are very complex and are not ‘user friendly’ especially in a
busy clinical setting.
87
Table 1.24: Scores to predict anticoagulation control
SAMe-TT2R2 Points PROSPER Points Williams et al

Sex (female) 1 Pneumonia 1 Age, y¥
Age (<60 years) 1 Renal dysfunction† 2 Systolic BP, mmHg¥
Medical history* 1 Oozing blood (bleeding history) 1 Body mass index, kg/m2¥
Treatment (interacting drugs, e.g., 1 Staying in hospital ≥7 days 1 Albumin, g/dL¥
amiodarone)
Tobacco use (within 2 years) 2 Pain medications 1 Neutrophil, %¥
Race (non-white) 2 No enhanced anticoagulation care‡ 4 Red blood cell count, x 106/mcl¥
Prescription for antibiotics 1 Red blood cell distribution width, %¥
Alcohol problem£
Anaemia£
Lung disease£
Stroke haemorrhagic£
Thrombocytopenia£
Venous thromboembolism£
Any antiarrhythmic£
Aspirin£
Cut offs
Scores 0-2 Good Scores 0-2 Good ≥4 poor factors TTR
INR INR <60%
control§ control§
Scores >2 Poor Scores >2 Poor INR ≥7 poor factors TTR
INR control§ <50%
control§
*Two or more of the following: hypertension, diabetes mellitus, coronary artery disease/myocardial infarction, peripheral arterial disease, congestive heart failure, previous stroke, pulmonary disease,
and hepatic or renal disease. ¥continuous variable; £binary variables; †Having records of acute kidney injury, chronic kidney disease, or end stage kidney disease prior 180 days
‡
lack of participation (no access or plan) in a dedicated anticoagulation management service when initiating a VKA; § on probability
88
1.5.1.2 Non-clinical factors
1.5.1.2.1 Patient factors
Studies have shown that patient’s comprehension and acceptance of the complex regimen
required with warfarin therapy might also affect the quality of VKA treatment (283). Evidences
have shown that in the general AF population, patients have minimal knowledge about their
medical condition (284-289), poor understanding of the benefits and risk of specific treatment
in particular OAC therapy (285-291), and are usually not aware of factors that can influence
the effectiveness and safety of the treatment (284-287). However, one recent survey
conducted in eight European countries showed different results (292). Most of their patients
(91-94%) have good knowledge of anticoagulation (in terms of indication and INR target
range). Furthermore, patients with college or university grades had lower frequent deviations
of their target INR range (2.8% vs. 5.1%, p<0.05) and had higher awareness (57% vs. 38.5%,
p<0.05) of the anticoagulation related risk of bleeding compared to those without schooling
respectively (292). However, this survey did not indicate the proportions of patients with
different education level thus this might affect their results (probably they had higher
proportions of patients with college or university grades).
Recently, the ESC guideline has recommended a tailored patient education in all phases of
AF management (3). Patient education is important to ensure accurate information is delivered
about AF and its treatment. Apart from that, obtaining feedback from patients regarding any
concerns or barriers (for example food and drug interactions, the need to modify lifestyle habits
and the risk of bleeding complications) that prevent them from taking warfarin (283) are also
crucial. The concerns should be addressed so that any barriers to medication taking can be
avoided and thus optimal outcome of anticoagulation therapy can be achieved. In the recent
years, several patient information tools (for example phone apps or structured face to face
89
educational session) have been developed focusing on oral anticoagulation therapy (293,
294). Evidences have shown that these educational tool or intervention was effective at
improving the quality of anticoagulation therapy (219, 293, 295, 296). The TREAT (219)
educational intervention (an educational-behavioural intervention), was effective at improving
the quality of anticoagulation therapy (TTR) among patients receiving it compared to usual
care (TTR at 6 months 76.2% vs. 71.3%, p=0.035 respectively) (293). Two recent surveys
conducted in Serbia (218) and Singapore (220) demonstrated that better knowledge, quality
of life, adherence rate and higher satisfaction to VKA therapy resulted in good TTR at follow
up; however these studies were not designed to measure the impact of an educational
intervention on TTR.
Despite overwhelming evidence showing the effectiveness of oral anticoagulation in stroke
prevention in AF, the demonstrable benefits of the OACs, in terms of stroke prevention, will
not be translated if patients do not adhere to, or fail to persist with their medications (226). A
recent extensive review of 30 studies (226) assessing adherence and persistence towards
OACs reported that within the five NOAC trials (132-135, 297), the discontinuation rate for
patients who were on warfarin was similar to each of the NOACs ranging between 13-34%.
(132-135, 297) In the same review, 13 ‘real-world’ [9 NOACs (298-306) and 4 warfarin (227,
307-309)] studies were also included (226). The adherence rate to warfarin was lower (40-
56%) in three of the warfarin studies (227, 307, 308), compared to 63-99% adherence in the
nine NOAC studies (298-306). Among all these studies, only two studies (227, 300)
demonstrated the impact of non-adherence on treatment outcomes. Although retrospective
and with a relatively short duration of follow up (median 0.67 years), Shore et al (300)
demonstrated that even a 10% decrease in adherence to dabigatran therapy was associated
with 13% increase in the combined outcome of all-cause mortality and stroke. In a larger
number of population (N=64, 661), with a slightly longer follow up (median 1.1 year), Yao et al
(227) showed an increased risk of stroke in patients with CHA2DS2-VASc score of ≥2 not taking
OACs for ≥6 months.
90
Other patient-related factors that might influence anticoagulation control are socioeconomic
status. Rose et al (200) has demonstrated that patients living in the poorest area (based on
the zip code of residence) was predicted to have poor TTR control compared to the wealthiest
during the first 6 months of warfarin therapy. This relationship persisted even after 6 months
being on warfarin therapy (200). In another population based study of 166,742 patients, lower
socioeconomic status (based on the median neighbourhood income quantiles) was also a risk
factor for bleeding and bleeding related to mortality among older individuals taking warfarin
therapy (233).
Common concerns among patients receiving VKA relate to food and drug interactions.
Patients receiving warfarin are advised to reduce intake of food that is rich in vitamin K, for
example, green leafy vegetables (spinach, broccoli etc.) as this might impact INR stability.
Vitamin K-rich food might counteract the anticoagulant effects of warfarin and dose adjustment
is required in patients presenting with low or high INR values after drastic changes in dietary
intake of vitamin K. A recent systematic review (234) (two intervention trials and nine
observational studies) summarising current evidence on the interaction between dietary
vitamin K intake and warfarin concluded that the evidence available does not support the
restriction of dietary intake of vitamin K but encourages patients to have a stable dietary habit
and avoidance of dramatic changes in dietary vitamin K (234).
Due to the narrow therapeutic index of warfarin, concomitant use of warfarin and herbal
remedies results in a major safety concern (235). Warfarin accounted for 26% of cases of
drug-herb interaction from clinical cases (310). A review on clinical evidences of herb and
warfarin interaction has highlighted clinical effects, severity of interaction and quality of clinical
evidences (235). They have identified thirty-eight selected herbs, four were evaluated with
Level I evidence as ‘highly probable’ to interact with warfarin, three were ‘probable interaction’
with Level II evidence, ten were ‘possible’ (Level III evidence) and twenty-one were ‘doubtful
91
interaction’ (Level IV evidence). The concomitant use of warfarin should be strongly avoided
in ‘highly probable’ (Cranberry, Soya, St John’s wort and Danshen) and ‘probable herbs’
(coenzyme Q10, Chinese Angelica, Ginger). Whereas for ‘possible’ (for example Ginko and
Chamomile) and ‘doubtful’ (for example Fenugreek and Parsley) interaction, for safety reason,
close monitoring of INR is recommended (235). For example, Cranberry juice commonly used
for blood and digestive disorder has been found to be linked to a major bleeding and high INR
(due to potentiation of warfarin effect) in a case report in the US involving a man who took
warfarin after drinking 710 ml of cranberry juice (311). Despite the available case reports about
warfarin and herb interaction, the intensity of the interaction might be overestimated. Future
studies or trials are needed to ascertain the magnitude and the clinical impact of these
interactions.
The impact of fasting on anticoagulant control is debatable. A prospective Singaporean (236)
study investigated the effect of fasting among 32 patients taking warfarin pre-Ramadhan,
during Ramadhan, and post Ramadhan. Although underpowered, a decrease in TTR was
seen from 81.0% to 69.6% before Ramadhan to during Ramadhan, respectively. In contrast,
another recent study (237) showed that TTR was better during the Ramadhan period
compared to pre-Ramadhan period (TTR 82.1% vs. 70%; p<0.001). More studies with larger
sample size are needed to ascertain the impact of fasting towards anticoagulation control
among warfarinised patients.
1.5.1.2.2 Physician factors
Bleeding risk is the most commonly cited (221) reason for not initiating or delaying warfarin
treatment. In some older patients, some physicians prefer to target a lower INR range (INR
1.6-2.5) to avoid bleeding complications as elderly (>65 years in HAS-BLED score) is a risk
factor for bleeding complications.
92
The benefit of oral anticoagulation in an elderly population has been demonstrated in several
studies. Focks et al (312) reported the rate of major bleeding and ischemic stroke of 2.8 and
2.3 per 100 patient years respectively, in his cohort of the very elderly (≥80years) AF patients.
Based on number needed to treat (NNT: 91) and number needed to harm (NNH: 22) a total of
four strokes/TIAs can be prevented based on every major bleed caused by VKA. Meanwhile,
Friberg et al (313) has shown a positive net clinical benefit from treatment with warfarin
(adjusted net clinical benefit >6%/y) in nearly all AF patients except those at lowest risk of
stroke as the benefit of preventing a stroke far outweighs the smaller risk of bleeding even in
patients with high HAS-BLED scores.
Bleeding risk should not be used as a reason to withhold anticoagulant treatment; instead
anticoagulation treatment should be used with caution and strict control even in patients at
high risk of bleeding complications (3, 314). Some physicians would prefer to avoid a major
bleeding event than to prevent stroke, whereas patients are prepared to accept the risk of
bleeding rather than to suffer from a stroke (221, 315). Currently there are five options for OAC
therapy for stroke prevention in AF (VKAs, dabigatran, apixaban, edoxaban, rivaroxaban). The
availability of NOACs allows physicians to choose the best NOAC that would fit into their
patient’s criteria. In the case of patients with high risk of bleeding (HAS-BLED>3), dabigatran
110 mg, apixaban 5mg or edoxaban 60 mg can be offered to patients (316).
1.5.1.2.3 Health care system factors
1.5.1.2.3.1 Hospital vs. community-based vs. clinical trial
To date, 16 studies have validated the SAMe-TT2R2 score in AF cohorts. These studies are
described in detail in the next section (section 1.6, page 101). In these studies (Table 1.25
and Figure 1.6), seven (199, 317-322) used hospital anticoagulation clinics to monitor
patient’s INR, four (323-326) stated hospital monitoring without specifying whether it was
performed by an anticoagulation clinic or not, three (199, 280, 327) were from clinical trial
93
settings and one from primary care (328) and cardiology outpatient clinic each. (329)
Anticoagulation control was highest in patients monitored by hospital anticoagulation clinics,
with TTRs ranging from 58% (319) to 78% (322). In the hospital setting (without information
on anticoagulant clinic involvement), TTR ranged from 38.2% (325) to 58% (326). In the
clinical trial settings, TTR was 53.6% (280) to 68.5%(327) and lastly, TTR was 69%(328) in
primary care.
A 2006 meta-analysis of 67 studies (222) compared the effect of study setting on
anticoagulation control in a mixed group of patients (on OAC for AF, valvular disease, VTE,
cerebrovascular disease, peripheral vascular disease) and showed that anticoagulation
control was significantly lower in the community setting compared to those from
anticoagulation clinic and in RCTs [unadjusted mean TTR 56.7 (51.5-62.0) vs. 65.6 (63.7-
67.7) vs. 66.4 (59.4-73.3), p>0.0001 respectively]. Similarly, another meta-analysis of eight
studies from 14 participating centres in the United States reported that AF patients who are
managed in the community setting had a lower mean TTR (51%) compared to those managed
by the anticoagulant clinic i.e. TTR 63% (223). In contrast, in the Auricula registry of 18,391
patients in 67 different centres, found no significant difference in the mean TTR among
hospital based centres versus community based centres (TTR 75.7% vs. 80.3% respectively)
and their mean TTR in the entire population was 76.2%, higher than reported in clinical trials
(206).
94
Table 1.25: Mean TTR% in studies validating the SAMe-TT2R2 score
a. Study design Population Method INR Mean (SD)/

b. Length of follow-up monitoring Median (IQR) TTR
Pivatto Junior, 2017, Brazil a. Retrospective 263 Anticoagulant clinic- 62.5 (44.2-79.5)
(317) b. 1 year hospital
Bernaitis, 2016, Singapore a. Retrospective 1137 Hospital based 58.0 (34.3)
(326) b. -
Chan, 2016, Hong Kong (325) a. Retrospective 1428 Hospital based 38.2 (24.4)
b. 4.7 ± 3.6 years (mean)
Gorzelak-Pabis, 2016, Poland a. Prospective 104 Hospital based 51 (32)
(324) b. -
Lip, 2016, USA (280) a. Prospective 229 Trial setting 53.6 (23)
b. 438 days on OAC
Lobos-Bejarano, 2016, Spain a. Retrospective 1524 Primary care 69 (17.7)
(328) b. >12 months
Proietti, 2016, Europe, Asia, a. Prospective 3665 Clinical trial centre 68.5 (55.17-79.32)
Australasia(327) b. 563 days (median)
Szymanski, 2016, Poland (323) a. Retrospective 211 Hospital 51.8 (25.0-71.2)
b. N/A
Abumuaileq, 2015, Spain (319) a. Retrospective 911 Anticoagulant clinic- PINRR 58 (18)
b. 10 months (mean) hospital
95
a. Study design Population Method INR Mean (SD)/
b. Length of follow-up monitoring Median (IQR) TTR
Roldán, 2015, Spain (318) a. Prospective 459 Anticoagulant clinic- 64 (17)
b. 6 months hospital
Ruiz-Ortiz, 2015, Spain (329) a. Retrospective 1056 Cardiology outpatient 63.8 (25.9)
b. 27 months (median) clinic-hospital
Gallego, 2014, Spain (322) a. Prospective 972 Anticoagulant clinic- 78 (19.98)
b. 952 days (median) hospital
Lip, 2014, France (330) a. Prospective 8120 Clinicians -hospital -
b. 1016±1018 days
(mean)
Poli, 2014, Italy (321) a. Prospective 1089 Anticoagulant clinic- 73 (62.5-82.0)
b. 4.6 years (mean) hospital
Skov, 2014, Denmark (320) a. Prospective 182 Anticoagulant clinic- 76 (-)
b. 1 year hospital
Apostolakis, 2013, UK (199) a. Retrospective and 1305 Derivation- clinical trial Derivation cohort 64.2
prospective Internal-clinical trial (18)
b. 3.5 years (mean) External: anticoagulant Internal validation 63.0
clinic-hospital (19)
External validation 66
(16)
96
90
Clinical trial
80 78 Hospital
76
73
Anticoagulant clinic-hospital
69 68.5
70
66
64.2 64 63.8 Primary Care
62.5
60 58 58 Cardiology outpatient clinic-hospital

53.6
51.8 51
Mean TTR%
50
40 38.2
30
20
10
0
ov
i
n
li
no
is
z
or
is
is
n
is
go
ti
sk
p
Po
rti
et
da
lie
ha
ab
ak
ak
at
Li
ni
Sk
ra
le
an
O
oi
ua
rn
ol
Ju
C
ol
ol
-P
al
ja
z-
Pr
m
R
Be
t
um
G
Be
G
os
os
ui
P.
y
Sz
R
Ap
Ap
L-
Ab
Studies
Figure 1.6: Mean TTR% among studies that validated the SAMe-TT2R2 score
97
1.5.1.2.3.2 Models of care
A variety of models of care known as usual care (UC) model, anticoagulant clinic (AC) model
and patient self-testing (PST) model (also known as home monitoring) were developed due to
the complexity of managing patients on warfarin therapy. In the UC model, patients are
managed by physicians without formal systematic monitoring procedures or policies to focus
on dose management. In the AC model, patients are usually managed by specialised nurses
or pharmacist under physicians’ supervision with systematic policies and procedures on
initiating, optimising and maintaining warfarin therapy. Lastly in PST model, patients self-
manage their own INR test at home using a portable point of care (POC) instrument. Patients
either receive instruction on dose from a healthcare provider or self-managed their own dose
(PSM) (331).
Studies have shown that different models of care might influence patient’s quality of
anticoagulation therapy and impact adverse clinical outcome. A recent systematic review (228)
of 25 studies (3 RCTs and 22 non-RCTS) involving 12,252 participants evaluated the quality
of anticoagulation control among pharmacist-managed anticoagulant services (PMAS)
compared with routine medical care. Quality of anticoagulation therapy was significantly better
in the PMAS group (TTR 66.9%) compared to usual care (TTR 56.7%), evidenced by a higher
TTR in the former compared to the latter in 23 of 25 studies. Adverse clinical outcomes were
also lower in the PMAS group compared to routine medical care, evidenced by lower or equal
risk of major bleeding (N=10 of 12 studies) and lower rates of thromboembolic events (in nine
out of 10 studies). (228) Another meta-analysis of eight RCTs (229) comparing the
effectiveness of PMAS versus other models of care (including UC by physicians, nurses and
other healthcare professional) in a mixed group of patients also showed better TTR in standard
therapeutic range for patients in the PMAS compared to UC group. However, safety and
mortality data were inconclusive; instead patients in PMAS group were most satisfied (229)
(Table 1.26). The findings from the two meta analyses showed the benefits of PMAS over UC
98
towards improving TTR however more studies are needed to confirm the benefits of PMAS in
terms of adverse clinical outcomes.
Evidence has shown that more frequent INR testing results in a reduction of thromboembolic
and bleeding events (332). With the difficulties associated with frequent INR visits to an
anticoagulant clinic in short interval of time, the concept of patient self-testing (PST) was
introduced. Since 2004, the American College of Chest Physicians (ACCP) have
recommended PST to be implemented ‘for patients who are motivated and can demonstrate
competency’ (Grade 2B) (333). Several meta-analyses of clinical trials showing the
advantages of PST/PSM have been conducted in 2011 (232) and 2012(231) (Table 1.26).
Bloomfield et al (232) demonstrated that in highly motivated adult patients requiring long term
anticoagulation therapy with warfarin, PST alone or in combination of PSM was associated
with significantly reduced risk of thromboembolic events (42%) and deaths (26%) without an
increased risk of major bleeding events compared to patients receiving UC (232). Similarly,
Heneghan et al (334) conducted a systematic review including 28 trials in 2016 also showed
a significant reduction of TE events in the PSM or PST groups [RR 0.58, (95% CI 0.45 to 0.75)];
but no significant reduction in major bleeding [RR 0.95, (95% CI, 0.80 to 1.12)] or all-cause
mortality [RR 0.85, (95% CI 0.71 to 1.01)] (231). Results from the two meta-analyses suggest
the benefits of PST and/or PSM towards reducing TE and mortality compared to UC.
Another small randomised trial (N=159) performed in the Netherlands reported that PST and
PSM patients had a significantly improved quality of life compared to PST patients or UC
(managed in the hospital) only (335).
99
Table 1.26: Models of care and anticoagulation control and/or clinical events
No of Populati Population Models TTR and/or clinical events

studies on, N
Manzoor 25 studies 12, 252 AF, VTE PMAS vs. TTR higher in PMAS
2017(228) 3 RCT UC (66.9%) vs. UC (56.7%) in
22 non RCT 23 of 25 studies
Zhou 2016 8 RCTs 1493 AF, VTE, valvular PMAS vs. TTR PMAS vs. UC: MD
(229) heart disease, UC 3.66 95% CI (2.2-5.11) for
CVA, INR 2.5 ± 0.5 (standard INR
cardiomyopathy, range)
mural thrombus TTR PMAS vs. UC: MD
and others 2.85 95% CI (-0.56 to 6.26)
for INR 2.5 ± 0.7 (expanded
INR range)
Heneghan 28 trials 8950 AF, valve PST and/or TE:

2016 (334) replacement, PSM vs. UC PST or PSM: (RR 0.58,
DVT 95% CI 0.45-0.74;
participants = 7594; 18
studies)
Mortality:
PST or PSM: (RR 0.85,
95% CI 0.71-1.01; 6358
participants, 11 studies)
Bloomfield 22 trials 8413 AF, mechanical PST and/or TE: OR 0.58, 95% CI (0.45-
2011(232) heart valve PSM vs. UC 0.75, =<0.001)
replacement Mortality: OR 0.74 95%CI
(0.63-0.87, p<0.001)
CI: confidence intervals; MD: mean difference; PMAS: pharmacist-managed anticoagulant service; PST: patient self-
testing; PSM: patient self-monitoring; UC: usual care
100
1.6 Use of the SAMe-TT2R2 score to predict anticoagulation control in
atrial fibrillation treated with vitamin K antagonists
The section summarises studies which have assessed and/or validated the SAMe-TT2R2
score in patients treated with VKA for AF. This section has been published in the Heart Rhythm
Journal (336).
The original purpose of developing the SAMe-TT2R2 score was to produce a simple clinical
schema which could be used routinely in everyday practice to help assess the likelihood of an
AF patient being able to achieve and maintain good anticoagulation control on VKA therapy,
using patient-related clinical parameters which are readily available. The availability of NOACs
worldwide has resulted in increased usage due to their advantages. These include faster
onset-of-action [average maximum effect approximately three hours after intake (337)
compared to VKA (onset 36-72 hours)], greater reduction in stroke/systemic embolism [+19%
compared to VKA(3)], avoidance of INR monitoring with NOACs(338), and absence of
achieving/maintaining adequate TTR (as with warfarin). Achieving a therapeutic INR can take
2-4 weeks and often longer (131). After termination of study drug in the NOAC trials, of those
patients switching to warfarin, <40% achieved a therapeutic INR within 15 days, and <80%
after 30 days(339); more strokes occurred during that period in the patients who went from
study drug to VKA than from VKA to VKA (339, 340). This strong argues for using NOACs
over VKAs where possible, however, VKAs are still widely used globally and will not disappear
from use especially for AF patients with severe renal impairment, moderate to severe mitral
stenosis or mechanical heart valves (3). In addition, in low- and middle-income countries
where cost plays an important role in options available for OAC treatment VKA is still the first-
line antithrombotic agent of choice, therefore the SAMe-TT2R2 score will remain an important
decision-making tool, currently and in the future, to guide physicians choice of anticoagulant
treatment (341).
101
1.6.1 SAMe-TT2R2 score validation studies
Current studies (N=16) assessing the SAMe-TT2R2 score in AF patients are summarised in
Table 1.27 and the baseline patient characteristics of these cohorts are presented in Table
1.28. The majority of the studies (N=9) (199, 280, 318, 320-322, 324, 327, 330) were
performed prospectively, with a follow-up duration ranging from six months (318) to 4.7 years
(325). Eleven of the studies were performed in European populations (199, 318-324, 327, 328,
330), three in Asian populations (325, 326, 342) and one in the American populations (280).
Proietti et al (327) studied a mixed indication clinical trial cohort including patients from Europe,
Asia and Australasia.
Most studies to date were performed in elderly (mean or median age ranging from 61 years to
76 years old) Western Caucasian populations, which mainly used warfarin (10 studies) (199,
280, 320, 321, 323, 325-327, 330, 342) as their OAC of choice. The majority of the patients
had multiple comorbidities with hypertension being the most common, except for the study by
Lip et al (330) where congestive heart failure was most common. All of the studies reported a
low prevalence of smoking status and use of amiodarone for rhythm control, with the exception
of Lip et al (330), with 35% of patients using amiodarone. As shown in Figure 1.7, as the
SAMe-TT2R2 score categories increase, the mean TTR of their study population decreases.
Five studies (280, 319, 325, 328, 343) investigated the relationship between components
included in the SAMe-TT2R2 score and TTR. Three studies (280, 319, 328) showed that female
sex was associated with poor anticoagulation control; one (319) showed that having ≥2
comorbidities was related to poor TTR and one (280) showed that black ethnicity (as well as
NYHA IV) was associated with poorer anticoagulation control. Chan et al (325) also reported
that having heart failure and diabetes mellitus independently predicts poor anticoagulation
control.
102
Table 1.27: Studies assessing the SAMe-TT2R2 score in atrial fibrillation cohorts
a. Study design Population SAMe-TT2R2 score distribution Percentage of

b. Mean follow-up a. Number (%); mean (SD) TTR (%) patients with
c. Method INR monitoring b. Mean (SD)/median (IQR) age dichotomised TTR
(range, years) (%)
c. Race/ethnicity
d. OAC used
Pivatto Junior(317) a. Retrospective a. 263 AF 0-1: 138 (52.5); 69.2 -
2017 a. 1 year a. 71.2 (64.1-78.5) ≥2: 125 (47.5); 56.3
Brazil b. Hospital OAC clinic b. White
c. 97.3% Warfarin
-
a. Retrospective a. 1137 AF 0-1:0
Bernaitis(326)
b. - b. 71 (63-77) 2: 339; 63.2 (34.1)
2016 c. Hospital c. Asian >2:798; 55.8 (34.1)
d. Warfarin
Singapore
Chan(325) a. Retrospective a. 1428 NVAF 2: 22 (14.3); 70† TTR≥70: 11

2016 b. 4.7 ± 3.6 years b. 76.2 (8.7) 3: 80 (51.9); 70 TTR<70: 89
Hong Kong c. Hospital c. Chinese 4: 41 (26.6); 70
d. Warfarin 5: 7 (4.5); 70
6: 4 (2.6);70
a. Prospective a. 104 AF with cognitive 0-1: 64 (26) -

Gorzelak-Pabis(324)
b. - impairment ≥2: 50 (28)
2016
c. Hospital b. 75 (10)
Poland
c. White
d. 61% Acenocoumarol
-
Lip(280) a. Prospective a. 229 AF 0-1:0.571 (0.22)
2016 b. 438 days b. 66.7 (11) ≥2: 0.498 (0.24)
USA c. Trial setting c. 80.3% White
d. Warfarin
103
a. Study design a. Number SAMe-TT2R2 score distribution Percentage of
b. Mean follow-up b. Mean (SD)/median (IQR) (%); mean (SD) TTR (%) patients with
c. Method INR monitoring age (range, years) dichotomised TTR
c. Race/ethnicity (%)
d. OAC used
Lobos-Bejarano(328) a. Retrospective a. 1524 NVAF 0-1: 69.6% (17.4) TTR≥65: 60.6
2016 b. >12 months b. 77.4 (8.7) ≥2: 66.6% (18.5) TTR<65: 39.4
Spain c. Primary care c. White
d. 94.8% Acenocoumarol
Proietti(327) a. Prospective a. 3665 AF 0-2: 2914 (80.4); TTR>70: 46.9

2016 b. Median 563 days b. 72(66-77) 69.05 (55.63-79.89) TTR≤70: 53.1
Europe, Asia, c. Mixed‡ >2: 710 (19.6);
(IQR 483-651)
d. Warfarin 66.55 (52.83-77.46)
Australasia
c. Trial setting
Szymanski (323) a. Retrospective a. 211 AF 0-1: 114 (54); 52.3 TTR>70: 25.2
2016 b. - b. 57.1 (10.2) ≥2: 97 (46); 51.3 TTR≤70: 74.8
Poland c. Hospital c. White
d. 75.4% warfarin
Abumuaileq(319) a. Retrospective a. 911 NVAF 0-1:672 (74); 59 (18)¶ PINRR>65:44

2015 b. 10 months b. 73 (11) ≥2: 239 (26); 54 (19)¶ PINRR≤65:55
Spain c. Hospital OAC clinic c. White
d. 93% Acenocoumarol
Roldán (318) a. Prospective a. 459 NVAF <2: 253 (55); 67 (18) TTR>65:54
2015 b. 6 months b. 76 (70-82) ≥2: 206 (44.8); 61 (16) TTR≤65:46
Spain c. Hospital OAC clinic c. White
d. Acenocoumarol
Ruiz-Ortiz(329) a. Retrospective a. 1056 NVAF 0-1:613 (58); 65.6 (26.2) TTR≥65:52.7

2015 b. Median 27 months b. 73.6 (9.8) ≥2: 443 (42); 61.3 (25.3) TTR<65:47.3
Spain c. Cardiology clinic c. White
d. Acenocoumarol
104
a. Study design a. Number SAMe-TT2R2 score distribution Percentage of
b. Mean follow-up b. Mean (SD)/median (IQR)age (%); mean (SD)TTR (%) patients with
c. Method INR monitoring c. Race/ethnicity dichotomised TTR
d. OAC used (%)
Gallego(322) a. Prospective a. 972 NVAF 0-1:431 (44); 79.67 (19.46) -
2014 b. Median 952 days b. 76 (70-82) ≥2: 332 (34); 78.4 (20.28)
Spain c. Hospital OAC clinic c. White >2:208 (21); 74.25 (20.24)
d. Acenocoumarol
Lip(330) a. Prospective a. 8120 AF†† 0-1: 4504 (55); 77(1.7)§ -

2014 b. 1016±1018 days b. 70 (15) ≥2: 2252 (28); 52(2.3)§
France c. Clinicians -hospital c. White >2:1364 (17); 43(3.2)§
d. Warfarin
Poli(321) a. Prospective a. 1089 AF 0-1:624 (57); 72.3 (15.3) -

2014 b. 4.6 years b. 75 (30-94) 2: 288 (26); 72.0 (15.6)
Italy c. Hospital OAC clinic c. White >2:177 (16); 68.2 (16.4)
d. Warfarin
Skov(320) a. Prospective a. 182 AF 0-1:105 (58); 76 -
2014 b. 1 year b. 70.2# ≥2: 77 (42); 76
Denmark c. Hospital OAC clinic c. White
d. Warfarin
Apostolakis(199) a. Retrospective and a. 1305 AF (Internal/External validation) Internal validation

2013 prospective b. 69(8)/74(10) 0: 242 (18); 0.66±0.16/0.7±0.13 TTR>70:35.7
United Kingdom b. 3.5 years c. 8.7%, 19.3 % non-white 1: 413 (31);0.65±0.18/0.66±0.17 TTR≤70:64.3
c. Clinical trial (internal- (internal/external-validation) 2: 303 (23);0.63±0.17/0.66±0.16 External validation
validation)/Hospital OAC d. Warfarin 3:185 (14); 0.59±0.22/0.65±0.17 TTR>70:44.1
clinic (external-validation) TTR≤70:55.9
AF: atrial fibrillation; CV: cardiovascular; INR: international normalised ratio; IQR: interquartile range; Max: maximum; MI: myocardial infarction; NVAF: non-valvular atrial fibrillation; OAC: oral
anticoagulant/anticoagulation; ROC: area under curve; SD: standard deviation; SAMe-TT2R2 score: sex (female), age (<60 years, medical history (≥2 of the following: hypertension, diabetes, coronary
artery disease or myocardial infarction, peripheral arterial disease, congestive heart failure, previous stroke, pulmonary, hepatic, or renal disease), treatment with interacting drugs (e.g. amiodarone[all
1 point], current tobacco use and race (non-white) [2 points]; TTR: time to therapeutic; TE: thromboembolism; VTE: venous thromboembolism †TTR presented as ≥70% and <70% not mean TTR;
‡mixed population: White, Black, Asian, other; §number of patients with labile INR, (%); ¶PINRR % (mean ± SD); #no SD or IQR reported; †† n=4637 on VKA; - not reported
105
Table 1.28: Baseline characteristics of all studies assessing the SAMe-TT2R2 score in AF population
N (%) Female Age Hypertension Diabetes HF Prior PAD Renal CAD COPD Smoking Previous Amiodaron
<60 stroke/TIA diseas bleeding e
e
Pivatto(317) 113 41 231 108 149 96 25 7 76 36 37 24 26
(43.0) (15.6) (87.8) (41.1) (56.7) (36.5) (9.5) (2.7) (28.9) (13.7) (14.1) (9.1) (9.9)
Bernaitis 448 172 677 343 88 45 - 156 271 - 84 - 78
(326) (39.4) (15.1) (59.5) (30.2) (7.7) (4.0) (13.7) (23.8) (7.4) (6.9)
Chan(325) 671 48.0 922 387 367 496 102 2.9 407 - 71.0 - 94
(52.5) (3.4) (64.6) (27.1) (25.7) (34.7) (7.1) (2.0) (28.5) (5.0) (6.6)
Gorzelak- 63 - 92 30 72 15 - - - - 20 - 8
Pabis(324) (60.6) (88.5) (28.8) (69.2) (14.0) (19.2) (7.7)
Lip(280) 47 57 206 106 208 26 31 - 178 - - - 46
(20.5) (24.9) (90.0) (46.3) (90.8) (11.4)/ (13.5) (77.7) (20.1)#
14 (6.1)
L.Bejarano 741 66 1223 473 392.0 209.0 99 92 286 - 100 134 100
(328) (48.6) (4.3) (80.2) (31.0) (25.7) (13.7) (6.5) (6.0) (18.8) (6.6) (8.8) (6.6)
Proietti(327) 1116 72§ 2812 860 1372 753 - - 1619 - 334 208 -
(30.5) (66-77) (76.7) (23.5) (37.4) (20.5) (44.2) (9.1) (5.7)
Szymanski 79 108 194 27 8.0 16 - - - - 31.0 - 17
(323) (37.4) (51.2) (91.9) (12.8) (3.8) (7.6) (14.7) (8.1)
Abumuaileq 306 - 678 220 343 103 92 36¶ 127 183 77 115 -
(319) (33.6) (74.4) (24.1) (37.7) (11.3) (10.1) (4) (13.9) (20.1) (8.5) (12.6)
Roldán(318) 237 38 368 141 87 67 - 51 70 50 38 37 72
(53.0) (8.0) (80.0) (31.0) (19.0) (15.0) (11.0) (15.0) (11.0) (8.0) (8.0) (16.0)
Ruiz- 443 - 884 321 235 150 - 153 215 176 76 56 102
Ortiz(329) (42.0) (83.7) (30.4) (22.2) (14.2) (14.5) (20.3) (16.7) (7.2) (5.3) †† (9.7)
Gallego 494 66 796 249 350 182 - 94 182 - 136 79 -
(322) (51.0) (7.0) (82.0) (26.0) (36.0) (19.0) (10.0) (19.0) (14.0) (8.0)
106
N (%) Female Age Hypertension Diabetes HF Prior PAD Renal CAD COPD Smoking Previous Amiodaro
<60 stroke/TI disease bleeding ne
A
Lip(330) 3,129 - 3,405 1,244 4,466 674 - 734 2,434 870 1,053 - 1,670
(39) (42.0) (15.0) (55.0) (8.0) (9.0) (30.0) (11.0) (13.0) (35.0)
Poli(321) 412 61 745 216 268 313 143 - 239 - 181 - 200
(37.8) (5.6) (68.7) (19.9) (24.7) (28.8) (13.2) (22.1) (16.6) (18.4)
Skov 54 23 - - - - - - - - 41 - 27
(320) (29.6) (12.6) (22.5) (14.8)
Apostola 382 147 692 200 197 130 57 53 173 - 64.0 - 129
kis(199)† (37.5) (14.4) (67.9) (19.6) (19.3) (12.8) (5.6) (5.2)‡ ‡ (17.0) (6.3) (12.7)
§§
Apostola 157 30.0 234 64 45 30.0 8 2.0 44 - 140 - 26

kis(199)‡ (67.1) (10.5) (81.8) (22.4) (15.7) (12.8) (2.8) (0.7)‡ ‡ (15.4) (49.0) (9.1)
§§
CAD: Coronary artery disease; HF: Heart failure; COPD: chronic obstructive pulmonary disease; eGFR: estimated glomerular filtration rate; MI: myocardial infarction; TIA:
transient ischemic attack; PAD: peripheral arterial disease
*internal validation †external validation; ‡ median age; § eGFR 30 ml/min/1.73m2; || antiarrhythmic; ¶ Major bleed; # hepatic/renal disease; ** history of MI
107
90
80
70
Mean TTR %
60 SAMe-TT2R2
score
50 categories
0-1
40
2
30 >2
20
10
-O
ng
6)
ov
go
-P
tis
n
ti
li
sk
ile
L-
Po
et
dá
01
G
Sk
R
Li
le
ai
an
ua
oi
al
ol
(2
rn
Pr
ym
um
G
R
Be
p
Li
Sz
Ab
SAMe-TT2R2 categories in validation studies
Figure 1.7: Mean TTR vs. SAMe-TT2R2 categories in validation studies
108
Six of the studies (199, 317, 319, 328-330) reported the predictive ability of the SAMe-TT2R2
score using C-statistics (Figure 1.8). Taken together, these validation studies suggest that the
SAMe-TT2R2 score is able to predict good or poor anticoagulation control among AF patients
better than chance, with C-statistics ranging from 0.56 (328) to 0.72. (199) Many risk scores
based on clinical factors such as CHADS2, CHA2DS2-VASc, Killip and TIMI scores show
broadly similar modest C-indexes (approx. 0.6) when used to predict patients categorised at
‘high risk’ who actually sustain clinical events (193, 314).
Six studies (319, 321, 322, 325, 327, 330) also examined if the SAMe-TT2R2 score could
discriminate AF patients with clinical events. Five (319, 322, 325, 327, 330) demonstrated
some positive associations for SAMe-TT2R2 score predicting clinical events, with C-statistics
ranging from 0.55 (330) to 0.62 (322) (Table 1.29). As seen in most of the studies, (199, 280,
318, 319, 321-330, 342) increasing SAMe-TT2R2 score demonstrated poorer TTR values
which might also translate into poorer outcomes. This can be evidenced by studies that
showed the SAMe-TT2R2 score relating to severe bleeding (322) and major bleeding (defined
by the Bleeding Academic Research Consortium) (330), stroke/TE (330), adverse
cardiovascular event (322) and death (322, 330) during follow up. In an observational study
performed in 911 non-valvular AF Spanish patients, the SAMe-TT2R2 score also successfully
predicted the composite outcome of major bleeding, TE complications and death (319). A
Chinese study also demonstrated that a SAMe-TT2R2 score of ≤2 vs. SAMe-TT2R2 of 3 vs.
SAMe-TT2R2 ≥4 is associated with lower annual stroke risk (3.49%/year vs. 4.56% per year
vs. 6.41%/year, respectively) (325) .
1.6.2 Importance of good anticoagulation control
Achieving good anticoagulant control (i.e. TTR≥65-70%) as recommended by guidelines (3,
344) is essential for managing AF patients treated with VKA. Indeed, numerous studies have
demonstrated that a high TTR translates into lower risk of stroke and bleeding (119, 120, 122,
195, 345). A systematic review demonstrated that a 7% and 12% improvement in TTR can
109
lead to a reduction in major bleeding and thromboembolic event, respectively, by 1 event per
100 patient years (120). A real-world study (119) of 27,458 warfarin treated patients for AF
(with at least 3 INR measurements) showed that in patients with good anticoagulation control
(TTR ≥70%), stroke risk was reduced to 79% compared to patients with poor INR control (TTR
≤30%). However, achieving and maintaining a therapeutic INR can be difficult to accomplish
and therefore, NOACs are preferred to VKA in the majority of patients requiring OAC initiation
(3).
1.6.3 Impact of different methods of calculating TTR
Fauchier and colleagues (346) have raised concern about the different methods used to
calculate TTR, whether to use TTR based on the Rosendaal method, percentage of INRs in
range (PINRR) (traditional method) or percentage of visits in range on a given date (cross-
sectional method), as these methods are not interchangeable. Overall, 14 studies (199, 280,
318, 320-329, 342) reported TTR using the Rosendaal method, only one study (319) utilised
PINRR method and another one reported ‘labile INR’ as their measure of anticoagulation
control (330). At the moment, there is no evidence showing which method of calculating
percentage of INR in range is best, as each method has its own unique strengths and
weaknesses (347). While TTR via the Rosendaal method calculates the exact percentage of
days the INR falls within range; its calculation is more complex than the others and is based
on linear extrapolation. In contrast, calculating TTR via the PINRR method is simpler as it only
looks at the number of INRs that fall within the therapeutic range divided by the total number
of INR tests undertaken. However, the PINNR method does not consider the actual number
of days of anticoagulant treatment and thus might underestimate control in patients with
inconsistent INR monitoring, patients who have temporary discontinued therapy and patients
with a long gap between each INR test, in contrast to the Rosendaal method where these
factors will be accounted for, resulting in a lower TTR.
The next section discusses about antithrombotic therapy in patients with valvular heart
disease.
110
SAMe-TT2R2 score validation studies
Predictive ability of SAMe-TT2R2 score on anticoagulation control with C-statistics (95% CI)
Figure 1.8: Predictive ability (C-statistics and 95% confidence intervals) of SAMe-TT2R2 towards
anticoagulation control in validation studies
111
Table 1.29: Predictive ability (C-statistics) of SAMe-TT2R2 for anticoagulation control and clinical events
Anticoagulation control, c-statistics (95% CI) Clinical events, c-statistics (95% CI)
Pivatto Junior (317) TTR≥65: 0.612 (0.544-0.681; p=0.002) -
Chan (325) - Stroke: 0.54 (0.52-0.57)

Abumuaileq (319) PINRR ≤70: 0.60 (0.56-0.64; p<0.01) Composite major bleeding, thromboembolic complication or
death:
0.57 (0.51-0.62); p=0.03
Ruiz-Ortiz (329) TTR≥65: 0.57 (0.53-0.60; p<0.0005) -
Gallego (322) - Adverse CV event: 0.62 (0.57-0.68; p<0.001)

Bleeding: 0.55 (0.49-0.62; p=0.117)
All-cause mortality: 0.62 (0.55-0.68; p<0.001)
Lip (330) Labile INR: 0.589 (0.574-0.603) Stroke/TE: 0.561 (0.547-0.575)
Severe bleeding: 0.552 (0.537-0.566)
Major BARC bleeding:0. 574 (0.560-0.589)
Death: 0.544 (0.530-0.559)
Apostolakis (199) TTR 31% internal 0.72 (0.64-0.795) -
TTR 36% external 0.70 (0.57-0.82)
112
1.7 Antithrombotic therapy in atrial fibrillation associated with valvular
heart disease
AF and valvular heart disease (VHD) often coincide and are present in about 2.5% of patients
in industrialized countries (348-350). However, the management of patients with both
conditions have not been well addressed in large clinical trials. Patients with valvular heart
disease are often excluded from most of the clinical trials due to the complexity of their
management strategy. Due to this, there is a lack of definitive guidance on how best to manage
this group of patients (349, 351).
Recently, new definition to ‘valvular AF’ has been proposed based on the type of oral
anticoagulation to be used in AF patients with valvular heart disease. The term valvular AF is
outdated; the new term is Evaluated Heart valves, Rheumatic or artificial (EHRA) type I VHD
and EHRA type II VHD (Table 1.30) (349, 351). These new terms will be used throughout the
thesis except when it is cited as ‘valvular AF’ from original studies.
Table 1.30: Classification of AF patients with valvular heart disease [taken directly
from (356, 358)]
Definition OAC therapy Valve type

Evaluated Heart valves, VKA only • Moderate to severe mitral
Rheumatic or artificial (EHRA) stenosis of rheumatic origin
type I VHD • Mechanical prosthetic valve
replacement
Evaluated Heart valves, VKA or NOAC • Mitral regurgitation
Rheumatic or artificial (EHRA) • Mitral valve repair
type II VHD considering • Aortic stenosis
CHA2DS2-VASc score • Aortic regurgitation
• Tricuspid regurgitation
• Tricuspid stenosis
• Pulmonary regurgitation
• Pulmonary stenosis
• Bio prosthetic valve
replacement
• Transaortic valve
intervention (TAVI)
113
1.7.1 Epidemiology of valvular heart disease with atrial fibrillation
Large differences can be seen in the epidemiology of VHD across different types of VHD and
between low and high income countries (352). Rheumatic heart disease (RHD) is the most
common cause of morbidity and mortality in low income countries while calcific aortic valve
disease (CAVD) carries the greatest burden of VHD in high income countries (352). A review
reported the prevalence of RHD between 46 per 100,000 in northern India while higher
prevalence can be seen in the Solomon Islands, 2400,000 per 100,000 (353). In contrast, in
the United States, aortic stenosis (a spectrum of CAVD) accounts for 45% of all deaths from
VHD and was the main driver of VHD-related deaths over the past 30 years (354). The risk of
ischaemic stroke can be up to 17 times greater in AF patients with rheumatic heart disease
compared to patients with AF alone without any significant valvular heart disease (355).
However, limited information is available on the prevalence and incidence of AF associated
with VHD. The RELY-AF registry (enrolled AF patients at 164 sites in 46 countries) reported
the presence of RHD among AF patients as 31.5% in India, 21.5% in Africa and 2.2% in North
America (356). Whereas the ORBIT-AF registry, a multicentre, prospective, outpatient hospital
registry included 176 US practices with 9748 AF patients, demonstrated a prevalence of
27.7% with significant VHD in this population. Among them, 4.1% had mitral
stenosis/mechanical valve, 4.7% had bioprosthetic valves or balloon valvuloplasty or prior
valve repair while a higher prevalence (18.9%) of patients with aortic regurgitation/aortic
stenosis, mitral regurgitation or tricuspid regurgitation (357) was found, consistent with other
high income countries (354).
1.7.2 Anticoagulation therapy in AF patients with valvular heart disease
Thrombotic events are the most common cause of mortality and morbidity especially after
surgery for VHD. This risk is especially higher within the first 3 months, for both bioprosthetic
and mechanical devices (348). Therefore, antithrombotic therapy is required to prevent
thrombotic events in VHD patients after surgical intervention. Effective measures should also
114
be made to control modifiable risk factors (for example effective blood pressure control in
hypertensive patients) to reduce the risk of thromboembolism, together with the prescription
of antithrombotic drugs (358). Anticoagulation therapy is required lifelong in patients receiving
a mechanical valve as this confers a life-long thrombotic risk. In addition, AF, a common
arrhythmia in VHD, also requires life-long anticoagulation (3, 348, 349, 358). Thus, patients
with both AF and mechanical/bioprosthetic device are at risk of thromboembolic and bleeding
complications if their anticoagulation therapy is not well optimised (348).
1.7.2.1 Anticoagulation therapy in mechanical heart valves
Exposure to the artificial valve and tissue injury caused by the presence of a prosthetic valve
activates the intrinsic and extrinsic coagulation pathways thus inducing the formation of
thrombin which in turn facilitates thrombus generation (349, 351). This condition is more
pronounced in AF patients, patients with mild to severe stenosis and aortic stenosis, as all
these conditions affect the blood flow turbulence thus triggering the coagulation cascade,
intensifying the propensity to thrombosis (349, 351). VKA antagonists prevent the coagulation
cascade at both the intrinsic and extrinsic pathways thus preventing the formation of thrombin.
The European guidelines (3, 349, 351) recommend the use of VKA in patients undergoing
mechanical valve transplantation (regardless of presence of AF) and AF patients with
moderate to severe mitral stenosis with good anticoagulation control, TTR >65-70% (3, 349,
351). The use of VKA should be monitored according to the INR range and targets based on
the prosthesis thrombogenicity and patient-related risk factors (Table 1.31). Treatment
duration with anticoagulation therapy depends on several factors. Patients with mechanical
valves and those with bioprosthetic valves or native valve disease (aortic and mitral stenosis,
aortic, mitral and tricuspid regurgitation) with additional (≥2) stroke risk factors require lifelong
anticoagulation treatment (Class 1 recommendation) (3, 351). However, the current ACC/AHA
guideline (359) recommends VKA therapy with the addition of aspirin (at least 75-100 mg/day)
to all patients receiving mechanical valves (Class 1A recommendation), whereas ACCP (360)
115
recommends addition of aspirin only in high-risk patients with careful monitoring of the
bleeding risk (359, 360).
Table 1.31:Target INR values for VKA among patients with prosthetic valves [taken
directly from (358, 368)]
Prosthesis Valve type INR target in INR target in

thrombogenicity patients with patients without
related factors risk factor (361)
≥1*(351)
Low Carbomedics. Medtronic 3.0 2.5
Hall, St Jude Medical, ON-
X
Medium Other bileaflet valves 3.5 3.0
High Lillehei-kaster, 4.0 3.5
omniscience, Starr-
Edwards, Bjork-Shiley and
other tilting-disc valves
*Risk factors: mitral or tricuspid valve replacement, previous thromboembolism, AF, mitral stenosis of any degree,
left ventricular ejection fraction <35%
A meta-analysis from 1994 (362) of 13,088 patients on antithrombotic therapy vs. no
antithrombotic therapy investigated thromboembolic and bleeding complications among
patients receiving a mechanical heart prosthesis. In this study, the incidence of major
embolism (causing death, residual neurological deficit or peripheral ischemia causing surgery)
without any antithrombotic therapy was 4.0 per 100 patient-years (95% CI 2.9-5.2) while this
incidence was further reduced to 1 per 100 patient-years (95% CI 1.0-1.1) in patients with VKA
and 2.2 per 100 patient-years (95% CI 1.4-3.1) for those on antiplatelet therapy (362). For
bleeding complications, OAC therapy increased the incidence of total bleeding (cerebral,
intracranial, bleeding causing death or hospitalisation and minor bleeding); 1.9 per 100 patient-
years (95% CI 1.7-2.0) and the addition of antiplatelet therapy further increased the incidence
of total bleeding; 4.6 (95% CI 3.1-6.4) (362).
Another more recent meta-analysis of RCTs by Massel et al (363), comparing VKA alone vs.
combination VKA and antiplatelet, of 4122 mechanical valve replacement patients found a
reduced risk of TE event [OR 0.43 (0.32-0.59, p<0.00001] and mortality [OR 0.57 (0.42-0.78;
116
p=0.0004)] in patients with the combination therapy compared to VKA alone. However there
was increased risk of major bleeding with the addition of antiplatelet therapy [OR 1.58 (1.14-
2.18); p=0.006)] compared to anticoagulation therapy (363).
The use of NOACs in patients with mechanical valve prosthesis is contraindicated (351, 358).
There is only one trial, the RE-ALIGN trial, a phase 2 dose study (364) of patients with
mechanical heart prosthesis (aortic and mitral) that evaluated the use of dabigatran versus
warfarin. Based on the kidney function, the selection dose of dabigatran was 150, 220 or
300mg twice daily to achieve a trough level of 50ng/ml which is based on the pharmacokinetic
model from the RE-LY trial. Unfortunately, the trial was terminated early after randomisation
of 252 of 405 patients because there was an excess of thromboembolic and bleeding events
in the dabigatran group (364). Several explanations have been suggested including
inadequate concentration of dabigatran in plasma, varied pharmacodynamics properties of
dabigatran and warfarin and over reactive contact coagulation pathway in the early
postoperative period induced by the sewing ring (351). No other studies of NOACs (factor Xa)
has been tested subsequently. Due to this, currently all patients with mechanical valve
prosthesis should be anticoagulated with a VKA (3, 349, 351).
117
1.7.2.2 Anticoagulation therapy in bioprosthetic valves
Reports have shown that thromboembolism associated with bioprosthetic valves ranged from
0.2%-3.3% per year (360, 365, 366) whereas higher risk can be found among the valves in
the mitral, compared to the aortic, position. Similar to that seen in mechanical prostheses, but
to a lower extent, TE risk is also higher within the first 3-months post-surgery (367, 368). Due
to that, compared to OACs, the use of low dose aspirin is now favoured for those with surgical
aortic bioprosthesis (without other indications for OAC, for example AF) for the first three
months post-surgery; however, this is based on a low-level of evidence (class IIa, level C)
(358). OACs may be considered for the first 3 months after surgery in aortic bioprosthesis
patients with class IIB, level C evidence (358, 369-371). In contrast, those with mitral or
tricuspid valve replacement with bioprosthesis should still be considered for OAC therapy for
the first 3 months after surgical intervention (358).
Lifelong anticoagulation therapy is required in patients with bioprosthesis and with additional
risk factors such as venous thrombosis, AF, hypercoagulable state and severely impaired left
ventricular function with low evidence (class I, level C) (351, 358). When long term
anticoagulation therapy is needed, VKA should be favoured in patients with bioprosthesis
(358). Despite the lack of RCT, (351, 372-374) NOACs can be used instead of warfarin in AF
patients with bioprosthesis after the third month of the post-operative period (351, 358).
The Dabigatran Versus Warfarin After Bioprosthesis Valve Replacement for the Management
of Atrial Fibrillation Postoperatively (DAWA) pilot study (375) compared dabigatran vs.
warfarin post bioprosthetic valve replacement in AF patients however due to low enrolment
(N=27; 15 dabigatran and 12 warfarin), the trial was terminated earlier and no concrete
conclusion was made. One patient from the warfarin group and no patient from the dabigatran
group developed intra cardiac thrombus and ischaemic stroke respectively after 90 days of
randomisation (375). Other small studies (373, 374) also suggested NOACs as reasonable
alternatives to VKA therapy in patients with AF and VHD, but more studies are needed to
confirm its efficacy and safety profile (351).
118
The ARISTOTLE (376) and ENGAGE-AF (373) trials included 82 and 191 patients with a
bioprosthesis, respectively. In both trials, the incidence of stroke/systemic embolism was
similar; between 1.19-2.9%/year for those receiving either apixaban or low/high dose
edoxaban compared to 1.7%/year incidence with warfarin therapy (351, 373, 376). Compared
to warfarin, edoxaban low and high dose was associated with similar risk of stroke/SE [HR
0.37 (0.10-1.42) and HR 0.53 (0.16-1.78) respectively compared to warfarin]; however only
low dose edoxaban was associated with a lower risk of major bleeding [HR 0.12 (0.01-0.95)];
low and high dose of edoxaban with lower risk of all-cause mortality [high dose edoxaban vs.
warfarin: HR 0.46 (0.23-0.91) and low dose edoxaban vs. warfarin: HR 0.43 (0.21-0.88)
respectively] (351, 373, 376).
1.7.2.3 Anticoagulation control in patients with valvular heart disease (VHD)
To date, only four (377-380) studies evaluated anticoagulation control by TTR in patients
undergoing valve replacement therapy. Two studies from Italy and Denmark (377, 378) that
evaluated TTR among mechanical heart valve patients showed relatively low TTR ranging
from 55-60% while very good TTR of 71-73% can be seen in another two Swedish studies
(379, 380).
Meanwhile, in 2002 one study (381) assessed the effect of anticoagulation control on long
term survival after valve replacement with a Medtronic Hall valve among 1532 patients
receiving single valve replacement at either the aortic or mitral position. Anticoagulation control
variability (ACV) was used and defined as the percentage of INRs outside the 2.0-4.0 range
[patients with aortic valve replacement (AVR) had a range of 2.0-3.0 while patients with mitral
valve replacement (MVR) had a higher range of 3.0-4.0]. The ACV was further divided into 3
equal sized groups: low (0-19.9% ACV), intermediate (20-29.9% ACV) and high ACV (≥30%
ACV); with higher ACV reflecting poor anticoagulation control. Overall, 75.5% of the collected
INRs were within the target range. Of those INRs outside of the therapeutic range, 12.0% were
below 2.0 and 12.5% were above 4.0. Survival at 15 years was reduced in the high ACV group
119
with AVR but was similar among the low and intermediate group (28% vs. 59% vs. 55%;
p<0.001, respectively). Similarly, as with MVR at 15 years, the survival rate was 56%, 42%
and 24%; p<0.001 for the low, intermediate and high ACV groups, respectively; also,
significantly reduced in the high ACV group. On multivariate analysis, ACV per 20% increase
was associated with increased mortality [HR 1.8, coefficient 0.595; p=0.001]. In this study,
although the quality of anticoagulation control was not assessed via the Rosendaal or the
percentage of INRs in range methods, better quality of anticoagulation assessed by ACV was
associated with reduced mortality (381).
In essence, VHD patients, particularly those with surgical prostheses (regardless of AF), are
at increased risk of thromboembolic complications. Long term anticoagulation therapy with
VKA is recommended for those with a mechanical prosthesis; whereas those with a
bioprosthesis (without additional risk factors) require at least three months of anticoagulation
therapy (VKA/NOACs) after surgical intervention. Measures should be taken to ensure the
quality of anticoagulation control is optimised level to prevent TE and bleeding complications.
120
1.8 Aims and objectives
This thesis will include three studies with the main objective of examining anticoagulation
control in AF patients from different cohorts. Specific objectives are outlined below. Secondary
objectives will be mentioned in each specific study.
1. The original aim was to assess the impact of at behavioural-educational intervention
(TREAT-2) on TTR, among patients identified as less likely to establish and maintain
adequate TTR (SAMe-TT2R2 score >2), receiving warfarin and comparing these
patients against those receiving warfarin and usual care alone (SAMe-TT2R2 score 0-
2). However, due to a change in clinical practice regarding the prescription of NOACs
instead of warfarin since 2015, the original aim of this study changed with the focus on
assessing depression, anxiety, knowledge about AF, beliefs about medication, and
quality of life among newly anticoagulated AF patients in this cohort [see Chapter 2].
2. To examine the quality of VKA control (measured by TTR), predictors of
anticoagulation control, and the relationship between INR control and adverse clinical
outcomes [thromboembolic (stoke/TIA and systemic embolism), bleeding events,
cardiovascular hospitalisation, all-cause mortality and ≥1 composite endpoints
(MACE)] in AF patients in a multi-ethnic cohort [see Chapter 3].
3. To examine the quality of VKA control (measured by TTR), predictors of
anticoagulation control, and the relationship between INR control and adverse clinical
outcomes [thromboembolic (stoke/TIA and systemic embolism), bleeding events,
cardiovascular hospitalisation, all-cause mortality and ≥1 composite endpoints
(MACE)] in patients with operated valvular heart disease (VHD), with and without AF
[see Chapter 4].
121
Chapter 2. A prospective study examining non-vitamin K antagonist
oral anticoagulants (NOACs) versus warfarin based on the SAMe-
TT2R2 score strata in anticoagulant-naïve patients with atrial
fibrillation: the TREAT-2 study
2.1 Abstract
Introduction:
To ensure efficacy and safety of OAC therapy with VKA (for example warfarin), the therapeutic
range of INR 2.0-3.0 must be achieved. In clinical practice, this is often poorly controlled and
could be due to poor adherence, inadequate knowledge and awareness of the importance of
OAC therapy. The TREAT intervention (219, 382), a one-off educational-behavioural session,
demonstrated a significant improvement in the TTR compared to patients receiving usual care
alone.
Objective:
The original aim of this study was to examine the impact of a behavioural-educational
intervention (TREAT-2) on TTR among warfarin patients with SAMe-TT2R2 score >2 (those
predicted to have poor response to warfarin therapy) and compare their TTR among patients
with SAMe-TT2R2 score 0-2 (those predicted to have good response warfarin therapy).
Medication adherence via pill count method was assessed for patients receiving NOAC
therapy. Secondary endpoints included assessment of patients’ depression, anxiety,
knowledge of AF and its treatment, beliefs about medication and quality of life using validated
questionnaires at baseline and six months follow up. Due to a change in clinical practice
regarding the prescription of NOACs instead of warfarin since 2015, there were insufficient
patients initiated on warfarin therapy within the Trust. Therefore, the comparison of the impact
122
of the TREAT-2 intervention on TTR among warfarin-treated patients could not be examined.
Instead the results focus on the secondary outcomes.
Methods:
Prospective, observational and longitudinal study design was employed among patients
newly-initiating OAC therapy for stroke prevention in AF (either warfarin or a NOAC). The
Patient Health Questionnaire-9 (PHQ-9), Generalised Anxiety Disorder-7 (GAD-7), AF
knowledge scale, Beliefs about medication (BMQ) and Atrial Fibrillation Effect on Quality of
Life (AFEQT) questionnaires were completed among 139 newly anticoagulated AF patients at
baseline and 105 patients at six months follow up. The parameters were compared
descriptively between warfarin and NOACs patients at both time points. The change in scores
between baseline and follow-up were analysed in those who completed the questionnaires at
both time points (N=105).
Results:
At baseline the overall median (IQR) depression and anxiety scores were 4.0 (1.0-8.0) and
1.0 (0-5.0) respectively. The mean (SD) AF knowledge score was 5.7 (1.7). Patients had
positive perceptions about their medications evident by the mean (SD) positive necessity-
concern differential 5.8 (4.1) and poor overall quality of life score, 66.7 (53.7-77.8). There were
no significant differences in the depression, anxiety and beliefs about medication scores over
time. However, significantly higher proportions of patients answered correctly in the question
assessing the consequences of AF (88.6% vs. 50.5%; p<0.001) and symptoms score from
the quality of life questionnaire has significantly improved at follow up compared to baseline
(83.3 vs. 79.2; p=0.02)] respectively. Median (IQR) TTR at follow up for patients on warfarin
was 77.3% (54.4-84.7) and adherence for 70% of NOAC patients via pill count method was
100%.
123
Conclusion:
Newly anticoagulated AF patients appear to have low levels of anxiety, depression, poor AF
knowledge, positive perceptions about their medication and poor overall quality of life at
baseline and these parameters remained the same after six months of follow up. However,
significantly more patients were aware of the consequences of AF and symptoms of AF (by
the AFEQT questionnaire) improved at follow up. Future research is required in order to
determine the impact of educational and behavioural interventions towards improving
knowledge, emotional health, quality of life and thus preventing adverse clinical outcomes in
patients receiving different types of OAC therapy.
124
2.2 Background and rationale
Adequate OAC, either with VKA or NOAC is essential for effective stroke prevention in patients
with AF with ≥1 additional stroke risk factor (173). NOACs offer efficacy, safety and relative
convenience compared to the VKAs (130). However, cost considerations result in variable
policies regarding NOACs in different healthcare systems, ranging from unrestricted
reimbursement at one end of the spectrum to full payment by AF patients at the opposite end.
Some healthcare systems have even adopted ‘conditional authorisation’ of NOACs
prescription, based on (say) 6 months of low quality of anticoagulation with VKAs, as
measured by the TTR during a pre-defined period of VKA trial, whilst others still search for
optimal criterion for patient selection in the local setting. Indeed, the SAMe-TT2R2 score could
help aid individual decision-making regarding the choice between VKAs or NOACs in routine
clinical practice.
Previous studies by the Birmingham group (284, 383, 384) and others (385, 386) have found
that many AF patients possess very little knowledge of their disease and do not understand
the risks and/or benefits of anticoagulant therapy, particularly among ethnic minority patients
(383, 384). This may contribute to poor INR control, given the complexity of the warfarin
regimen, with dosing adjustments, drug-, food- and alcohol- interactions. Few studies have
intervened to improve adherence with, and understanding of, warfarin therapy. Thus, a more
structured education intervention among patients who are predicted to be less likely to achieve
good INR control (SAMe-TT2R2 score of >2) may be an alternative treatment strategy which
could help improve their TTR.
Patient education has been found to improve INR control (296). Indeed, an earlier pilot study
of a brief educational intervention (284) demonstrated a significant improvement in the
awareness of target therapeutic INR (p<0.0001) and factors which may affect INR levels
(p=0.005), with a trend towards improvement in awareness of the benefits of anticoagulants
125
and bleeding risks. Further, the TREAT intervention (see Table 2.1 for the components of the
intervention), a one-off educational-behavioural session, delivered by a health psychologist,
demonstrated a significant improvement in the TTR compared to patients receiving usual care
alone (76.2% vs. 71.3% respectively; p=0.035) (219, 382).
However, a recent Cochrane systematic review (387) assessing the impact of knowledge and
behavioural intervention on TTR showed equivocal results. The mean difference of TTR
between patients receiving educational and self-monitoring interventions compared to usual
care was 6.31 (95% CI -5.63 to 18.28). This suggest that although TTR appears to be higher
in the intervention group vs. usual care, analysis of the pooled data was not in favour
(statistically) of the former compared to the latter. Nonetheless, these results were based on
only two trials (N=69) with very low quality of evidence, assessing the impact of self-monitoring
and education intervention vs. usual care on TTR. Thus, further trials are needed to
investigate the benefits of similar interventions (both educational and behavioural) on
anticoagulation control in AF patients.
Improving understanding about a disease and its treatment allows patients to make informed
decisions about the management of their condition and treatment may make a significant
difference to clinical outcomes. Whilst NOACs are a valid alternative to warfarin, the latter will
still continue to be used as a treatment for AF, and interventions that can improve
anticoagulation control are essential to reduce the risk of adverse events.
126
Table 2.1: Components of the TREAT intervention
Content
Educational Booklet • AF causes and consequences
• Warfarin and its metabolism
• Stroke risk and risk of bleeding on treatment
• Lifestyle changes (diet, alcohol, lifestyle changes)
Patient DVD • AF: causes, consequences, side effects, treatment options

• Warfarin: INR monitoring, lifestyle changes
Delivered by ‘expert • Patient barriers: psychological, physical
patient’ narratives • Consultant Q&A: common questions and answers
and consultant Q&A
Patient worksheet Including: Calculate your own risk of stroke; personal barriers to
warfarin uptake; and discussion of personal goals for lifestyle
changes
Self-monitoring diary Two-week diary monitoring including: Diet; Alcohol intake (in units);
Medications; and INR outcomes
2.2.1 Study objective

The initial objective of this study was to perform a prospective observational intervention of
NOAC versus warfarin based on SAMe-TT2R2 score strata in anticoagulant-naïve patients with
AF. In addition, to evaluate the impact of the TREAT-2 educational and behavioural
intervention on TTR among patients identified as less likely to establish and maintain adequate
TTR (SAMe-TT2R2 score >2) receiving warfarin and comparing these patients against those
receiving warfarin and usual care alone (those with a SAMe-TT2R2 score ≤2).
Secondary objectives included assessment of patients’ depression, anxiety, knowledge of AF
and its treatment, beliefs about medication and quality of life using validated questionnaires
at baseline and six months follow up.
127
2.3 Methods
Study design
Prospective observational study with a 16-month inclusion period and a 6-month follow-up.
Patients
Anticoagulant-naïve AF patients referred for OAC therapy (see Figure 2.1), were recruited
from three different sources: (1) OAC clinic at Sandwell and West Birmingham Hospitals
(SWBH), (2) AF/Cardiology Clinic at SWBH and (3) OAC clinic, University Hospitals
Birmingham (UHB).
Inclusion criteria
Male and female adult (aged ≥18 years) patients with electrocardiographically documented
AF without Evaluated Heart valves, Rheumatic or Artificial (EHRA) type 1 VHD, and at least
one additional risk factor for stroke (based on the CHA2DS2VASc score)(146), who were OAC-
naïve (having never taken OAC) and eligible for OAC were considered for inclusion. Men with
a CHA2DS2VASc score of ≥1 and women with a CHA2DS2VASc score of ≥2 were eligible for
OAC therapy.
Exclusion criteria
The exclusion criteria were: (1) aged <18 years old, (2) any contraindication to OAC, (3)
prosthetic cardiac valve or significant VHD with an indication for heart surgery, (4) likelihood
of intermittent or permanent discontinuation of OAC during follow-up (e.g., due to major
surgery or post-AF ablation), (5) active malignancy, (6) cognitive impairment, (7) any disease
likely to cause their death within 6 months and (8) unable to provide written informed consent.
128
Ethical approval
This study involved patients from two NHS sites, thus REC approval was applied for. Ethical
approvals were obtained from the West Midlands South Birmingham Research Ethics
Committee [REC; (REC reference: 16/WM/0339)], the Health Research Authority [HRA; (IRAS
ID: 193145); 26th September 2016), and also SWBH Research and Development (R&D; R&D
reference: 16CARD06;13th October 2016). The University Hospitals Birmingham was added
as a site later and approval from the REC and UHB R&D was obtained (Reference number
RRK6149; 12th December 2017). Approval letters can be found in Appendix 5. The University
of Birmingham acted as the sponsor for this study.
Patient recruitment
Recruitment of patients lasted for 16 months starting from 15th October 2016 to 3rd March
2018. However, the process of obtaining ethical approval at the UHB site began later on and
was a lengthy process. Thus, in the 3-month period of recruitment at the UHB site (3rd January
– 3rd March 2018), only a limited number of patients were recruited (25 patients were screened
for eligibility and only nine patients agreed to participate) (Figure 2.1).
2.3.1 Procedure
After initiation of OAC (either warfarin or a NOAC) therapy, patients were seen by an
anticoagulant nurse/healthcare professional for an educational session regarding their AF and
anticoagulation therapy as per usual care. After that, patients were approached by the
researcher to discuss the study and if the patient agreed to participate, written informed
consent was obtained and the baseline questionnaire was administered. Individual baseline
SAMe-TT2R2 score was calculated and patients were allocated to one of four groups on the
basis of their SAMe-TT2R2 score and OAC (warfarin or NOAC) (see Figure 2.1). Patients with
a SAMe-TT2R2 score ≤2 who were prescribed warfarin (dose-adjusted to achieve a target INR
of 2.0 to 3.0) were assigned to Group 1 and patients with a SAMe-TT2R2 score ≤2 who were
129
prescribed NOAC were assigned to Group 2. Patients with a SAMe-TT2R2 score >2 who were
prescribed warfarin were assigned to Group 3, to receive the intensive education (TREAT-2).
Patients with a SAMe-TT2R2 score >2 who were prescribed NOAC (apixaban, dabigatran,
edoxaban, or rivaroxaban) were assigned to Group 4. Baseline demographic and clinical
information, including medical, medication history and laboratory information was recorded
from the hospital records onto a proforma (Appendix 1, A1.1). All patients were informed
about AF and the need for anticoagulant therapy by a healthcare professional using the
standard warfarin or NAOC-specific education checklist at baseline as per usual care. All
patients on warfarin also received the standard Yellow book to identify that they were on
warfarin.
Patients in Group 3 would receive a group intervention (between 2-4 patients plus carer/family
member) based on the session developed for the TREAT study delivered by the researcher
within 4 weeks of warfarin initiation. In addition, patients would receive an educational booklet,
self-monitoring diary, worksheet and alert card.
INR monitoring
INR monitoring was performed by the Anticoagulation Services at SWBH and UHB. All
patients who received warfarin (usual care and intensive education arms) attended the
anticoagulant outpatient clinic at the respective hospital to have their INR checked using a
capillary sample. The frequency of the INR visits was at the discretion of the OAC clinic (the
OAC clinic staff were blinded to the intervention arm the patient is allocated to enable as
‘naturalistic’ as possible follow-up and monitoring). Every INR result from baseline to the end
of the study (6-months) was recorded. The proportion of time each patient spent in the
therapeutic INR range (2.0 to 3.0) (TTR) was calculated by the Rosendaal (using linear
interpolation method between two consecutive INR values) and the percentage of INR in
range (PINRR) methods (dividing the number of INRs that falls in range by the total number
of INR tests). INR data was utilised from months 1 to 6 follow-up (to allow attainment of the
130
correct dose of warfarin during the first four weeks). More details of the calculation of TTR can
be seen in section 3.3.4.1.1, page 178.
Six-month follow-up
All patients were followed up via phone call at 6 months. Patients receiving NOACs were
asked to post back all their pill boxes and blister packs to the researcher. Medication
adherence of patients in Groups 2 and 4 were reviewed at 6- months via pill count.
The battery of questionnaires (PHQ-9, GAD-7, AF knowledge scores, BMQ and AFEQT) was
posted to all participants at follow up with a stamped addressed envelope for return. If returned
questionnaires were not fully completed, the researcher contacted the patient by telephone to
facilitate 100% completion of the questionnaires. Patients were sent a reminder questionnaire
at both baseline and follow up if they did not respond within 2-3 weeks of receiving the original
set of questionnaires.
2.3.2 Assessment of medication adherence
Medication adherence was assessed among patients receiving NOAC therapy via the pill
count method. Patients were asked to keep their NOAC boxes and blisters from the point they
were recruited into the study until six months of follow up. They were also given stamped-
addressed envelopes to return their empty pill boxes and blister packages. Upon receiving the
packages, the number of pills remaining in the blister packages were counted. In addition to
pill counting, the patients were also asked two questions to assess adherence via a phone
call: 1. ‘Do you sometimes forget to take your blood-thinning pills?’ (yes/no answer); 2. ‘Over
the past 2 weeks, were there any days when you did not take your blood-thinning medicine?’
(yes/no answer). These two questions were adapted from the self-report questions from the
Morisky Green Levine adherence scale, which showed concurrent and predictive validity on
131
blood pressure control in majority of patients with good adherence to antihypertensive
medications (388). Although the pill count method is an indirect measure of adherence, it has
higher accuracy compared to other subjective methods (389), cost effective (389) and is
commonly used in clinical trials (390, 391).
132
Patient recruitment
Oral anticoagulant AF/cardiology Oral anticoagulant

(OAC) clinic, SWBH clinic, SWBH (OAC) clinic, UHB
OAC-naïve patients with non-valvular

atrial fibrillation
Physician to start OAC of choice in clinic
Patient selection based on

SAMe-TT2R2 score
N=256
Observational
follow up through- SAMe-TT2R2 SAMe-TT2R2
out 6 months score ≤2 score >2 Physician choice
to assign to
group 3 or 4
Group 3 Group 4
Group 1 Group 2
VKA* plus
VKA* NOAC TREAT NOAC
intervention
N=64 N=64
N=64 N=64
*All INR measurements under the jurisdiction of SWBH/UHB OAC clinic
Figure 2.1: Study design and patient selection flow chart
133
2.3.3 Study outcomes
The primary outcome was the proportion of time spent in the therapeutic INR range, 2.0 to
3.0, at 6 months for the two groups commencing warfarin and medication adherence (via pill
count) for patients receiving a NOAC. The following secondary outcomes were examined: (1)
depression, (2) anxiety, (3) patients’ knowledge of AF, (4) beliefs about medication and (5)
quality of life.
Ancillary descriptive analyses explored the incidence of bleeding, stroke/TIA, cardiovascular
hospitalisation and death (given that the study was not powered to detect these differences).
Exploratory analyses determined whether the incidence of bleeding, stroke/TIA,
cardiovascular hospitalisation, death and composites (≥1) of these events were similar in
patients with SAMe-TT2R2 0-2 compared to SAMe-TT2R2 >2.
The number of patients with strokes/TIA, bleeding, thromboembolic, CV hospitalisations and
death events were determined from hospital records. Stroke was defined as a focal neurologic
deficit, from a non-traumatic cause, lasting at least 24 hours and further categorized as
ischemic (with or without haemorrhagic transformation), haemorrhagic, or of uncertain type (in
the case of patients who did not undergo brain imaging or in whom an autopsy was not
performed). Systemic embolism was defined as a thromboembolic event outside the brain,
retina, heart or lungs. Stroke and systemic embolism were later combined as thromboembolic
events (TE). Major bleeding was classified according to the ISTH criteria (392): fatal bleeding,
and/or symptomatic bleeding in a critical area or organ, such as intracranial, intraspinal,
intraocular, retroperitoneal, intra-articular or pericardial, or intramuscular with compartment
syndrome, and/or bleeding causing a fall in haemoglobin level of 20 g/L or more, or leading to
transfusion of two or more units of whole blood or red cells) (392). Clinically relevant non-
major bleeding (CRNMB) was defined as clinically overt bleeding that did not satisfy the criteria
for major bleeding and that led to hospital admission, physician guided medical or surgical
treatment, or a change in antithrombotic therapy (392). Major bleeding and CRNMB were
combined as bleeding events. Cardiovascular hospitalisation was defined as a hospitalisation
134
with a cardiovascular cause: i) heart failure, MI, new angina, non-fatal cardiac arrest,
ventricular arrhythmia, uncontrolled AF/atrial flutter, supraventricular arrhythmia, ii) valve
surgery, coronary artery bypass graft (CABG) surgery, percutaneous transluminal coronary
angioplasty (PTCA) surgery, pacemaker/ implantable cardioverter-defibrillator (ICD) insertion,
carotid endarterectomy, peripheral angioplasty/surgery, limb amputation (393) AND as
recorded in patient’s medical records.
2.3.3.1 Questionnaires
The following questionnaires were given to the patients to complete at baseline and 6-months
later to assess depression and anxiety, knowledge of AF, beliefs about medication, and quality
of life (see Appendix 2 for full questionnaires). These questionnaires were chosen as they
are validated questionnaires and have been used in other cohorts of chronic diseases
including asthma, diabetes and cardiovascular diseases including AF (394-396) (397) (398)
(399).
2.3.3.1.1 Patient Health Questionnaire (PHQ-9)
Depression was assessed using the Patient Health Questionnaire (PHQ-9) (400). The PHQ-
9 is a 9-item scale that contains the diagnostic criteria for depression based on the DSM-IV
criteria. It is calculated by assigning scores of 0 to 3 to the response categories of “not at all,”
“several days,” “more than half the days,” “and nearly every day,” respectively; scores range
from 0-27. Scores of 0 indicated no depression and ≥15 signify the presence of major
depression. Scores of 5, 10, 15, and 20 represents thresholds demarcating the lower limits of
mild, moderate, moderately severe and severe depression, respectively (400).
135
2.3.3.2.1 The Generalised Anxiety Disorder 7-item (GAD-7)
The Generalised Anxiety Disorder 7-item (GAD-7) questionnaire was used to assess anxiety.
It is a 7-item scale and is calculated by assigning scores of 0 to 3 to the response categories
of “not at all,” “several days,” “more than half the days,” “and nearly every day,” respectively,
and adding the scores together. A total score for the 7 items ranges from 0 to 21. Scores of 0
indicated no anxiety and ≥10 represents the presence of generalised anxiety disorder (GAD).
Scores of 5, 10 and 15 are taken as the cut off points for mild, moderate and severe anxiety,
respectively (401). The GAD-7 and PHQ-9 scores were chosen because they were
recommended by the NICE guidelines as one of the valid measures of anxiety and depression
severity (400, 401) among primary care patients (402, 403). Furthermore, both scales have
been validated in patients with cardiovascular diseases, (394-396) with good sensitivity and
specificity to detect major depressive disorder (81% and 77%, respectively) (395) and
generalised anxiety disorder (75% and 89%, respectively) (404).
2.3.3.3.1 Knowledge of AF
Patients’ knowledge of AF was assessed using the Atrial Fibrillation Knowledge scale. (397)
This scale consists of 11-items concerning AF in general (3-items), symptoms recognition, (3-
items) treatment (3-items) and general attitudes towards AF (2-items). For each question,
patient can choose one of three options; only one answer is correct. Scores range from 0 to
11 or 0-100%, with higher correct scores denoting better knowledge of AF(397). This scale
has a border line reliability score with Cronbach α of 0.58 in its original derivation cohort. This
scale combines three important aspects of AF knowledge: AF management, symptoms and
antithrombotic therapy. It can also be used to detect gaps in knowledge and attitude towards
AF management (397).
136
2.3.3.4.1 Beliefs about Medicine Questionnaire (BMQ)
To assess patients’ beliefs about medication, the Beliefs about Medicine Questionnaire (BMQ)
was used (398). This is an 18-item questionnaire consisting of two parts, one assessing
patients’ belief about their own medicine (BMQ-specific) and the other assessing patients’
beliefs about medicine in general (BMQ-general). The BMQ-specific covers two themes the
specific-necessity theme evaluates patients’ view about the importance and necessity of their
medicines, whilst specific-concern theme comprises patients’ beliefs about potential harm and
adverse effect of their own medicines. Each sub-scale has a score ranging from 5 to 25.
Patients can choose if they ‘strongly disagree’, ‘disagree’, ‘uncertain’, ‘agree’ and ‘strongly
agree’ with the statements regarding their view of medicines. A high score on the ‘necessity’
theme indicates that patients think their medicines are important to them; a high score on the
‘concern’ theme means that patients are worried and concerned about their own medicines.
The difference between the necessity and concern domain is obtained by subtracting the two
values. Positive values indicate that patients perceive their medication as more important than
their concerns about potential side effects of the medication and vice versa for negative
values. Likewise, BMQ-general part has two themes as well; general overuse theme assesses
how patients perceive the extent of medicine usage, and the general harm theme represents
patients’ beliefs about the harmful nature of medicines in general. The scores of the last two
themes range from 4 to 20; a high score in each theme means negative perception about
medicines in general (398). This scale is valid, reliable and has been validated in AF and other
cardiovascular diseases, renal, and diabetic populations (398).
2.3.3.5.1 The Atrial Fibrillation Effect on Quality–of-life (AFEQT)
The Atrial Fibrillation Effect on Quality–of-life (AFEQT) questionnaire was used to assess AF
patients’ quality of life (405). It is a 20-item scale that is further divided to assess symptoms
(4-items), daily activities (8-items), treatment concern (6–items) and lastly treatment
137
satisfaction (2–items). Responses were expressed using a 7-point Likert scale ranging from
the most severe limitations/symptoms to no limitation/symptom. The raw score of 1 to 7 was
transformed to a 0 to 100 scale. A score of 0 indicates lowest quality of life and a score of 100
indicates highest quality of life. Thus, higher scores on the AFEQT instrument indicate better
health status. The responsiveness of this instrument is its ability to detect clinically meaningful
changes in a patients’ health status over time. Changes in the AFEQT overall and domain
scores from baseline to subsequent 6 months were used to evaluate responsiveness to
change over time (405). The AFEQT questionnaire was shown to adequately assess quality
of life in AF populations evident by demonstrating robust content validity in one systematic
review (399).
2.3.4 Hypothesis
It was originally hypothesised that patients with a SAMe-TT2R2 score >2 receiving warfarin
who received the TREAT-2 intervention (Group 3) would have a significant improvement in
TTR compared to those patients with a SAMe-TT2R2 score 0-2 receiving warfarin only.
Secondly, it was hypothesised that patients with a SAMe-TT2R2 score >2 would have more
depression and anxiety symptoms, poor knowledge of AF, negative beliefs or perception
towards medication and have poorer quality of life than those with SAMe-TT2R2 score 0-2.
Unfortunately, due to the change in clinical practice during the study, where prescription rates
for NOACs for stroke prevention in AF increased sharply from 2015, there were only 13
patients (Group 1 and 3) prescribed with warfarin therapy of which only four (Group 3) were
eligible for the TREAT-2 intervention. However, these four patients did not agree to participate
in the TREAT educational intervention within 4 weeks of warfarin initiation; but only agreed to
do the questionnaire (Figure 2.2). Therefore, the planned analysis of comparing TTR between
Group 1 (warfarin and usual care) and Group 3 (warfarin and TREAT-2 educational
intervention) could not be undertaken. Hence the analyses focus on the secondary outcomes
from the questionnaire data, adverse events and medication adherence in NOAC patients.
138
Patients were grouped according to the type of OAC they received (warfarin or NOAC) for the
questionnaire analyses.
Research questions
• Do patients with a SAMe-TT2R2 score >2
o Have poorer TTR?
o Have more depression and anxiety symptoms, poorer knowledge of AF and
quality of life and negative perceptions of medication?
o Have more thromboembolic and bleeding events, CV hospitalisations, death
and a composite of these events than those with SAMe-TT2R2 score 0-2?
Power calculations
Power for the primary endpoint of INR control, evidenced by TTR was calculated based on
the results of the TREAT (219) study. In the TREAT study, patients receiving the intensive
educational intervention (N=43) and usual care (N=54) had a mean (SD) TTR of 78.5% (20.1)
and 66.7% (21.8), respectively. Therefore, a sample size of 54 patients in each of the warfarin
groups will provide at least 90% power to detect similar differences at a significance level of
0.05. The same number of patients were recruited for the NOAC groups. To allow for a 20%
attrition rate, 64 patients per group were needed, resulting in a total sample size of 256
patients. Data was analysed using IBM SPSS for Windows (Version 23.0) (406).
2.3.5 Statistical analysis
Following a test of statistical normality, by histogram plot method and the Kolmogorov-Smirnov
test where a bell-shaped distribution in the former and p-values >0.05 in the latter were
indicative of normally distributed data. Continuous variables were presented as mean (SD)
and for non-parametric data, median with interquartile range (IQR, 25th to 75th quartile) were
reported. Categorical variables were reported as counts with percentages. Descriptive
139
statistics were presented for baseline demographic and clinical information. Categorical
variables were analysed using the chi-square statistic or the Fisher exact test (where expected
frequencies are less than five in any cell).
All data were analysed by intention-to-treat. The primary endpoint, TTR, was determined by
the method of linear interpolation using the Rosendaal and the PINRR methods, with INR data
from months 1 to 6. Overall TTR was also calculated incorporating all available INRs.
Differences in the overall TTR and TTR excluding the inception period (INR value from day 1
to day 30) were examined using the Wilcoxon-signed ranked test and were reported as median
(IQR) as they were not normally distributed. P-values <0.05 were considered statistically
significant.
Data for the secondary endpoints of (1) depression, (2) anxiety, (3) patients’ knowledge of AF,
(4) beliefs about medication and (5) quality of life at the two time-points (baseline, 6- months)
were presented descriptively and graphically to illustrate the change in these variables over
time. In this section, patients were grouped according to the type of OAC they received, either
warfarin or NOAC, instead of the original SAMe-TT2R2 groups. As mentioned previously, there
were too few patients in Groups 1 and 3 (N=9 and N=4, respectively) compared to Groups 2
and 44 (N=102 and N=24, respectively), thus making the comparison between SAMe-TT2R2
groups inappropriate. However, the results arranged by the original groupings are provided in
Appendix 3. Paired t-test and Wilcoxon-signed ranked test were used for normally and non-
normally distributed data, respectively, to investigate the changes for the questionnaire
variables over time for patients who completed the questionnaires at both baseline and six
months (N=105).
The number of patients with thromboembolism, major bleeding, CV hospitalisation and all-
cause death at 6- months follow up were presented as absolute numbers and percentages.
The events were also compared in relation to the SAMe-TT2R2 score categories (0-2 vs. >2)
and were presented as proportions.
140
2.4 Results
During the period of patient recruitment, 598 AF patients (573 SWBH and 25 UHB) were
screened for eligibility. One hundred and seventy-four patients refused to participate and 255
patients were not eligible due to OAC-experience (N=123), did not attend appointment
(N=111), no documented evidence of AF (N=14), cancer (N=4), and cognitive impairment
(N=3) (Figure 2.2). Of those eligible to participate (N=343), 169 (49.2%) agreed and gave
written informed consent. However, only 139 (40.5%) (SWBH 132 and UHB 7) patients
returned the baseline questionnaires. At six months follow up, 105 (75.5%) patients returned
their questionnaires, with only 67 (70%) NOAC patients returning the empty pill boxes and
blister packs (Figure 2.2).
2.4.1 Baseline demographics and clinical characteristics of AF patients
Table 2.2 presents the demographics and clinical characteristics of 139 AF patients who were
prescribed OAC therapy at baseline stratified by their baseline SAMe-TT2R2 score. There were
111 patients (79.9%) with the SAMe-TT2R2 score of 0-2, of which 9 patients were prescribed
warfarin (Group 1) and 102 patients were prescribed NOACs (Group 2). Only 28 patients
(20%) had a SAMe-TT2R2 >2, of which only 4 patients were prescribed warfarin (Group 3) and
24 patients received a NOAC (Group 4).
In the overall population, the mean (SD) age was 72.0 (8.5), 56.1% were male, the majority
were white (88.8%), married (55.4%), and had secondary school level education (76.3%).
Slightly more than half experienced no AF symptoms (54.7%) according to the modified
European Heart Rhythm Association (mEHRA) classification system and most (77.0%) had
paroxysmal AF. Hypertension (75.5%) was the most common co-morbidity followed by
diabetes (28.8%) and chronic kidney disease (26.6%) [defined as creatinine clearance (CrCl)
<60ml/min or as stated in the medical notes]. The overall mean (SD) CHA2DS2-VASc and
HAS-BLED scores were 3.3 (1.5) and 1.9 (1.1) respectively (Table 2.2).
141
Screened for eligibility N=598 (SWBH= 573; UHB =25)
Excluded, N=255:
• Not OAC naïve: 123
• Did not attend appointment: 111
• No AF: 14
• Cancer: 4
• Cognitive impairment: 3
Number of patients eligible: 343 Did not return

• Agreed to participate and gave consent, N=169 baseline
(160 SWBH; 9 UHB) questionnaire,
• Rejected: 174 N=30
Answered baseline questionnaire, N=139

(132 SWBH; 7 UHB)
SAMe-TT2R2 0-2 SAMe-TT2R2 >2
Group 1: Group 2: Group 3: Group 4:

N=9 N=102 N=4 N=24
Warfarin NOACs
N=13 N=126
Answered 6-month questionnaire,

N=105
Warfarin NOACs
N=9 N=96
Returned pill counts for NOAC

patients only, N=67
Figure 2.2: Flowchart of patient’s inclusion and follow-up in the study
142
Table 2.2: Baseline demographics and clinical characteristics of newly anticoagulated AF patients stratified by the SAMe-TT2R2
score

Overall Group 1 Group 2 Group 3 Group 4
N=139 N=9 N=102 N=4 N=24
Age at OAC Mean age (SD) 72.0 (8.5) 72 (6.2) 73.8 (7.9) 68.5 (13.3) 67.2 (9.0)
initiation
Age groups ≤64 21 (15.1) 0 10 (9.8) 1 (25.0) 10 (41.7)
65-74 65 (46.8) 7 (77.8) 48 (47.1) 2 (50.0) 8 (33.3)
≥75 53 (38.1) 2 (22.2) 44 (43.1) 1 (25.0) 6 (25.0)
Sex Female 61 (43.9) 4 (44.4) 41 (40.2) 2 (50.0) 14 (58.3)
Ethnic groups White 123 (88.8) 8 (88.9) 102 (100) 2 (50.0) 11 (45.8)
South-Asian 7 (5.0) 1 (11.1) 0 0 6 (25.0)
Afro-Caribbean 9 (6.5) 0 0 2 (50.0) 7 (29.2)
Marital status Married 77 (55.4) 5 (55.6) 55 (53.9) 2 (50.0) 15 (62.5)
Single 19 (13.7) 1 (11.1) 16 (15.7) 1 (25.0) 1 (4.2)
Divorced/separated 12 (8.6) 1 (11.1) 7 (6.9) 1 (25.0) 3 (12.5)
Widowed 31 (22.3) 2 (22.2) 24 (23.5) 0 5 (20.8)
Educational Primary school 5 (3.6) 0 3 (2.9) 0 2 (8.3)
status Secondary school 106 (76.3) 8 (88.9) 75 (73.5) 4 (100) 19 (79.2)
College 28 (20.1) 1 (11.1) 24 (23.5) 0 3 (12.5)
Age leaving Mean age (SD) 16.7 (5.8) 16.7 (3.6) 17.2 (6.2) 15.5 (0.6) 15.9 (4.7)
education
Alcohol intake Alcohol>14unit/day 23 (16.5) 0 20 (19.6) 1 (25.0) 2 (8.3)
Smoking status Smoking/ex-smoker 17 (12.2) 0 4 (3.9) 2 (50.0) 11 (45.8)
143
Table 2.2 continued
N=139 N=9 N=102 N=4 N=24
Modified EHRA Class 1 (none) 76 (54.7) 5 (55.6) 52 (51.0) 4 (100) 15 (62.5)
class Class 2 (mild) 40 (28.8) 3 (33.3) 31 (30.4) 0 6 (25.0)
Class 3 (severe) 23 (16.5) 1 (11.1) 19 (18.6) 0 3 (12.5)
AF type Paroxysmal 107 (77.0) 7 (77.8) 77 (75.5) 3 (75.0) 20 (83.3)
Persistent 9 (6.5) 0 7 (6.9) 0 2 (8.3)
Permanent 23 (16.5) 2 (22.2) 18 (17.6) 1 (25.0) 2 (8.3)
Past medical Heart failure 10 (7.2) 1 (11.1) 7 (6.9) 1 (25.0) 1 (4.2)
history Hypertension 105 (75.5) 6 (66.7) 76 (74.5) 4 (100.0) 19 (79.2)
Diabetes 40 (28.8) 3 (33.3) 28 (27.5) 2 (50.0) 7 (29.2)
Stroke/TIA 23 (16.5) 3 (33.3) 13 (12.7) 0 7 (29.2)
Vascular disease* 21 (15.1) 1 (11.1) 13 (12.7) 1 (25.0) 6 (25.0)
Lung disease# 26 (18.7) 3 (33.3) 15 (14.7) 1 (25.0) 7 (29.2)
Kidney disease† 37 (26.6) 4 (44.4) 24 (23.5) 2 (50.0) 7 (29.2)
Anaemia 29 (20.9) 1 (11.1) 25 (24.5) 0 3 (12.5)
Previous bleeding 7 (5.0) 0 6 (5.9) 0 1 (4.2)
CHA2DS2-VASc Mean (SD) 3.3 (1.5) 3.6 (1.9) 3.2 (1.3) 3.5 (2.6) 3.4 (1.7)
score
CHA2DS2-VASc Low risk 0 0 0 0 0
score categories 22 (15.8) 2 (22.2) 13 (12.7) 1 (25.0) 6 (25.0)
Intermediate
High risk 117 (84.2) 7 (77.8) 89 (87.3) 3 (75.0) 18 (75.0)
HAS-BLED score Mean 1.9 (1.1) 1.9 (0.8) 2.0 (1.1) 1.8 (0.5) 1.8 (1.2)
HAS-BLED score Low risk (0-2) 108 (77.7) 7 (77.8) 79 (77.5) 4 (100) 18 (75.0)
categories High risk (≥3) 31 (22.3) 2 (22.2) 23 (22.5) 0 6 (25.0)
SAMe-TT2R2 Mean (SD) 1.7 (1.2) 1.3 (0.7) 1.1 (0.7) 3.5 (0.6) 3.7 (0.8)
score
144
Table 2.2 continued
N=139 N=9 N=102 N=4 N=24
Current Warfarin 13 (9.4) 9 (100) 0 4 (100) 0
medications 126 (90.6) 0 102 (100) 0 24 (100)
NOACs
Beta-blocker 77 (55.4) 5 (55.6) 53 (52.0) 3 (75.0) 16 (66.7)
ACEI/ARB 72 (51.8) 3 (33.3) 54 (52.9) 2 (50.0) 13 (54.2)
Diuretics 44 (31.7) 3 (33.3) 29 (28.4) 2 (50.0) 10 (41.7)
Amiodarone 3 (2.2) 0 1 (1.0) 0 2 (8.3)
Concurrent antiplatelet 7 (5.0) 0 2 (2.0) 0 5 (20.8)
Digoxin 12 (8.6) 2 (22.2) 8 (7.8) 0 2 (8.3)
Calcium channel 54 (38.8) 3 (33.3) 39 (38.2) 2 (50.0) 10 (41.7)
blocker
Statins 98 (70.5) 7 (77.8) 68 (66.7) 3 (75.0) 20 (83.3)
Group 1: SAMe-TT2R2 score 0-2 + warfarin; Group 2: SAMe-TT2R2 score 0-2 +NOAC; Group 3: SAMe-TT2R2 score >2 +warfarin; SAMe-TT2R2 score >2+NOAC
ACEI/ARB: angiotensin converting enzyme inhibitor/ angiotensin receptor blockade; AF: atrial fibrillation; CHA2DS2-VASc score - Congestive heart failure/left ventricular dysfunction, Hypertension,
Age ≥75years [2 points], Diabetes, Stroke [2 points], Vascular disease, Age 65–74 years, and Sex category (female). Total scores range between 0-9; low risk CHA2DS2-VASc score: 0 male; 1 female,
intermediate: 1male, ≥2 female, high risk CHA2DS2-VASc score: ≥2 male; ≥3 female; TIA: transient ischemic attack; eGFR: estimated glomerular filtration rate, ml/min/1.73 m2; HAS-BLED score –
uncontrolled Hypertension: systolic ≥160 mmHg, Abnormal renal/liver function, Stroke, Bleeding history or predisposition, Labile INR ratio/TTR <60, Drugs/alcohol concomitantly. Total scores range
between 0-9; low risk of bleeding range between 0-2 and high risk of bleeding ≥3; modified European Heart Rhythm Association symptom scale (mEHRA): 1 no symptoms, 2 mild and moderate, 3
severe, 4 disabling; SAMe-TT2R2 score – Sex female, Age<60, Medical history (more than two comorbidities), Treatment (interacting drug, e.g. Amiodarone), Tobacco use (doubled) and Race (non-
white, doubled). Total scores ranged from 0-8; probable good response to VKA therapy range between 0-2 and probable poor response to VKA therapy ranged from ≥3; SD: standard deviation
* Vascular disease: prior myocardial infarction, peripheral artery disease or aortic plaque; # Lung disease: obstructive and restrictive diagnosed lung conditions; †eGFR <60ml/min or as noted in
medical notes
145
2.4.2 Psychological measures, knowledge and beliefs about medication of
AF patients overall and according to OAC groups (warfarin vs. NOACs)
2.4.2.1 Depression and anxiety
At baseline, the median (IQR) depression and anxiety scores in the whole cohort were 4.0 (1.0-
8.0) and 1.0 (0-5.0), respectively (Table 2.3). Most patients had none or minimal symptoms of
depression (57.6%) or anxiety (71.9%) (Table 2.3). The median (IQR) depression and anxiety
scores were higher in patients receiving a NOAC than those receiving warfarin, [depression [4.0
(1.0-8.0) vs. 2.0 (0-4.5)] and anxiety [1.0 (0-5.0) vs. 0 (0-4.5)], respectively.
At the 6-month follow-up, the overall median (IQR) scores of PHQ-9 and GAD-7 were the same
[depression 4.0 (0-9.0) and anxiety 1.0 (0-5.0), respectively] (Table 2.3). There were no significant
differences in the median (IQR) depression and anxiety scores (Table 2.4 and Figure 2.3) and
the proportion of AF patients in different categories of depression (Figure 2.4) and anxiety (Figure
2.5) across the two time points.
The prevalence of major depression (PHQ-9 ≥15) at baseline and follow-up was 6.7% (Table
2.4). Meanwhile, the prevalence of generalised anxiety disorder (GAD-7 ≥10) was 11.4% at
baseline but declined slightly to 9.5% at follow up (Table 2.4). There was no significant difference
between the proportion of patients with major depression and generalised anxiety disorder over
time (Table 2.4 and Figure 2.6).
146
Table 2.3: Baseline and 6 months follow up psychological measures of AF patients overall and according to OAC groups
(warfarin vs. NOACs)
Baseline (N=139) Follow up (N=105)

Median (IQR) Overall, Warfarin NOACs Overall, Warfarin NOACs
N=139 N=13 N=126 N=105 N=9 N=96
PHQ-9 (9 items; scores range from 0-27)
Total score 4.0 (1.0-8.0) 2.0 (0-4.5) 4.0 (1.0-8.0) 4.0 (0-9.0) 3.0 (0-6.0) 4.0 (0.3-9.0)
Minimal 0-4, N (%) 80 (57.6) 10 (76.9) 70 (55.6) 56 (54.4) 7 (77.8) 51 (53.1)
Mild 5-9, N (%) 31 (22.3) 2 (15.4) 29 (23.0) 25 (24.3) 1 (11.1) 24 (25.0)
Moderate 10-14, N (%) 20 (14.4) 0 20 (15.9) 15 (14.6) 0 15 (15.6)
Moderately severe 15- 5 (3.6) 0 5 (4.0) 6 (5.8) 1 (11.1) 5 (5.2)
19, N (%)
Severe depression 20- 3 (2.2) 1 (7.7) 2 (1.6) 1 (1.0) 0 1 (1.0)
27, N (%)
GAD-7 (7 items; scores range from 0-21)
Total score 1.0 (0-5.0) 0 (0-4.5) 1.0 (0-5.0) 1.0 (0-5.0) 0 (0-5.0) 1.0 (0-5.0)
Minimal 0-4, N (%) 100 (71.9) 10 (76.9) 90 (71.4) 68 (66.0) 6 (66.7) 64 (66.7)
Mild 5-9, N (%) 23 (16.5) 1 (7.7) 22 (17.5) 25 (24.3) 2 (22.2) 23 (24.0)
Moderate 10-14, N (%) 8 (5.8) 2 (15.4) 6 (4.8) 6 (5.8) 0 6 (6.3)
Severe anxiety 15-21, 8 (5.8) 0 8 (6.3) 1.0 (1.0) 1 (11.1) 3 (3.1)
N (%)
GAD-7: The Generalised Anxiety Disorder 7-item; IQR: interquartile range; NOACs: non-vitamin K oral anticoagulants; OAC: oral anticoagulants; PHQ-9: Patient
Health Questionnaire
147
Table 2.4: Changes in psychological measures between baseline and 6 months follow up among overall AF patients
(N=105)
Median (IQR) change in score Baseline Follow up Changes in score Differences over time
N=105 N=105 over time p- value*
PHQ-9 (9 items)
Median (IQR) score 4.0 (1.0-8.5) 4.0 (0-9.0) 0 (-2.5 to 1.5) 0.53
Major depression PHQ-9≥15, N (%) 7 (6.7) 7 (6.7) - 1.00
GAD-7 (7 items)
Median (IQR) score 1.0 (0-5.0) 1.0 (0-5.0) 0 (-1.0 to 1.0) 0.67
Major anxiety disorder, GAD-7≥10, N 12 (11.4) 10 (9.5) - 0.822
(%)
*Wilcoxon-signed ranked test; AF: atrial fibrillation; PHQ-9: Patient Health Questionnaire; GAD-7: The Generalised Anxiety Disorder 7-item; IQR: interquartile range
148
90
*p=0.02 83.3 83.3
79.2
80
75 75
72.2
70 66.7 66.7
58.3
60
Mean/median score
54.2
50
40
Baseline
30
Follow up
19.2 19.1
20
13 12.9
10.4 10.7
8.3 8.2
10 5.8 5.9 6.2 6.1
4 4
1 1
0
ity
n*
rn
re
rn
-9
-7
se
ty
s
ia
om
dg
si
ar
co
ce
ce
Q
AD
io
nt
tiv
ru
es
PH
le
ct
H
re
pt
on
ls
on
ve
ac
G
ow
fa
ec
ffe
-G
ba
O
tc
ily
tis
Sy
N
6
kn
di
Q
lo
-G
-S
en
da
=8
sa
-S
BM
G
rn
T
Q
AF
*N
Q
EQ
T
ce
BM
T
BM
at
EQ
EQ
BM
EQ
on
tre
AF
AF
AF
-c
AF
T
ty
EQ
si
es
AF
ec
N
Figure 2.3: Mean/median score at baseline and 6 months follow up for depression, anxiety, knowledge of AF, beliefs about
medication and quality of life among overall AF patients who completed the questionnaire at both time points (N=105)
149
90
78.1 79.0
80
70
60
% AF patients
50 Baseline
40 Follow up
30
20 15.2 14.3
10 4.8 5.7
1.9 1
0
Min/mild (0-9) Moderate (10-14) Moderately severe (15-19) Severe depression (20-27)
Figure 2.5: PHQ-9 scores in categories among overall AF patients at baseline and 6
months follow up (N=105)
100
90.6
88.6
90
80
70
60
% AF patients
50 Baseline
Follow up
40
30
20
10 6.7 5.7 4.8 3.8
0
Min/mild (0-9) Moderate (10-14) Severe anxiety (15-21)
Figure 2.4: GAD-7 scores in categories among overall AF patients at baseline and 6
150
12
11.4
Major depression
Generalised anxiety disorder
10 9.5
6.7 6.7
% AF patients
0
Baseline Follow up
Figure 2.6: Major depression (PHQ-9 ≥15) and generalised anxiety disorder (GAD-7≥10) among overall
AF patients at baseline and 6 months follow up (N=105)
151
2.4.3 AF knowledge
At baseline, the overall mean (SD) score, for AF knowledge was low at 5.7 (1.7) with warfarin
patients obtaining slightly higher mean scores overall compared to NOAC patients (Table 2.5).
At follow up, the overall mean (SD) score for AF knowledge was similar to the score at
baseline, at 5.9 (1.9) (Table 2.5). There were no significant differences in the overall mean AF
knowledge scores between baseline and 6-months (Table 2.6).
Each question of the AF knowledge scale was also analysed specifically among patients who
answered the questionnaires at both time points (N=105) (Table 2.7 and Figure 2.7). The
majority of patients correctly answered the ‘reason for OAC prescription’ (90.5%) and on the
question ‘regarding physical activity’ (91.4%) at baseline. However, significantly fewer patients
correctly answered the ‘reason for OAC prescription’ at follow up (74.3%) compared to
baseline (p=0.002). Conversely, significantly higher proportions of patients were aware of ‘the
consequences of AF’ at follow up (88.6%) compared to baseline (50.5%; p<0.001) (Table 2.7
and Figure 2.7).
152
Table 2.5: Baseline and 6 months follow up knowledge scale of AF patients overall and according to OAC groups (warfarin vs.
NOACs)
AF knowledge scale (11 items; scores range from 0-100%)

Baseline (N=139) Follow up (N=105)
Mean (SD), % Overall Warfarin NOACs Overall Warfarin NOACs
N=139 N=13 N=126 N=105 N=9 N=96
Total overall scores 5.7 (1.7) 6.8 (1.2) 5.6 (1.7) 5.9 (1.9) 7.0 (1.3) 5.8 (1.9)
(min-max: 0-11)
Total overall scores, % 52.0 (15.4) 61.5 (11.2) 51.0 (15.4) 53.9 (16.9) 63.6 (11.1) 52.9 (17.0)
AF: atrial fibrillation; NOACs: non-vitamin K oral anticoagulants; OAC: oral anticoagulants; SD: standard deviation
Table 2.6: Change in knowledge scale between baseline and 6 months follow up among overall AF patients (N=105)
AF knowledge scale Baseline Follow up Changes in score Differences over time p-

(11 items; scores range from 0- N=105 N=105 over time value*
100%)
Total overall score 5.8 (1.7) 5.9 (1.9) -0.1 (2.0) 0.50
(min-max: 0-11)
Total overall score, % 52.6 (15.5) 53.9 (16.9) -1.2 (18.3) 0.50
*Paired t-test; AF: atrial fibrillation
153
Table 2.7: Specific questions in the AF knowledge scale and percentages of patients with correct response at baseline and 6 months
follow up (N=105)
Questions in AF knowledge score Baseline Follow up p-value*

AF in general
1. If AF is identified without any complaints, patients should immediately visit hospital 18 (17.1) 17 (16.2) 1.00
2. It is risky if patients do not feel his/her AF 45 (42.9) 33 (31.4) 0.06
3. AF is a rare condition 8 (7.6) 8 (7.6) 1.00
Symptoms recognition
1. What are the trigger factors for AF 47 (44.8) 58 (55.2) 0.07
2. Why is it important to take my medications properly 50 (47.6) 44 (41.9) 0.43
3. What is atrial fibrillation? 64 (61.0) 69 (65.7) 0.53
AF treatment
1. Why patients using oral anticoagulation should be careful with the use of alcohol 71 (67.6) 72 (68.6) 1.00
2. What is the function of anticoagulation clinic 61 (58.1) 57 (54.3) 0.61
3. Why is oral anticoagulation prescribed in certain patients with AF 95 (90.5) 78 (74.3) 0.002
AF general attitude
1. Statements regarding physical exercise 96 (91.4) 93 (88.6) 0.61
2. Statements regarding the danger associated with AF 53 (50.5) 93 (88.6) <0.001
*chi-square; AF: atrial fibrillation
154
Baseline
Follow up
100 p=0.002
p<0.001
90.5 91.4
88.6 88.6
90
% of patients with correct answer
80
74.3
70 67.6 68.6
65.7
61
60 58.1
55.2 54.3
50.5
50 47.6
44.8
42.9 41.9
40
31.4
30
20 17.1 16.2
10 7.6 7.6
0
Patient should attend It is risky if patients do AF is a rare condition Trigger factors for AF The importance of AF definition Reason for being Function of Reason for OAC Statements on Statements on danger
hospital if there is no not feel his/her AF taking medication careful with alcohol anticoagulation clinic prescription physical exercise associated with AF
AF symptoms properly while taking OAC
Figure 2.7: Proportion of AF patients with correct answers in each specific question at baseline and 6 months follow up (N=105)
155
2.4.4 Beliefs about medication
At baseline, the overall mean (SD) score in the general overuse and general harm domains were
10.5 (2.9) and 8.6 (2.9), respectively. In terms of patients’ specific beliefs about their
anticoagulation therapy, the mean (SD) specific-necessity score was higher [19.0 (3.0)] than the
specific-concern score [13.3 (3.5)], with a positive necessity-concern differential [5.8 (4.1)]. This
indicates that patients perceived their medications are being more important than their concerns
regarding medications. NOAC patients obtained slightly higher scores on the general harm and
specific-concerns domains at baseline compared to warfarin patients (Table 2.8).
At six months follow up, the scores in the general and specific domains remained the same.
Similar scores were seen among warfarin and NOAC patients in the general overuse and general
harm domains, while NOAC patients achieved slightly higher scores on the specific necessity and
necessity-concern differential (Table 2.8). There was no significant difference in either the
general and specific domains over time (Table 2.9).
156
Table 2.8: Baseline and 6 months follow up scores on beliefs about medication of AF patients overall and according to
OAC groups (warfarin vs. NOACs)
Baseline scores (N=139) Follow up scores (N=105)

Baseline scores, Overall, Warfarin NOACs Overall, Warfarin NOACs
mean (SD) N=139 N=13 N=126 N=105 N=9 N=96
Beliefs about medication (BMQ; 18 items)
BMQ general (scores range from 4-20)
General overuse 10.5 (2.9) 10.5 (2.9) 10.5 (2.9) 10.7 (2.9) 10.8 (1.4) 10.7 (3.0)
(4-20)
General harm 8.6 (2.9) 7.4 (1.7) 8.7 (2.9) 8.2 (2.4) 8.4 (1.2) 8.2 (2.5)
(4-20)
BMQ specific (scores range from 5-25)
Specific necessity 19.0 (3.0) 19.2 (2.4) 19.0 (3.1) 19.1 (3.1) 18.8 (1.4) 19.1 (3.2)
(5-25)
Specific concern 13.3 (3.5) 12.2 (2.7) 13.4 (3.6) 12.9 (3.8) 13.3 (4.3) 12.9 (3.8)
(5-25)
Necessity-concern 5.8 (4.1) 6.9 (3.9) 5.7 (4.2) 6.1 (4.4) 5.4 (4.2) 6.2 (4.4)
differential
BMQ: Beliefs about medication; NOACs: non-vitamin K oral anticoagulants; OAC: oral anticoagulants; SD: Standard deviation
157
Table 2.9: Change in score on beliefs about medication between baseline and 6 months follow up among overall AF
patients (N=105)
BMQ (18 items) Baseline Follow up Changes in score Differences over

mean (SD) score N=105 N=105 over time time p- value*
‡
BMQ general
General overuse 10.4 (3.0) 10.7 (2.9) -0.4 (2.8) 0.20
General harm 8.3 (2.8) 8.2 (2.4) 0.1 (2.4) 0.78
BMQ specific‡
Specific necessity 19.2 (3.0) 19.1 (3.1) 0.1 (2.4) 0.54
Specific concern 13.0 (3.5) 12.9 (3.8) 0.0 (3.5) 0.91
Necessity-concern differential 6.2 (4.0) 6.1 (4.4) 0.1 (3.9) 0.78
*paired t-test; AF: atrial fibrillation; BMQ: Beliefs about medication; SD: standard deviation
158
2.4.5 Quality-of-life
The overall global median (IQR) baseline score for quality of life was 66.7 (53.7-77.8). Patients
scored lowest in the daily activity domain [60.4 (39.6-79.2)] and highest in the AF symptoms
domain [79.2 (58.3-95.8)]. NOAC patients had higher scores in the AF symptoms and treatment
satisfaction domains and the overall global score compared to warfarin patients (Table 2.10).
At follow up, the median (IQR) global score was also 66.7 (49.1-81.9) and patients continued to
score lowest in the daily activity domain (Table 2.10). Similarly, NOAC patients had higher scores
in the AF symptoms and treatment satisfaction domains compared to warfarin patients. Among
patients who completed the questionnaires at both time-points, there was a significant increase
in AF symptoms score at follow-up [83.3 (64.6-100) vs. 79.2 (54.2-95.8); p=0.02], with no
significant change in the other domains over time (Table 2.11 and Figure 2.8).
159
Table 2.10: Baseline and 6 months follow up quality of life scores of AF patients overall and according to OAC groups
(warfarin vs. NOACs)
AFEQT (20 items; scores range from 0-100)

Baseline measures (N=139) Follow up measures (N=105)
Median (IQR) Overall, Warfarin NOACs Overall, Warfarin NOACs
N=139 N=13 N=126 N=105 N=9 N=96
AF symptoms (0-100) 79.2 (58.3-95.8) 58.3 (41.7-97.9) 79.2 (58.3-95.8) 83.3 (64.6-100) 75.0 (54.2-97.9) 83.3 (66.7-100)
Daily activity (0-100) 60.4 (39.6-79.2) 68.8 (31.3-82.3) 60.4 (39.6-77.6) 54.2 (34.4-77.1) 66.7 (21.9-82.3) 54.2 (35.4-76.6)
Treatment concern 75.0 (52.8-86.1) 75.0 (68.1-83.3) 75.0 (52.8-88.9) 72.2 (58.3-88.9) 72.2 (52.8-88.9) 72.2 (58.3-88.9)
(0-100)
Treatment Satisfaction† 75.0 (66.7-83.3) 66.7 (56.3-85.4) 75.0 (66.7-83.3) 83.3 (66.7-91.7) 75.0 (66.7-91.7) 83.3 (66.7-89.6)
(0-100)
Overall global score 66.7 (53.7-77.8) 60.2 (50.9-80.1) 66.7 (53.7-77.8) 66.7 (49.1-81.9) 75.0 (47.2-81.5) 66.2 (48.6-82.2)
(0-100)
† N=111; AFEQT: Atrial Fibrillation Effect on Quality–of-life; AF: atrial fibrillation; OAC: oral anticoagulants; IQR: interquartile range; NOACs: non-vitamin K oral
anticoagulants
160
Table 2.11: Change in quality of life scores between baseline and 6 months follow up among overall AF patients (N=105)
Median (IQR) change in Baseline Follow up Changes in score p- value*

score N=105 N=105 over time
AFEQT (20 items)
Median (IQR) AF 79.2 (54.2-95.8) 83.3 (64.6-100) -4.8 (22.2) 0.02
symptoms
Median (IQR) daily activity 58.3 (38.5-80.2) 54.2 (34.4-77.1) 2.8 (18.9) 0.33
Median (IQR) treatment 75.0 (52.8-86.1) 72.2 (58.3-88.9) -0.90 (23.2) 0.57
concern
Median (IQR) treatment 75.0 (66.7-83.3) 83.3 (66.7-83.3) -4.3 (22.2) 0.07
†
satisfaction
Median (IQR) overall 66.7 (51.9-78.0) 66.7 (49.1-81.9) -0.3 (16.2) 0.65
global score
†
N=86 at both time points; *Wilcoxon-signed ranked test; AFEQT: Atrial Fibrillation Effect on Quality–of-life; AF: atrial fibrillation; IQR: interquartile range
161
p=0.02
90
83.3 83.3
79.2
80
75.0 75.0
72.2
70 66.7 66.7
60 58.3
54.2
50
Baseline
40 Follow up
30
20
10
0
Symptoms Daily activity Treatment concern Treatment Satisfaction Overall global score
Figure 2.8: Quality of life domain scores assessed by the AFEQT questionnaire in AF patients overall at baseline and 6
162
2.4.6 Time in therapeutic range, medication adherence and adverse
clinical outcome at 6 months follow up
The overall median (IQR) TTR for warfarin patients at 6 months using the Rosendaal and
PINRR methods were 62.6 (49.0-70.3) and 43.1 (33.0-50.0), respectively. When the inception
period (the first 4 weeks of warfarin treatment) was excluded, TTR and PINRR were
significantly better, 77.3 (54.4-84.7; p=0.004) and 56.4 (45.8-66.7; p=0.004), respectively.
Among the 126 patients on NOACs at baseline, 96 patients reached the six months follow up.
Among them, 67 patients (70%) returned their pill boxes and blister packages at follow up and
the pill count demonstrated 100% adherence. Six patients had their medications prepared
from the pharmacy as weekly blister packs thus they are not able to keep their empty NOAC
packages/boxes. Moreover, upon questioning, majority of patients (86.5%) claimed that they
never forget to take their NOAC medication and only one patient claimed they have forgotten
to take it over the past two weeks.
At follow up, there were only 13 patients with ≥1 of the composites of bleeding, CV
hospitalisations and death. No thromboembolic events occurred, three patients experienced
bleeding events (2 major bleed and 1 CRNMB bleed), 8 patients had CV hospitalisation and
3 patients died. When the events were stratified according to the SAMe-TT2R2 score
categories (0-2 vs. >2), a non-significant higher proportion of patients with SAMe-TT2R2 >2
experienced CV hospitalisation (7.1% vs. 5.4%), death (3.6% vs. 1.8%) and ≥1 of the
composites of bleeding, CV hospitalisations and death (10.7% vs. 9.0%) compared to those
with SAMe-TT2R2 0-2 respectively. However, all bleeding events (2.7%) occurred among those
with SAMe-TT2R2 0-2.
163
2.5 Discussion
AF patients who were newly anticoagulated with either warfarin or NOACs for stroke
prevention do not appear to have significant depressive or anxiety symptoms at baseline.
However, patients had poor knowledge of AF and its treatment, perceived that OAC
medication is important and this outweighed their medication concerns, and they also had
poor quality of life. At six months, there were no significant changes in depression, anxiety,
beliefs about medication and quality of life. However, more patients were aware of the
consequences of AF but fewer patients understood the reason of OAC prescription and
patients had improvements in AF symptoms at follow up compared to baseline.
In this study, few patients reported depressive and anxiety symptoms at baseline and follow
up; consistent with previous studies in the West Midlands assessing depression and anxiety
among AF patients using validated questionnaires (99, 407) (219). In the overall population,
more patients experienced symptoms of generalised anxiety rather than major depression at
baseline (11.4% vs. 6.7%, respectively) and follow up (9.5% vs. 6.7%, respectively). This
finding is consistent with studies by Lane et al (99) and Clarkesmith et al (219). Both studies
(99, 219) reported a higher prevalence of anxiety (38.5-41.5%) than depression (25.5%)
among their AF cohorts at baseline, but were limited by smaller sample sizes (N=70 to 97)
and used different questionnaires to assess symptoms of depression and anxiety (BDI, STAI
and HADS) than the present study. Taken together, these results suggest that anxiety is the
predominant affective trait among AF patients, which might influence the patient’s quality of
life (96, 99, 219). The presence of higher anxiety symptoms rather than depressive symptoms
in the current cohort is difficult to explain but may be influenced by the presence of
comorbidities as well as AF symptoms and commencing OAC therapy. A recent systematic
review (408) of eight studies (AF vs. control group) assessed the role of psychological factors
in AF using validated questionnaires. From the five studies (95, 96, 409-411) assessing the
164
role of depression in AF patients included in the review, only one study (409) showed a
significant difference in the depression level among AF patients compared to healthy controls
(effect size 3.08; 95% CI 2.63-3.57) (409). For trait anxiety, two studies (96, 410) showed
higher levels of trait anxiety in AF patients compared to hypertensive patients (96) and patients
with supraventricular tachycardia (410), although the differences were small for both studies
[0.34; 95% CI 0.07-0.61 and 0.41 (0.02-0.80) respectively]. As a result, no clear conclusion
was made due to small number of studies and methodological differences (408). For example,
types of AF patients studied were different (paroxysmal vs. persistent vs. permanent), different
type of questionnaires were used to assess depression and anxiety and conclusions from
each study were inconsistent. Hence, future studies in this area are needed to draw a clear
conclusion on the involvement of psychological distress in AF patients.
In general, patients in the current study have positive perception towards OAC medication
evidenced by high mean (SD) specific-necessity scores [19.0 (3.0)] and a positive necessity-
concern differential [5.8 (4.1)] at baseline. This remained unchanged at follow up indicating
that patients perceive their OAC medication as more important than their concerns about it.
This finding is similar to the TREAT study (219) in the UK and in another study in Palestine
(412). Both studies showed higher specific-necessity beliefs than specific-concern beliefs and
this was associated with better treatment adherence in both studies (219, 412) and better TTR
in the TREAT education intervention group compared to the usual care group (219). In this
cohort, patients also disagreed that their medications are harmful. This further strengthens the
positive beliefs that patients have towards the importance of taking their medication which
could potentially impact their adherence level. In the adherence assessment among patients
receiving NOACs via the pill count method at six months, 70% of the patients reported 100%
adherence. However, the remainder (N=29) did not return any boxes/blister packages as they
had forgotten to keep them for the study purposes. In contrast, one study from Saudi Arabia
(413) among patients with multiple chronic diseases (diabetes, hypertension, asthma; N=408)
showed higher general harm score [13.6 (2.3)] than the current study and this was significantly
165
more prevalent among non-adherent patients compared to high adherent patients. This
suggests that where patients have negative thoughts about their medication, they are more
likely to become non-adherent. In this study, no significant differences were seen in both the
general and specific domains of the BMQ over time. Although the available evidence
demonstrates consistent results with the current studies, caution must be exercised as there
could be differences in terms of medication beliefs, usage and cultural differences between
studies conducted in Western countries compared to Middle Eastern countries (412) (413).
In terms of AF knowledge, AF patients in this cohort had poorer knowledge about AF and its
treatment reflected by an average score of 5.7 (1.7) and 5.9 (1.9) at baseline and follow up,
respectively. There were no significant differences in the AF knowledge score over time.
Nevertheless, on assessing each specific question on the AF knowledge questionnaire, more
patients were aware of the consequences of AF at follow-up but fewer understood the reasons
of OAC prescription at follow up compared to baseline. Clearly, there is a gap in AF knowledge
among patients which changes over time in terms of the need of OAC therapy for stroke
prevention. These findings are in keeping with other studies investigating knowledge among
AF patients (283, 284, 286, 287). Studies showed that AF patients have inadequate
knowledge of their condition (284-287), poor understanding of the benefits and risk of their
treatment, (285-287) specifically anticoagulation therapy and not aware of the factors that
could impact the effectiveness and safety of treatment (284-286). Thus, this could influence
their ability to make informed choices of the treatment options and prevent them from being
actively involved in management of their own treatment. Lane and colleagues (283) have
included strategies to improve knowledge among AF patients entailing: greater awareness
among the public of what is AF and the repercussions, better patient support and provisions
of educational materials, enhanced understanding of the patients’ needs among medical
professions, improved communication between physicians and patients and including
patients’ preferences during the discussion of treatment options (283). Studies (219, 386)
have also shown that with better education, emotional distress can be reduced, adherence
166
and concordance can be enhanced (219, 284, 289) and quality of life can be improved
resulting in better treatment outcomes (219, 386).
Overall, AF patients had poor quality of life evident by a median score of 66.7% for the overall
global score and this remained the same at 6 months follow up; similar to other studies (99,
219). In contrast, one trial (386) in the Netherlands (N=712) showed higher baseline QoL score
in AF patients randomised to the nurse-led group (72%) (with psychosocial support and
education intervention at 3,6 and 12 months) but similar QoL scores in the usual care group
(68%) compared to the current study. In their study, QoL, including anxiety and depression,
improved significantly after 1 year of follow up, irrespective of treatment group. However,
patients in the nurse-led group had better knowledge at follow up and quality of life was
significantly correlated with knowledge. Taken together, these findings suggest the benefits of
a structured educational intervention (involving educational reinforcement and psychosocial
support) in promoting better knowledge, emotional health and quality of life. In the current
cohort, patients seem to have the greatest limitation in their physical health compared to
emotional and clinical health at both baseline and follow up (evident by lowest score in the
daily activity domain). Multiple observational studies worldwide (109, 111-113, 116) have
shown similar results where AF patients appear to have the greatest impairment in their
physical health compared to their emotional health. This could be influenced by ageing,
symptoms severity, number of comorbidities, all of which could affect the patients’ ability to
conduct physical activity and thereby reduce their quality of life (QoL) (386) (109, 111-113,
116). In this study, significant improvement was only seen in the AF symptoms domain at
follow up, suggesting that patients were less affected by their symptoms at this point. This
could probably be influenced by effective symptom management with either pharmacological
or non-pharmacological interventions (cardioversion or ablation) although this is purely
speculative. Another explanation is that maybe patients had had time to come to terms with
their diagnosis and were therefore less bothered by their symptoms.
167
There were only thirteen patients on warfarin therapy and their overall median (IQR) TTR at
six months was suboptimal at 62.6% (49.0-70.3) however, TTR was significantly better and at
the optimal level after excluding the inception period (the first four weeks of therapy to allow
attainment of warfarin dose) with 77.3% (54.4-84.7; p=0.004); a finding similar to previous
study(219). In terms of adverse events, there were no TE events, however, three patients had
bleeding events (2 major bleed and 1 CRNMB), eight patients were hospitalised for CV reason
and two patients died at follow up. All adverse events occurred among NOAC patients.
2.5.1 Limitations
This study has several limitations. Firstly, the study contained only 13 patients on warfarin and
no patients attended the TREAT educational intervention thus the original planned analyses
and hypothesis testing were not possible, due primarily to a change in clinical practice
regarding NOAC prescription over warfarin for AF patients newly initiating OAC. The
recruitment target was not achieved (N=256) based on the pre-specified SAMe-TT2R2 groups
(N=64 in each group). Thus, the results for patients on warfarin or NOAC were presented
descriptively and no significance testing was undertaken. Secondly, although two centres
were utilised for patient recruitment, the overall number of patients included was low compared
to other longitudinal studies assessing psychological measures, knowledge and quality of life
among AF patients (108, 386, 414). However, the results from the current study are in keeping
with previous studies within the West Midlands reporting low levels of anxiety, depression and
knowledge among AF patients (99, 407). In addition, there were only 16 non-white patients,
thus the results may not be applicable to all AF participants. Thirdly, only 105 (76%) patients
completed the 6-month questionnaires; the study had 24% attrition rate at 6 months. It is
possible that those who did not return the 6-month questionnaires may have experienced
more health problems and/or worsening emotional health and quality of life. The results may
have been different if all respondents completed the 6-month questionnaires. However, the 6-
month response rate was 76%, which is higher than many previous questionnaire studies (96,
219, 415, 416). Fourth, among the 96 NOAC patients who reached the follow up only 70%
168
returned their empty pill boxes/packages. The pill count method is not the most robust method
of assessing adherence as patients may not have taken all the pills but simply returned the
empty boxes and blister packs, however this method is simple, low cost and used in many
clinical trials (389). To date, direct measures of adherence including measurement of drugs in
plasma or urine were considered the most accurate way to assess adherence however it is
expensive, difficult to perform, is dependent on the test used and drug metabolism (389).
2.5.2 Clinical implications and future research
This study will add to the existing literature on the psychological health, quality of life and AF
knowledge of patients with AF who receive OAC. Whilst NOACs have sought to overcome the
inherent difficulties experienced by patients prescribed with warfarin, the psychological impact
on patients (as well as warfarin patients) has not been thoroughly investigated to date.
Together, these results will provide important insights into patient’s feelings, beliefs and
awareness about the OAC, and the impact on their QoL. This information can be used to
create innovative strategies to improve health outcomes in AF patients receiving OAC therapy
for stroke prevention.
More in-depth studies investigating the psychological impact of OAC therapy in AF patients
are needed. Validated questionnaires to collect data on the emotional impact (anxiety and
depression) and quality of life could be routinely included as endpoints in large
RCTs/multinational registries so that this information is available alongside other endpoints.
Furthermore, the TREAT-2 intervention could be extended to other developing countries like
Malaysia where warfarin is the main OAC of choice for stroke prevention in AF. In Malaysia,
the majority of VKA patients are being managed by clinical pharmacist in the hospital setting
in the Medication Therapy Adherence Clinic (MTAC) (417). This clinic aims to optimise
medication therapy, improve medication adherence and prevent/reduce problems related to
medication with a pre-specified protocol that incorporates a patient education checklist, INR
169
testing interval, dosage adjustment and warfarin dispensing (417). To date, no trials are
available in Malaysia to determine the impact of MTAC towards TTR however, one
retrospective cohort study (418) of mainly AF patients showed that TTR was significantly
better in the MTAC compared to usual care group (TTR 65.1 vs. 48.3; p<0.05 respectively)
while another study (419) of AF patients showed no difference in TTR between the two
respective groups. It would be of interest to incorporate the components of TREAT-2
intervention into the MTAC protocol and design a trial to examine TTR in the TREAT-2
intervention group + MTAC (N=50) vs. usual care group (N=50) in Malaysia. The outcome of
this study could determine the benefits of the added TREAT-2 intervention into the current
MTAC protocol and the impact of such interventions towards psychological health, quality of
life and knowledge among patients receiving it.
2.6 Conclusion
Newly anticoagulated AF patients appear to have low levels of anxiety, depression, poor AF
knowledge, positive perceptions about their medication and poor overall quality of life at
baseline which remains unchanged at six months follow up. However, more patients were
aware of the consequences of AF and AF symptoms (by the AFEQT questionnaire) improved
significantly at follow up.
170
Chapter 3. Anticoagulation control in different ethnic groups
receiving vitamin K antagonist therapy for stroke prevention in
atrial fibrillation: the West Birmingham AF Project
3.1 Abstract
Introduction: Efficacy and safety of VKAs is optimised in AF patients when the INR is 2.0-
3.0. Anticoagulation control comparing different ethnic groups has not been well-assessed,
although epidemiological studies suggest poorer INR control in non-white cohorts.
Objective: To examine the quality of VKA control (TTR), predictors of anticoagulation control
and the prevalence of adverse clinical outcomes [thromboembolic (stoke/TIA and systemic
embolism), bleeding events, cardiovascular hospitalisation and all-cause mortality] in AF
patients in a multi-ethnic cohort at one acute Trust in the West Midlands, United Kingdom.
Ancillary analysis was also undertaken to investigate TTR among elderly (≥80 vs. <80 years)
and patients with different categories of kidney disease (eGFR≥90 vs. 60-89 vs.
≤59ml/min/1.73m2) within this cohort. Exploratory analyses investigated the relationship
between INR control and adverse clinical outcome.
Methods: All demographic and clinical data were collected retrospectively from the electronic
medical record database. VKA control was assessed retrospectively by TTR using the
Rosendaal method and percentage INRs in range (PINRR), among 991 White, Afro-
Caribbean and South-Asian AF patients. Predictors included patient’s demographics,
comorbidities and other clinical data and these were examined by multiple regression analysis.
The relationship between INR control and adverse clinical outcome was investigated with chi-
square.
Results: The overall mean (SD) age at warfarin initiation was 71.6 (9.4) years; 55% male;
mean (SD) CHA2DS2-VASc score 3.4 (1.6) and patients were followed up over a median of
171
5.2 years. The cohort consisted of 807 Whites, 102 South-Asians and 82 Afro-Caribbean
patients. Compared to Whites, mean (SD) TTR and PINRR were significantly lower in South-
Asians [TTR 67.9% vs. 60.5%, p<0.001; PINRR 58.8% vs. 51.6%, p<0.001 respectively] and
Afro-Caribbeans [TTR 67.9% vs. 61.1%, p<0.001; PINRR 58.8% vs. 53.1%, p<0.001
respectively], despite similar INR monitoring intensity. Whites had better anticoagulation
control, evidenced by a greater proportion with TTR ≥70% and PINRR ≥70%. Anticoagulation
control was significantly more likely to be sub-therapeutic (INR<2.0) among South-Asians and
Afro-Caribbeans compared to Whites (30.2%, 30.2%, and 24.7%, p<0.05 respectively).
Ancillary analyses showed that TTR was similar (66.6%) in patients ≥80 vs. <80 years and
there were no significant differences in TTR observed among patients with normal kidney
function (eGFR≥90), mild (eGFR 60-89) and mild-moderate-severe kidney disease (eGFR
≤59) [64.0% vs. 66.9% vs. 67.0%; p=0.053 respectively].
Logistic regression revealed that non-white ethnicity [OR 2.62 95%CI (1.67-4.10); p<0.001
and OR 3.47 (1.44-8.34); p=0.005] and anaemia [OR 1.65 95%CI (1.00-2.70); p=0.05 and OR
6.27 95%CI (1.89-20.94); p<0.003] were independent predictors of both TTR<70% and
PINRR<70%, respectively.
At 5.2 years, 329 (33.2%) patients experienced ≥1 major adverse clinical outcome (MACE).
Cardiovascular hospitalisations were significantly higher among South-Asians compared to
Whites (32.3% vs. 21.3%; p<0.05). Patients with CV hospitalisations were significantly more
likely to have poor TTR (TTR<70%, p=0.002 and TTR<65%, p=0.008).
Conclusions: Ethnic disparities in the quality of anticoagulation control are evident with
South-Asians and Afro-Caribbeans having poorer INR control compared to Whites. There
were no significant differences in TTR between elderly and younger patients (≥80 vs. <80
years) or between different categories of kidney disease. Non-white ethnicity and anaemia
remained the strongest independent predictor of poor TTR and PINRR. CV hospitalisations
were more prevalent among South-Asians and were associated with poor TTR.
172
3.2 Introduction
Vitamin K antagonists (e.g., warfarin) have been used for many decades for the prevention of
stroke in patients with AF (131). The efficacy and safety of VKA is determined by achieving
the target international normalised ratio (INR) of 2.0-3.0 in AF patients (131). To summarise
INR control over time, TTR can be calculated by various methods including Rosendaal’s (using
linear interpolation to assign INR value to each day between two consecutive INR values) and
the percentage of INRs within therapeutic range (PINRR) (420, 421). TTR is a significant and
important predictor of thromboembolic and bleeding outcomes in AF patients on VKA (3, 131,
173, 316). A recent European consensus document recommended an individual TTR of ≥70%
for optimal efficacy and safety outcomes whilst on a VKA (3), whilst the NICE guidelines on
AF recommend a TTR ≥65% (344).
Many factors can influence the quality of anticoagulation control thus affecting TTR, (see
Section 1.5.1, page 78 for a review of predictors of anticoagulation control) (199). Ethnicity
has been identified as one of the determinants of anticoagulation control in patients with AF
(199) and this has been incorporated in the SAMe-TT2R2 score (see Section 1.5.1.1.8, pages
83-85 for more information on ethnicity and TTR). Apart from ethnicity, studies have also
shown that increasing age (247, 252) and chronic kidney disease (263-270) were associated
with poor anticoagulation control [see Section 1.5.1 for more information on the impact of age
(pages 80-81) and kidney disease (pages 81-82 ) on TTR].
173
3.2.1 Study objectives
Birmingham is a multi-ethnic city mainly comprised of White British (53.1%), Afro-Caribbean
(4.4%), Pakistani (13.5%) and Indian (6%) citizens, and Sandwell and West Birmingham
Hospitals NHS Trust serves this tri-ethnic population. The main aim of this study was first to
investigate the quality of anticoagulant control; second, to examine predictors of
anticoagulation control (TTR) and lastly, to examine the prevalence of adverse clinical
outcomes in AF patients in a multi-ethnic cohort. Ancillary analyses were also undertaken to
investigate anticoagulation control among elderly patients (≥80 vs. <80 years) and patients
with different categories of kidney disease (eGFR≥90 vs. 60-89 vs. ≤59ml/min/1.73m2).
Exploratory analyses also investigated the relationship between INR control and adverse
clinical outcomes.
3.3 Methods
3.3.1 Study design
This is a retrospective cohort analysis of a multi-ethnic cohort of AF patients receiving VKA
therapy for stroke prevention at one acute Trust in the West Midlands, United Kingdom. Data
collection was undertaken from February to December 2016. AF patients receiving VKA
therapy for stroke prevention with a target INR range of 2.0-3.0 were selected from the DAWN
AC® anticoagulation management software; used by the Trust Anticoagulation Service to
manage anticoagulation therapy.
The DAWN AC® anticoagulation management software, a computer assisted dosage
program, is clinically validated software designed to manage large anticoagulation clinics in
an effective and safe way (422). It contains a complete anticoagulation decision support
package that includes induction, maintenance and bridging of warfarin therapy (422). It also
contains demographic and clinical information including reasons for anticoagulation, types of
174
anticoagulant used, target INR range, and the start date of anticoagulation, plus the dosing
algorithm, test interval and INR results including history and current INRs for VKA
management.
This study was considered as service evaluation by the SWBH Research and Development
department and therefore did not require REC approval. However, local R&D approval was
obtained (see email confirmation from SWBH R&D Department, Appendix 5).
3.3.2 Patient selection
An alphabetical list of patients was generated (N=2478) from DAWN® AC and patients were
selected at random by choosing every third, fifth and tenth patient from the patients list. A total
of 1070 patients were included constituting 43% of the whole population. The ethnic group
distribution of the population in this Trust (AF patients on VKA therapy) generated from the
DAWN database is as follows: 78% Whites, 7.4% South-Asian, 5.3% Afro-Caribbeans and
9.5% unknown/other ethnic background. For the purpose of this analysis, the present cohort
is representative of the whole AF and VKA cohort at the Trust, i.e., 81.4% White, 10.3% South-
Asian and 8.3% Afro-Caribbean.
Patients with EHRA type 1 VHD (N=45), unknown ethnicity (N=21), or unknown medical history
(N=13) were excluded from these analyses. EHRA type 1 VHD were defined as patients with
a diagnosis of moderate-to-severe mitral stenosis, rheumatic valvular disease or valve
replacement requiring VKA therapy. Thus, the final cohort comprises 991 patients (807 White,
102 South-Asian, and 82 Afro-Caribbean). Figure 3.1 presents the study design flow chart.
175
Atrial fibrillation patients on VKA therapy
selected from SWBH DAWN AC® Database
(N=2478)
Random patient’s selection according to

alphabetical order (N=1070)
Patient’s baseline demographic characteristics, past,

current medical and medication history recorded from
CDA and DAWN AC® databases
Patients excluded, N=79: EHRA type 1 VHD

(N=45), unknown ethnicity (N=21), unknown
medical background (N=13)
Final cohort, N=991
Elderly: * Ethnic groups: Chronic kidney disease: †

≥80 years vs. <80 years White, South-Asian, Afro- Group 1: eGFR ≥90
Caribbean population Group 2: eGFR 60-89
Group 3: eGFR ≤59
Outcome
Measure of Predictors of
Adverse clinical
anticoagulation anticoagulation
outcomes
control control
Time to therapeutic Percentage of INR Thromboembolic and bleeding

range (TTR) by within range (PINRR) events, CV hospitalization,
Rosendaal death
*Ancillary analysis 1; †Ancillary analysis 2; CDA: clinical data archive; CV: cardiovascular; INR: international
normalised ratio; TTR: time in therapeutic range; PINRR: percentage of INRs in range; SWBH: Sandwell and West
Birmingham Hospitals; VKA: vitamin K antagonist

176
3.3.3 Procedure
A proforma (see Appendix 4, Table A4.1) was used to collect all baseline demographic and
clinical characteristics of patients including medical history, medication, laboratory results and
also information on outcomes of interest, INR results and adverse clinical outcomes. All
demographics and clinical information were gathered from the Clinical Data Archive (CDA), an
electronic medical record database. The CHA2DS2-VASc score (146), HAS-BLED score (180)
and SAMe-TT2R2 score (199) were calculated for each patient based on the available
information and used to predict stroke, bleeding and anticoagulant control, respectively (see
Table 3.1).
Table 3.1: CHA2DS2-VASc, HAS-BLED, and SAMe-TT2R2 scores
CHA2DS2-VASc HAS-BLED
CHF or LVF ≤40% 1 Uncontrolled Hypertension† 1
Hypertension 1 Abnormal kidney/liver function 1/2
Age ≥75 years 2 Stroke 1
Diabetes 1 Bleeding‡ 1
Stroke/TIA/ thromboembolism 2 Labile INR§ 1
Vascular Disease* 1 Elderly ≥65 years 1
Age 65-74 1 Drugs/alcohol excess|| 1/2
Female sex 1
Total 9 9
SAMe-TT2R2
Sex (female) 1
Age (<60 years) 1
Medical history¶ 1
Treatment strategy# 1
Tobacco use** 2
Race (non-Caucasian) 2
Total 8
CHF: congestive heart failure; INR: international normalised ration; LVF: left ventricular function; TIA: transient ischemic attack; *
prior myocardial infarction, peripheral artery disease or aortic plaque; †systolic blood pressure ≥160mmHg; ‡ bleeding history,
anaemia or predisposition for bleeding; §poor time in therapeutic range (<60%); ||concurrent antiplatelet /non-steroidal anti-
inflammatory drugs, ≥8units alcohol/week; ¶Two or more of the following: hypertension, diabetes mellitus, coronary artery
disease/myocardial infarction, peripheral arterial disease, congestive heart failure, previous stroke, pulmonary disease, and
hepatic or renal disease; #interacting drugs, e.g., amiodarone; **within 2 years
177
3.3.4 Variables and definitions
3.3.4.1 Primary dependent/outcome variables:
3.3.4.1.1 Time in therapeutic range (TTR)
All available INR values from inception to 31st December 2016 or cessation of OAC therapy or
death were obtained from the CDA and DAWN AC® databases. The quality of anticoagulation
control was calculated as the proportion of TTR (INR 2.0-3.0), using the Rosendaal method
(421) (which uses linear interpolation to assign an INR value to each day between two
consecutive INR values) and PINRR, calculated by dividing the number of INR in range with
the total INR values (420, 423). TTR values were further dichotomized into TTR ≥70% and
TTR <70%, according to a recent European consensus document for optimal efficacy and
safety outcomes (3) whilst on a VKA, and ≥65% and <65% based on the NICE guidelines
(118). The proportion of sub-therapeutic (INR <2.0) and supra-therapeutic (INR >3.0) INRs
was also calculated. Years of follow up is defined as the duration of warfarin therapy denoted
from the first available INR on the DAWN AC® system until the present.
3.3.4.1.2 Adverse clinical outcome
The adverse clinical outcomes of interest were stroke, transient ischemic attack (TIA),
systemic embolism (SE), bleeding events [combination of major and non-clinically relevant
non-major bleeding (CRNMB)], cardiovascular (CV) hospitalization and all cause death and
these were obtained from the patients’ medical records, CDA. All analysis pertaining to
adverse clinical outcome were exploratory in nature. A composite endpoint of major adverse
clinical event (MACE) encompassed ≥1 of the following: stroke/TIA, systemic embolism,
bleeding, cardiovascular hospitalisation or death. Definitions of each outcome can be found in
Section 2.3.3, pages 134-135.
178
3.3.4.2 Independent variable: Ethnicity
Self-reported ethnicity was identified directly from the electronic medical records, CDA/DAWN®
AC, where available. The different ethnic groups identified were White British, White Irish,
White others, Asian British Indian, Asian British Pakistani, Asian British Bangladeshi, other
Asians, Black British Caribbean and Black British African, according to the UK Census. These
were then recoded into the three main ethnic groups; White, South Asian and Afro-Caribbean
respectively. Those without information on ethnicity were excluded from the study (N=21).
Ethnicity was an independent variable for the first part of the analysis examining the quality of
anticoagulation control among the different ethnic groups. However, for the second part of the
analysis, investigating the predictors of TTR, ethnicity was a covariate along with the other
demographics and clinical characteristics.
3.3.4.3 Predictors: Patient demographic and clinical factors
Patients’ age was calculated according from the date of the first of INR available from the VKA
therapy. Elderly patients were defined as patients who are ≥80 years. Information on gender
(male and female), smoking (smoking within 2 years and non-smoking) and alcohol history (no
alcohol, alcohol within recommended units i.e.; 14 units per week for both men and women or
above recommended units) were obtained directly from CDA. Smoking history was available
for 717 patients (72.4%). For calculation of the SAMe-TT2R2 score, that required information
on smoking status, all missing information on smoking status was coded as non-smoker
(N=274). Information on alcohol intake was only available for 58.0% patients, thus this variable
was excluded from further analysis. Comorbid conditions at baseline including hypertension,
heart failure, coronary artery disease/ischemic heart disease, stroke/TIA, prior bleeding history
and anaemia were obtained directly from CDA.
Other comorbid diseases like kidney disease, liver disease and anaemia were also assumed
based on available laboratory results at baseline. A correction factor to eGFR values was
179
made for Afro-Caribbean patients by multiplying the eGFR obtained from CDA by 1.21 (424).
Assumption of CKD was made if patients had eGFR <60ml/min or serum creatinine
>200umol/L; liver disease if abnormal liver function tests were reported; if alanine
transaminase/alkaline phosphatase (ALT/ALP) >x3 upper limit of normal (ULN), and anaemia
if the haemoglobin level was <135 g/L for males and <115 g/L for females. Laboratory results
and medications taken at baseline were obtained directly from the CDA.
3.3.4.3.1 Chronic kidney disease (CKD)
Further categorisation of kidney disease was made according to the ‘Kidney Disease
Improving Global Outcomes’ (KDIGO) GFR categories adapted by the NICE guidelines, with
five categories of kidney disease for the ancillary analysis (Table 3.2) (424).
Table 3.2: Categories of chronic kidney disease from the NICE guidelines (424)
GFR category GFR (ml/min) Terms

G1 ≥90 Normal or high
G2 60-89 Mildly decreased
G3a 45-59 Mild to moderately decreased
G3b 30-44 Moderate to severely decreased
G4 15-29 Severely decreased
G5 <15 Kidney failure
GFR: glomerular filtration rate
3.3.4.4 Other patient clinical factors (not included as predictors of

anticoagulation control)
Type of AF, including paroxysmal, persistent, long-standing persistent and permanent, was
obtained from the medical notes. If this information was not available, an assumption was
made based on the length of time since AF diagnosis and the pattern of ECG recordings
available with confirmation from a medical doctor and according to the ESC AF guidelines (3).
For example, if the available ECG showed multiple episodes of AF which lasted less than 48
hours, this was categorised as paroxysmal AF. If AF lasted longer than 7 days, including those
180
terminated by cardioversion (either with drugs or direct current cardioversion), this was
categorised as persistent AF. Whereas for long-standing persistent AF, the AF was
continuous, lasting ≥1 year and a decision had been made to adopt a rhythm control strategy.
Lastly permanent AF was defined when AF was accepted by both physician and patient and
a decision has been made to not continue with rhythm control therapy (3). Calculation of stroke
risk, bleeding risk and quality of anticoagulation control was made according to the CHA2DS2-
VASc score, HAS-BLED score, and TTR (Rosendaal and PINNR methods), respectively.
The individual CHA2DS2-VASc score (to predict stroke risk) was calculated as follows: one
point each for the presence of congestive heart failure, hypertension, diabetes mellitus,
vascular disease (defined as peripheral vascular disease, myocardial infarction or aortic
plaque), age 65-74 and female sex, and two points each for the presence of age≥75 years and
previous stroke/TIA (Table 3.1).
The individual HAS-BLED score (to predict bleeding risk) was calculated as follows: one point
each for the presence of uncontrolled hypertension (for all cases blood pressure was assumed
to be controlled as this is a requirement when on VKA therapy), abnormal kidney was defined
as serum creatinine >200umol/L, abnormal liver function as ALT/ALP >x3 ULN) and from the
past medical history (i.e., cirrhosis), previous stroke, prior bleeding within 12 months (including
recent diverticulitis, gastric ulcer or anaemia defined as haemoglobin level of <135 g/L for male
and <115 g/L for female and from the medical history), labile INR or TTR<60%, elderly >65
years and drugs (concurrent antiplatelet or NSAIDs) and/or alcohol (>14 units/week,
recommended by the current guidelines; modified from the original HAS-BLED score of >8
units/week). There was insufficient information about alcohol intake, and therefore this variable
was not included in the calculation of HAS-BLED score; the maximum HAS-BLED score was
eight. A HAS-BLED score of 0-2 was denoted as low risk of bleeding and ≥3 as high risk of
bleeding (Table 3.1).
181
The individual SAMe-TT2R2 score (to predict quality of anticoagulation control) was calculated
as follows: one point each for female sex; presence of ≥2 of the following medical conditions:
hypertension, diabetes mellitus, coronary artery disease or myocardial infarction, peripheral
artery disease, congestive heart failure, previous stroke, pulmonary disease, hepatic disease
or renal disease; treatment with interacting drugs e.g. amiodarone for rhythm control; and two
points each for tobacco use within 2 years (tobacco use is a combination of current smoking
and ex-smoking but with unknown duration as this information is not available); and race (non-
white), giving a possible score ranging from 0-8. Patients with SAMe-TT2R2 score of 0-2 were
defined as being likely to do well on VKA therapy and those with a SAMe-TT2R2 score >2 were
classified as at risk of suboptimal anticoagulation control (TTR<65%) (Table 3.1).
Normality tests were performed by histogram plot method and the Kolmogorov-Smirnov test
where a bell-shaped distribution in the former and p-values >0.05 in the latter were indicative
of normally distributed data. Normally distributed data were expressed as mean (standard
deviation, SD) and non-parametric data are presented as median (interquartile range, IQR).
Categorical variables were compared using the chi-square test or the Fisher’s Exact test (as
appropriate) and reported as counts with percentages. Continuous variables comparing >2
groups used the analysis of variance (ANOVA) test for normally distributed data, with post-hoc
tests as appropriate (e.g., Bonferroni), while, the Kruskal-Wallis test was utilised for non-
parametric data. Independent t-tests and Wilcoxon signed ranked tests were utilised for
comparing continuous variables within two groups for normally and non-normally distributed
data, respectively.
For baseline demographics, clinical characteristics and adverse clinical outcomes, data were
presented as descriptive statistics. Pearson correlation was utilised to investigate the
correlation between TTR and PINRR (normally distributed).
182
Linear regression analysis was conducted to investigate the predictors of TTR and PINRR as
a continuous variable. Logistic regression analysis was also used to investigate predictors of
poor TTR (TTR<70%) and PINRR (PINRR <70%) as categorical variables as this cut off
reflects poor anticoagulation control recommended by the European Guidelines (1). The
relationship between INR control (TTR<70% and TTR<65%) and adverse clinical outcomes
was investigated using the Chi-squared test and these analyses were exploratory. Predictors
of MACE events and composite endpoints (≥1 MACE) were examined using Cox proportional
hazard regression models. Survival analysis was also displayed using Kaplan-Meier curves.
P-values <0.05 were considered statistically significant. All analyses were confined to
complete cases only (except for smoking history) and were conducted using SPSS version
23.0 (IBM, NY, USA) (425).
183
3.4 Results
3.4.1 Baseline characteristics
3.4.1.1 By ethnicity
The final cohort included 991 AF patients receiving warfarin for stroke prevention, with a
median (IQR) length of follow up of 5.2 (3.2-7.0) years. The majority of the population were
White (N=807, 81.4%), with 10.3% South-Asian (N=102) and 8.3% Afro-Caribbean (N=82).
The overall mean age at warfarin initiation was 71.6 (9.4) years and South-Asian patients were
significantly younger than Whites and Afro-Caribbeans (p=<0.05 for group comparison); half
the population were male (55.3%) (see Table 3.3). Overall, hypertension was the most
common co-morbidity (79.2%), followed by chronic kidney disease (37.3%) and diabetes
mellitus (20.6%). There was a significant difference in prevalence of diabetes (p<0.001),
anaemia (p<0.001), and vascular disease (p=0.007) between ethnic groups. Diabetes mellitus
and anaemia were significantly more prevalent among South-Asians and Afro-Caribbeans
(p<0.05 for group comparison) compared to Whites and vascular disease was significantly
more prevalent among South-Asian compared to Whites and Afro-Caribbeans (p<0.05 for
group comparison). Smoking status was only available in 72.4% patients. Smoking (current
or ex-smoker within 2 years) appeared to be more prevalent among Whites (51.6%) compared
to South-Asian (14.6%) and Afro-Caribbeans (25.9%; p<0.05 group comparison). Type of AF
differed by ethnicity (p=0.004), with a greater proportion of Afro-Caribbeans (40.2%) found to
be in persistent AF compared to South-Asians and Whites. The only overall significant
difference (p<0.001) in medications was calcium channel blocker prescription with a greater
prevalence among Afro-Caribbeans compared to South-Asians and Whites.
The overall mean (SD) CHA2DS2-VASc score was 3.4 (1.6) and was significantly higher among
Afro-Caribbeans compared to Whites and South-Asians [3.9 (1.7) vs. 3.3 (1.6) vs. 3.6 (1.7),
respectively; p<0.05 for group comparison]. The overall mean (SD) HAS-BLED score was 1.5
(0.9) and was significantly higher in both South-Asians and Afro-Caribbeans compared to
184
Whites [mean (SD) HAS-BLED score of 1.8 (0.9) vs. 1.7 (1.0) vs. 1.5 (0.9) respectively; p<0.05
for group comparison]. The SAMe-TT2R2 score, was significantly higher among South-Asians
and Afro-Caribbeans compared to Whites (p<0.05 for group comparison) (see Table 3.3).
185
Table 3.3: Baseline characteristics of the study population overall and stratified by ethnicity and age (≥80 vs. <80 years)
Total, White, South-Asian, Afro-Caribbean, Age ≥80 Age <80 Overall Overall
N=991 N=807 N=102 N=82 years, years, p-value p-value
N=205 N=786 ethnicity age
Mean (SD) age 71.6 (9.4) 71.9 (9.3) 68.2 (9.9) a 72.9 (9.3) c - - <0.001 -
<65 209 (21.1) 166 (20.6) 29 (28.4) 14 (17.1) - - 0.004 -
65-74 355 (35.8) 287 (35.6) 45 (44.1) 23 (28.0) - - -
≥75 427 (43.1) 354 (43.9) 28 (27.5) a, c 45 (54.9) c - - -
Female 443 (44.7) 343 (42.5) 46 (45.1) c 54 (66.0) b, c 120 (58.5) 323 (41.1) <0.001
<0.001
Male 548 (55.3) 464 (57.5) 56 (54.9) 28 (34.1) 85 (41.5) 463 (58.9)
White - - - - 176 (85.9) 631 (80.3) -
South-Asian - - - - 10 (4.9) 92 (11.7) - 0.016
Afro-Caribbean - - - - 19 (9.3) 63 (8.0) -
Heart failure 138 (13.9) 109 (13.5) 14 (13.7) 15 (18.3) 31 (15.1) 107 (13.6) 0.49 0.66
Hypertension 785 (79.2) 631 (78.2) 82 (80.4) 72 (87.8) 176 (85.9) 609 (77.5) 0.12 0.011
Diabetes 204 (20.6) 132 (16.4) 44 (43.1) a 28 (34.1) b 38 (18.5) 166 (21.1) <0.001 0.47
Stroke/TIA 179 (18.1) 145 (18.0) 23 (22.5) 11 (13.4) 40 (19.5) 139 (17.7) 0.27 0.61
VTE 38 (3.8) 32 (4.0) 3 (2.9) 3 (3.7) 7 (3.4) 31 (3.9) 0.88 0.88
PAD 26 (2.6) 24 (3.0) 1 (1.0) 1 (1.2) 8 (3.9) 18 (2.3) 0.35 0.30
Vascular disease* 163 (16.4) 123 (15.2) 28 (27.5) a 12 (14.6) b 37 (18.0) 126 (16.0) 0.007 0.56
Lung disease# 196 (19.8) 165 (20.4) 12 (11.8) 19 (23.2) 34 (16.6) 162 (20.6) 0.08 0.23
186
Table 3.3 continued
Total, White, South-Asian, Afro- Age ≥80 years, Age <80 Overall p- Overall p-
N=991 N=807 N=102 Caribbean, N=205 years, value value
N=82 N=786 ethnicity age
Cardiomyopathy‡ 30 (3.0) 25 (3.1) 2 (2.0) 3 (3.7) 4 (2.0) 26 (3.3) 0.77 0.44
Kidney disease† 370 (37.3) 308 (38.2) 40 (39.2) 22 (26.8) 103 (50.2) 267 (34.0) 0.12 <0.001
Anaemia 145 (14.6) 101 (12.5) 28 (27.5)a 16 (19.5)b 34 (16.6) 111 (14.1) <0.001 0.44
Smoker/ex-smoker 326 (45.5) 300 (51.6) 12 (14.6)a 14 (25.9)b 49 (33.8) 277 (48.4) <0.001 0.002
(N=717)
Paroxysmal 274 (27.6) 225 (27.9) 31 (30.4) 18 (22.0) 48 (23.4) 226 (28.8)
Persistent 229 (23.1) 174 (21.6) 22 (21.6)c 33 (40.2)b, c 47 (22.9) 182 (23.2) 0.004 0.26
Permanent 488 (49.2) 408 (50.6) 49 (48.0) 31 (37.8) 110 (53.7) 378 (48.1)
ACEI/ARB 561 (56.6) 449 (55.6) 62 (60.8) 50 (61.0) 115 (56.1) 446 (56.7) 0.43 0.93
Beta-blocker 455 (45.9) 360 (44.6) 57 (55.9) 38 (46.3) 87 (42.4) 368 (46.8) 0.10 0.30
CCB 350 (35.3) 264 (32.7) 39 (38.2)c 47 (57.3)b, c 82 (40.0) 268 (34.1) <0.001 0.14
Digoxin 226 (22.8) 194 (24.0) 18 (17.6) 14 (17.1) 43 (21.0) 183 (23.3) 0.15 0.54
Diuretics 439 (44.3) 351 (43.5) 42 (41.2) 46 (56.1) 120 (58.8) 319 (40.6) 0.07 <0.001
Amiodarone 58 (5.9) 52 (6.4) 3 (2.9) 3 (3.7) 7 (3.4) 51 (6.5) 0.25 0.13
Concurrent 46 (4.6) 38 (4.7) 6 (5.9) 2 (2.4) 10 (4.9) 36 (4.6) 0.53 1.00

antiplatelet
187
Table 3.3 continued
Total, White, South-Asian, Afro- Age ≥80 years, Age <80 years, Overall p- Overall
N=991 N=807 N=102 Caribbean, N=205 N=786 value p-value
N=82 ethnicity age
Mean (SD) CHA2DS2- 3.4 (1.6) 3.3 (1.6) 3.6 (1.7) 3.9 (1.7)b 4.4 (1.3) 3.1 (1.6) 0.002 <0.001
VASc
0 29 (2.9) 25 (3.1) 2 (2.0) 2 (2.4) 0 29 (3.7)
1 81 (8.2) 72 (8.9) 5 (4.9) 4 (4.9) 0 81 (10.3)
2 185 (18.7) 150 (18.6) 27 (26.5) 8 (9.8) 12 (5.9) 173 (22.0)
3 238 (24.0) 200 (24.8) 20 19.6) 18 (22.0) 37 (18.0) 201 (25.6)
4 225 (22.7) 183 (22.7) 17 (16.7) 25 (30.5) 68 (33.2) 157 (20.0)
0.001 <0.001
5 124 (12.5) 101 (12.5) 12 (11.8) 11 (13.4) 46 (22.4) 78 (9.9)
6 82 (8.3) 62 (7.7) 13 (12.7) 7 (8.5) 28 (13.7) 54 (6.9)
7 22 (2.2) 11 (1.4) 6 (5.9) 5 (6.1) 12 (5.9) 10 (1.3)
8 4 (0.4) 2 (0.2) 0 2 (2.4) 1 (0.5) 3 (0.4)
9 1 (0.1) 1 (0.1) 0 0 1 (0.5) 0
CHA2DS2-VASc 39 (3.9) 32 (4.0) 3 (2.9) 4 (4.9) 0 39 (5.0)
categories: Low risk
Intermediate 71 (7.2) 65 (8.1) 4 (3.9) 2 (2.4) 0 71 (9.0) 0.214 <0.001
High risk 881 (88.9) 710 (88.0) 95 (93.1) 76 (92.7) 205 (100) 676 (86.0)
Mean (SD) HAS-BLED 1.5 (0.9) 1.5 (0.9) 1.8 (0.9)a 1.7 (1.0)b 1.8 (0.8) 1.5 (0.9) <0.001 <0.001
0 93 (9.4) 79 (9.8) 8 (7.8) 6 (7.3) 0 93 (11.8)

1 443 (44.7) 379 (47.0) 34 (33.3) 30 (36.6) 97 (47.3) 346 (44.0)
2 326 (32.9) 263 (32.6) 32 (31.4) 31 (37.8) 68 (33.2) 258 (32.8)
<0.001 <0.001
3 107 (10.8) 71 (8.8) 25 (24.5) 11 (13.4) 32 (15.6) 75 (9.5)
4 20 (2.0) 13 (1.6) 3 (2.9) 4 (4.9) 7 (3.4) 13 (1.7)
5 2 (0.2) 2 (0.2) 0 0 1 (0.5) 1 (0.1)
188
Table 3.3 continued
Total, White, South-Asian, Afro- Age ≥80 years, Age <80 years, Overall p- Overall
N=991 N=807 N=102 Caribbean, N=205 N=786 value p-value
N=82 ethnicity age
HAS-BLED categories 862 (87.0) 721 (89.3) 74 (72.5)a 67 (81.7) 165 (80.5) 697 (88.7)
Low risk <0.001 0.003
High risk 129 (13.0) 86 (10.7) 28 (27.5)a, c 15 (18.3)b, c 40 (19.5) 89 (11.3)
Mean SAMe-TT2R2 2.3 (1.4) 2.0 (1.2) 3.7 (0.9)a 3.8 (0.9)b 2.2 (1.2) 2.4 (1.4) <0.001 0.04
score
0 74 (7.5) 74 (9.2) 0 0 14 (6.8) 60 (7.6)
1 231 (28.6) 0 0 0 49 (23.9) 182 (23.2)
2 246 (24.8) 236 (29.2) 6 (5.9) 4 (4.9) 72 (35.1) 174 (22.1)
3 235 (23.7) 167 (20.7) 42 (41.2) 26 (31.7) 36 (17.6) 199 (25.3)
<0.001 0.004
4 162 (16.3) 87 (10.8) 37 (36.3) 38 (46.3) 31 (15.1) 131 (16.7)
5 35 (3.5) 12 (1.5) 13 (12.7) 10 (12.2) 3 (1.5) 32 (4.1)
6 6 (0.6) 0 3 (2.9) 3 (3.7) 0 6 (0.8)
7 2 (0.2) 0 1 (1.0) 1 (1.2) 0 2 (0.3)
SAMe-TT2R2 551 (55.6) 541 (67.0) 6 (5.9)a 4 (4.9)b 135 (65.9) 416 (52.9) <0.001 0.001
categories 0-2
>2 440 (44.4) 266 (33.0) 96 (94.1)a 78 (95.1)b 70 (34.1) 370 (47.1)
ACEI/ARB: angiotensin converting enzyme inhibitor/ angiotensin receptor blockade; CCB: calcium channel blocker; CHA2DS2-VASc score - Congestive heart failure/left ventricular dysfunction,
Hypertension, Age ≥75years [2 points], Diabetes, Stroke [2 points], Vascular disease, Age 65–74 years, and Sex category (female). Total scores range between 0-9; low risk CHA2DS2-VASc score: 0,
intermediate: 1, high risk CHA2DS2-VASc score: ≥2; TIA: transient ischemic attack; TE: thromboembolism; HAS-BLED score – uncontrolled Hypertension: systolic ≥160 mmHg, Abnormal renal/liver
function, Stroke, Bleeding history or predisposition, Labile INR ratio/TTR <60, Drugs/alcohol concomitantly. Total scores range between 0-9; low risk of bleeding range between 0-2 and high risk of
bleeding ≥3; SAMe-TT2R2 score – Sex female, Age<60, Medical history (more than two comorbidities), Treatment (interacting drug, e.g. Amiodarone), Tobacco use (doubled) and Race (non-white,
doubled). Total scores ranged from 0-8; probable good response to VKA therapy range between 0-2 and probable poor response to VKA therapy ranged from ≥3.
* Vascular disease: prior myocardial infarction, peripheral artery disease or aortic plaque; † Kidney disease: eGFR<60 ml/min; ‡ Cardiomyopathy: dilated, restrictive and obstructive myocardial conditions;
# Lung disease Includes obstructive and restrictive diagnosed lung conditions;
a
significant difference between White and South-Asian groups (p<0.05); b significant difference between White and Afro-Caribbean groups (p<0.05); c significant difference between South-Asian and
Afro-Caribbean groups (p<0.05)
189
3.4.1.2 By age (≥80 years and <80 years)
Baseline characteristics of the population grouped by ≥80 years and <80 years are also shown
in Table 3.3. There were 205 patients (20.6%) who were ≥80 years old and the majority were
female (58.5%; p<0.001) and of white ethnicity (85.9%; p=0.016). Hypertension (85.9%;
p=0.011) and chronic kidney disease (50.2%; p=<0.001) were significantly more prevalent
among elderly patients, whereas smoking history (48.4%; p=0.002) was significantly more
prevalent in patients aged <80 years. As expected, the mean (SD) CHA2DS2-VASc [4.4 (1.3);
p<0.001] and HAS-BLED score [1.8 (0.8); <0.001] were significantly higher among elderly
patients, whereas the mean SAMe-TT2R2 score was significantly higher in the younger
population.
3.4.1 By kidney disease categories
There were 974 patients with eGFR results available at baseline. The distribution of patients
according to the classification of the kidney disease is shown in Table 3.4. These categories
were further dichotomised into normal kidney function with eGFR ≥90 ml/min/1.73m2; mildly
decreased with eGFR 60-89 ml/min/1.73m2 and combination of group 3a, 3b, 4 and 5 (mild-
moderately-severe and kidney failure) with eGFR ≤59 ml/min/1.73m2.
Table 3.4: Distribution of patients in the current cohort according to the categories of
kidney disease, N=974
GFR category GFR N (%) New dichotomised group in this N (%)

(ml/min/1.73m2) cohort, eGFR (ml/min/1.73m2)
G1 ≥90 133 (13.7) Normal (≥90) 133 (13.7)
G2 60-89 491 (50.4) Mild (60-89) 491 (50.4)
G3a 45-59 225 (23.1) Mild-moderate-severe and kidney 350 (35.9)
G3b 30-44 98 (10.1) failure (≤59)
G4 15-29 24 (2.5)
G5 <15 3 (0.3)
GFR: glomerular filtration rate
190
For the purpose of the thesis, the description of the different kidney function groups will be
patients with eGFR ≥90, eGFR 60-89 and eGFR ≤59. The median (IQR) eGFR at baseline
was 66.0 (54.0-79.0) ml/min/1.73m2. Half of the population had eGFR 60-89 ml/min/1.73m2,
while 35.9% had an eGFR ≤59 ml/min/1.73m2.
Compared to patients with normal kidney function, AF patients with eGFR 60-89 and eGFR
≤59 were significantly older (p<0.05 for group comparison). There were significantly more
females and those of white ethnicity in patients with eGFR ≤59 compared to eGFR ≥90 (p<0.05
for group comparison) (Table 3.5). Heart failure, hypertension and anaemia were significantly
more prevalent in patients with eGFR ≤59 compared to eGFR ≥90 and eGFR 60-89 (p<0.05
for group comparison). Meanwhile, stroke/TIA was significantly more prevalent in patients with
eGFR ≥90 and eGFR ≤59 compared to eGFR 60-89 (p<0.05 for group comparison). A
significant proportion of AF patients with eGFR ≥90 had concomitant lung disease and also
smoked (or used to smoke within 2 years) compared eGFR 60-89 and eGFR ≤59 (p<0.05 for
group comparison) (Table 3.5).
In addition, the use of ACE/ARB and diuretics were significantly more prevalent in patients
with eGFR ≤59 (p<0.05 for group comparison) while the use of beta blockers (p<0.05 for group
comparison) was significantly more prevalent in patients with eGFR ≥90. Mean (SD) CHA2DS2-
VASc score [3.9 (1.6); p<0.05 group comparison] and HAS-BLED score [1.7 (0.9); p<0.05 for
group comparison] was significantly higher in patients with eGFR ≤59 while mean SAMe-TT2R2
score was significantly higher in patients with eGFR ≥90 (Table 3.5).
191
Table 3.5: Baseline characteristics of overall population with eGFR results and according to three categories of kidney disease
Total, eGFR≥90 ml/min eGFR 60-89 ml/min eGFR ≤59 ml/min Overall
N (%) N=974 N=133 N=491 N=350
p-value
a b, c
Mean (SD) 71.6 (9.4) 67.5 (10.1) 70.7 (9.4) 74.4 (8.4) <0.001
b
<65 205 (21.0) 46 (34.6) 116 (23.6) 43 (12.3)
Age
65-74 351 (36.0) 49 (36.8) 191 (38.9) 111 (31.7) <0.001
≥75 418 (42.9) 38 (28.6) 184 (37.5) 196 (56.0)b
Female 437 (44.9) 47 (35.3) 209 (42.6) 181 (51.7) b
Sex 0.002
Male 537 (55.1) 86 (64.7) 282 (57.4) 169 (48.3) b
White 792 (81.3) 89 (66.9) 403 (82.1)a 300 (85.7)b
Ethnic groups South-Asian 102 (10.5) 13 (9.8) 52 (10.6) 37 (10.6) <0.001
a b
Afro-Caribbean 80 (8.2) 31 (23.3) 36 (7.3) 13 (3.7)
Heart failure 135 (13.9) 15 (11.1) 55 (11.2) 65 (18.6)b,c 0.006
Hypertension 774 (79.5) 101 (75.9) 376 (76.6) 297 (84.9)b,c 0.008
Diabetes 202 (20.7) 26 (19.5) 100 (20.4) 76 (21.7) 0.84
Stroke/TIA 175 (18.0) 26 (19.5) 70 (14.3)a 79 (22.6)b,c 0.007
VTE 36 (3.7) 4 (3.0) 22 (4.5) 10 (2.9) 0.42
Medical history PAD 26 (2.7) 3 (2.3) 8 (1.6) 15 (4.3) 0.06
Vascular disease* 160 (16.4) 16 (12.0) 76 (15.5) 68 (19.4) 0.11
# a b,c
Lung disease 194 (19.9) 36 (27.1) 101 (20.6) 57 (16.3) 0.03
‡
Cardiomyopathy 29 (3.0) 4 (3.0) 16 (3.3) 9 (2.6) 0.85
Anaemia 145 (14.9) 17 (12.8) 62 (12.6) 66 (18.9)b,c 0.03
Smoker/ex-smoker 319 (45.1) 54 (51.9) 169 (47.5) 96 (38.7) 0.03
(N=708)
192
Table 3.5 continued
N=974 N=133 N=491 N=350
p-value
Paroxysmal 274 (28.1) 37 (27.8) 143 (29.1) 94 (26.9)
Types of AF Persistent 228 (23.4) 37 (27.8) 114 (23.2) 179 (51.1) 0.59
Permanent 472 (48.5) 59 (44.4) 234 (47.7) 77 (22.0)
ACEI/ARB 553 (56.8) 67 (50.4) 270 (55.0) 216 (61.7) 0.04
a c
Beta-blocker 454 (46.6) 69 (51.9) 207 (42.2) 178 (50.9) 0.02
CCB 347 (35.6) 45 (33.8) 175 (35.6) 127 (36.3) 0.88
Digoxin 223 (22.9) 25 (18.8) 108 (22.0) 90 (25.7) 0.22
Medications
Diuretics 433 (44.5) 42 (31.6) 192 (39.1) 199 (56.9)b,c <0.001
Amiodarone 58 (6.0) 6 (4.5) 27 (5.5) 25 (7.1) 0.46
Concurrent 44 (4.5) 6 (4.5) 19 (3.9) 19 (5.4) 0.56
antiplatelet
Mean (SD) 3.4 (1.6) 2.9 (1.5) 3.2 (1.6) 3.9 (1.6) b,c <0.001
0 27 (2.8) 9 (6.8) 16 (3.3) 2 (0.6)
1 80 (8.2) 14 (10.5) 51 (10.4) 15 (4.3)
2 183 (18.8) 29 (21.8) 104 (21.2) 50 (14.3)
3 235 (24.1) 35 (26.3) 125 (25.5) 75 (21.4)
CHA2DS2-VASc
4 222 (22.8) 25 (18.8) 101 (20.6) 96 (27.4)
score <0.001
5 120 (12.3) 14 (10.5) 55 (11.2) 51 (14.6)
6 81 (8.3) 6 (4.5) 30 (6.1) 45 (12.9)
7 21 (2.2) 1 (0.8) 8 (1.6) 12 (3.4)
8 4 (0.4) 0 1 (0.2) 3 (0.9)
9 1 (0.1) 0 0 1 (0.1)
193
Table 3.5 continued
N=974 N=133 N=491 N=350
p-value
b
CHA2DS2-VASc Low risk (0) 37 (3.8) 11 (8.3) 23 (4.7) 3 (0.9)
score categories Intermediate (1) 70 (7.2) 12 (9.0) 44 (9.0) 14 (4.0)b,c <0.001
High risk (≥2) 867 (89.0) 110 (82.7) 424 (86.4) 333 (95.1)b
HAS-BLED score Mean 1.5 (0.9) 1.5 (0.9) 1.4 (0.8) 1.7 (0.9)c <0.001
0 91 (9.3) 14 (10.5) 61 (12.4) 16 (9.3)
1 434 (44.6) 56 (42.1) 236 (48.1) 142 (40.6)
2 322 (33.1) 46 (34.6) 148 (30.1) 128 (36.6)
<0.001
3 106 (10.9) 14 (10.5) 41 (8.4) 51 (14.6)
4 19 (2.0) 2 (1.5) 5 (1.0) 12 (3.4)
5 2 (0.2) 1 (0.8) 0 1 (0.3)
HAS-BLED score Low risk (0-2) 847 (87.0) 116 (87.2) 445 (90.6) 286 (81.7)
0.001
categories High risk (≥3) 127 (13.0) 17 (12.8) 46 (9.4) 64 (18.3)
Mean 2.3 (1.3) 2.7 (1.5) 2.2 (1.3)a 2.4 (1.2)c <0.001
0 69 (7.1) 12 (9.0) 51 (10.4) 6 (1.7)
1 230 (23.6) 20 (15.0) 123 (25.1) 87 (24.9)
2 244 (25.1) 24 (18.0) 114 (23.2) 106 (30.3)
SAMe-TT2R2
3 228 (23.4) 36 (27.1) 119 (34.2) 73 (20.9)
score <0.001
4 160 (16.4) 29 (21.8) 65 (13.2) 66 (18.9)
5 35 (3.6) 9 (6.8) 16 (3.3) 10 (2.9)
6 6 (0.6) 2 (1.5) 3 (0.6) 1 (0.3)
7 2 (0.2) 1 (0.8) 0 1 (0.3)
194
Table 3.5 continued
N=974 N=133 N=491 N=350 p-value
SAMe-TT2R2 0-2 543 (55.7) 56 (42.1) 288 (58.7) 199 (56.9)
0.003
score categories >2 431 (44.3) 77 (57.9) 203 (41.3) 151 (43.1)
eGFR ≥90ml/min/1.73m2- normal kidney function; eGFR 60-89 ml/min/1.73m2- mild kidney disease; eGFR ≤59 ml/min/1.73m2- mild-moderate-severe and kidney failure
ACEI/ARB: angiotensin converting enzyme inhibitor/ angiotensin receptor blockade; CCB: calcium channel blocker; CHA2DS2-VASc score - Congestive heart failure/left ventricular dysfunction,
Hypertension, Age ≥75years [2 points], Diabetes, Stroke [2 points], Vascular disease, Age 65–74 years, and Sex category (female). Total scores range between 0-9; low risk CHA2DS2-VASc score:
0, intermediate 1, high risk CHA2DS2-VASc score: ≥2; TIA: transient ischemic attack; TE: thromboembolism; eGFR: estimated glomerular filtration rate, ml/min/1.73 m2; HAS-BLED score – uncontrolled
Hypertension: systolic ≥160 mmHg, Abnormal renal/liver function, Stroke, Bleeding history or predisposition, Labile INR ratio/TTR <60, Drugs/alcohol concomitantly. Total scores range between 0-9;
low risk of bleeding range between 0-2 and high risk of bleeding ≥3; SAMe-TT2R2 score – Sex female, Age<60, Medical history (more than two comorbidities), Treatment (interacting drug, e.g.
Amiodarone), Tobacco use (doubled) and Race (non-white, doubled). Total scores ranged from 0-8; probable good response to VKA therapy range between 0-2 and probable poor response to VKA
therapy ranged from ≥3.
* Vascular disease: prior myocardial infarction, peripheral artery disease or aortic plaque; ‡ Cardiomyopathy: dilated, restrictive and obstructive myocardial conditions; # Lung disease: obstructive and
restrictive diagnosed lung conditions;
a: significant difference between eGFR≥90 and eGFR 60-89, p<0.05

b: significant difference between eGFR≥90 and eGFR ≤59; p<0.05
c: significant difference between eGFR 60-89, and eGFR ≤59; p<0.05
195
3.4.2 Measures of anticoagulation control
In this cohort, 96% of patients were VKA naïve. Table 3.6 presents the measures of
anticoagulation control in the overall population and by ethnic group. Overall mean (SD) TTR
and PINRR values were 66.6% (13.3) and 57.6% (11.2) respectively. TTR significantly
correlated with PINRR (r=0.773, p<0.001). White patients had significantly higher mean TTR
values compared to Afro-Caribbean and South-Asian patients, based on both mean TTR (by
Rosendaal method; p<0.05 group comparison) and PINRR (p<0.05 for group comparison)
(Figure 3.2). When TTR and PINRR was dichotomised (<70% vs. ≥70%) the same trend was
observed (Figure 3.3). The mean (SD) number of INR tests used to calculate TTR was 58.7
(25.5) and was similar across ethnic groups. The proportion of sub-therapeutic INRs (<2.0)
was significantly greater among South-Asians and Afro-Caribbeans, with no differences in the
proportion of supra-therapeutic INRs by ethnicity (Figure 3.4). Overall, 29.6% and 4.1% of the
population had at least one INR value above 5.0 and 8.0, respectively (Figure 3.5).
196
Table 3.6: Measures of anticoagulation control overall and by ethnic group
F- X2 p-
South-Asian, Afro-Caribbean,
Measures of anticoagulation control, N (%) Total, N=991 White, N=807 value value value
N=102 N=82
Mean (SD) TTR Rosendaal 66.6 (13.2) 67.9 (12.8) 60.5 (12.8)a 61.3 (14.2)b 22.7 - <0.001
TTR<70% 550 (55.5) 417 (51.7) 75 (73.5) 58 (70.7) <0.001
a b
25.9
TTR≥70% 441 (44.5) 390 (48.3) 27 (26.5) 24 (29.3)
TTR<65% 400 (40.4) 294 (36.4) 59 (57.8) 47 (57.3) <0.001
27.9
TTR≥65% 591 (59.6) 513 (63.6) 43 (42.2)a 35 (42.7)b
Mean (SD) PINRR 57.6 (11.2) 58.8 (10.8) 51.6 (10.9)a 53.1 (11.6)b 27.3 - <0.001
PINRR<70% 851 (85.9) 677 (83.9) 99 (97.1) 75 (91.5) <0.001
a, c b,c
PINRR ≥70% 140 (14.1) 130 (16.1) 3 (2.9) 7 (8.5) 15.3
PINRR <65% 736 (74.3) 576 (71.4) 91 (89.2) 69 (84.1) <0.001
a b 19.6
PINRR ≥65% 255 (25.7) 231 (28.6) 11 (10.8) 13 (15.9)
Mean (SD) number of INR tests 58.7 (25.5) 59.4 (24.6) 55.0 (24.6) 56.5 (33.7) 1.7 - 0.18
a b
Mean (SD) percentage INRs<2 25.7 (10.0) 24.7 (9.5) 30.1 (11.2) 30.2 (10.9) 23.8 - <0.001
Mean (SD) percentage INRs>3 16.6 (7.2) 16.5 (7.3) 17.9 (7.1) 16.5 (6.2) 1.9 - 0.16
INR>5 293 (29.6) 239 (29.6) 31 (30.4) 23 (28.0) - 0.125 0.94
INR>8 41 (4.1) 36 (4.5) 3 (2.9) 2 (2.4) - 1.177 0.56
a b
Median (IQR) years of follow-up 5.2 (3.2-7.0) 5.5 (3.4-7.0) 4.3 (2.6-6.7) 4.0 (2.4-6.1) - - <0.001
TTR: Time in therapeutic range; PINRR: Percentage of INRs within range; INR: international Normalised Ratio; SD: standard deviation; IQR: interquartile range
a
significant difference between White and South-Asian groups (p<0.05); b significant difference between White and Afro-Caribbean groups (p<0.05); c significant difference between South-Asian and Afro-Caribbean
groups (p<0.05)
197
87.5 Mean % PINRR Mean INR test Mean % INR <2.0 Mean %TTR
*p<0.001 across the groups

68*
70
61* 61*
59* 59
57
55 53*
52*
52.5
Mean
35 30* 30*
25*
17.5
0
White South-Asian Afro-Caribbean
Ethnic groups
Figure 3.2: Measures of anticoagulation control (including TTR by Rosendaal and PINRR method) by
ethnicity
60%
*p<0.001 across the groups

50% 48%*
40%
% patients
29%*
30% 27%*
TTR≥70%
PINRR ≥70%
20%
16%*
10% 9%*
3%*
0%
Ethnic groups
Figure 3.3: Percentage of patients by ethnic group with a therapeutic (TTR≥70%) TTR by the
Rosendaal and PINRR method
198
70.0%
58.8%
60.0%
51.6% 53.1%
50.0% *p<0.05
% of INRs
40.0%
* INRs<2.0
*
30.1% 30.2%
30.0% * INRs>3.0
24.7%
PINRR
20.0% 16.5% 17.9% 16.5%
10.0%
0.0%
Ethnic groups
Figure 3.4: Percentage of INRs within therapeutic range (PINRR), and below (INR<2.0) and above
(INR >3.0) therapeutic range
35%
30%
30%
30% 28%
25%
% of patients
20%
INR>5
15%
INR>8
10%
5%
5% 3% 2%
0%
Ethnic groups
Figure 3.5: Percentage of patients with an INR >5.0 and >8.0
199
Measures of anticoagulation control continued
3.4.2.2 By age (≥80 and <80 years)
Table 3.7 presents the measures of anticoagulation control when patients were grouped
according to age ≥80 and <80 years. The quality of anticoagulation control by both measures,
TTR (Rosendaal method) [66.6% in ≥80 and <80 years age group] and PINRR [57.1% in the
≥80 years old vs. 57.7% in <80 years old] were similar between the two age categories. Elderly
patients had significantly fewer INR visits (mean 51 vs. 61 visits; p<0.001 for ≥80 vs. <80
years, respectively) and a lower duration of follow up (Figure 3.6). Good TTR (defined as TTR
and PINRR ≥70%) was 44% and 14% in those aged ≥80 years and <80 years, respectively;
over half of the elderly population did not achieve the optimal percentage TTR advocated by
clinical guidelines. No significant differences in sub-therapeutic or supra-therapeutic INRs
were observed by age (≥80 and <80 years).
Table 3.7: Measures of anticoagulation control among overall population and in

patients aged ≥80 and <80 years
N, (%) Age ≥80, Age <80,

p-value
N=205 N=786
Mean TTR (SD) 66.6 (13.8) 66.6 (13.1) 1.00
TTR<70 114 (55.6) 436 (55.5)
1.00
TTR≥70 91 (44.4) 350 (44.5)
Mean PINRR (SD) 57.1 (11.6) 57.7 (11.1) 0.54
PINRR<70 176 (85.9) 675 (85.9)
1.00
PINRR≥70 29 (14.1) 111 (14.1)
Mean (SD) number of visits 51.2 (22.7) 60.7 (25.8) <0.001
Mean (SD) percentage of 26.6 (9.8) 25.5 (24.5) 0.17
INRs<2
Mean (SD) percentage of 16.4 (15.6) 16.7 (7.1) 0.60
INRs>3
INR>5 70 (34.1) 223 (28.4) 0.13
INR>8 10 (4.9) 31 (3.9) 0.69
Median (IQR) years follow up 4.4 (2.6-6.2) 5.7 (3.3-7.1) <0.001

TTR: Time in therapeutic range, PINRR: Percentage of INRs within range; INR: international Normalised Ratio; SD: standard
deviation; IQR: interquartile range
200
p<0.001
70 67
67 61
60 57 57
51
50
Mean TTR %
% TTR/PINRR
Mean PINRR %
40
No. of visits
30
20
10
0
≥80 years <80 years
Age
Figure 3.6: Mean percentage TTR and PINRR and number of visits among patients aged ≥ 80 and <80 years
201
Quality of anticoagulation control continued
3.4.2.3 By categories of kidney disease
Table 3.8 presents the measures of anticoagulation control according to the different
categories of kidney disease. Overall, there was no statistically significant difference in the
quality of anticoagulation control, measured by TTR (Rosendaal’s method), by kidney disease
groups, although TTR was higher in patients with eGFR 60-89 and eGFR ≤59 (67% in both
groups) compared to those in with eGFR ≥90 (64%; overall p=0.053) (Figure 3.7). There was
a significant trend in higher PINRR in patients with eGFR 60-89 and eGFR ≤59 (PINRR 58%
in both groups) compared to those with eGFR ≥90 (PINRR 55%; all p<0.05 for group
comparison) (Table 3.8 and Figure 3.7).
Higher proportions of patients with eGFR 60-89 and eGFR ≤59 achieved good TTR compared
to eGFR ≥90 (p<0.05 for group comparison). Sub-therapeutic INRs (INR <2.0) (Table 3.8 and
Figure 3.8) and the proportion of patients with at least one INR >8.0 were significantly more
prevalent in patients with eGFR ≥90 compared to eGFR 60-89 (p<0.05 for group comparison).
Meanwhile, the proportion of supra-therapeutic INRS >3.0 did not differ by kidney function
groups (Table 3.8 and Figure 3.8).
202
Table 3.8: Measures of anticoagulation control among different categories of kidney disease, N=974
N, (%) All eGFR≥90 ml/min eGFR 60-89 ml/min eGFR ≤59 ml/min Overall
(n=974) N=133 N=491 N=350 p-value
Mean TTR (SD) 66.5 (13.2) 64.0 (14.1) 66.9 (12.7) 67.0 (13.4) 0.053
a b
TTR<70 542 (55.6) 87 (65.4) 263 (53.6) 192 (54.9)
0.05
TTR≥70 432 (44.4) 46 (34.6) 228 (46.4) a 158 (45.1)b
Mean PINRR (SD) 57.4 (11.1) 55.0 (11.5) 57.8 (10.6)a 57.8 (11.7)b 0.02
PINRR<70 842 (86.4) 120 (90.2) 427 (87.0) 295 (84.3)
0.21
PINRR≥70 132 (13.6) 13 (9.8) 64 (13.0) 55 (15.7)
Median (IQR) number 59.0 (41.0-74.0) 61.0 (39.5-78.5) 60.0 (42.0-74.0) 57.5 (41.0-72.0) 0.37
visits*
Mean (SD) percentage of 25.9 (9.9) 27.8 (9.9) 25.4 (9.8)a 25.7 (10.2) 0.05
INRs<2
Mean (SD) percentage of 16.6 (7.2) 17.0 (8.3) 16.6 (7.0) 16.6 (7.1) 0.84
INRs>3
INR>5 287 (29.5) 41 (30.8) 129 (26.3) 117 (33.4) 0.08
INR>8 40 (4.1) 11 (8.3) 14 (2.9)a 15 (4.3)b,c 0.02
Median (IQR) years follow 5.2 (3.2-7.0) 5.0 (2.9-7.0) 5.6 (3.3-7.1) 5.0 (3.2-6.9) 0.18
up*
eGFR ≥90ml/min/1.73m2- normal kidney function; eGFR 60-89 ml/min/1.73m2 - Mild kidney disease; eGFR ≤59 ml/min/1.73m2- mild-moderate-severe and kidney failure
eGFR: estimated glomerular filtration rate, ml/min/1.73 m2; INR: international normalised ratio; IQR: interquartile range; SD: standard deviation; TTR: time in therapeutic range; PINRR: percentage of
INRs in range; *Kruskal-Wallis test was utilised to compare median across the groups
a: significant difference between eGFR≥90 and eGFR 60-89; p<0.05

b: significant difference between eGFR≥90 and eGFR ≤59; p<0.05
c: significant difference between eGFR 60-89 and eGFR ≤59, p<0.05
203
80
*p=0.02
Anticoagulation control measures
70 67 67
64
61 * 60 *
58 58
60 55 * 55
50
40 Mean TTR %
Mean PINRR%
30
Median no of visits
20
10
0
eGFR≥90 eGFR 60-89 eGFR ≤59
eGFR categories
Figure 3.7: Measures of anticoagulation control in different categories of kidney disease
70
60 58 58
Anticoagulation control measures
55
50
*p=0.05
40
Mean PINRR%
30 28 * * Mean INR <2.0
* 26
25
Mean INR >3.0
20 17 17 17
10
0
eGFR≥90 eGFR 60-89 eGFR ≤59
eGFR categories
Figure 3.8: The proportion of INRs within, below and above therapeutic range in different
categories of kidney disease
204
3.4.3 Predictors of time in therapeutic range (TTR) by the Rosendaal and
PINRR methods
In the unadjusted analyses for predictors of TTR and PINRR (as continuous variables), non-
white ethnicity, heart failure, vascular disease, anaemia, and bleeding history, with the addition
of diabetes for PINRR, negatively predicted TTR and PINRR (Table 3.9 and 3.11).
After adjusting for demographic and clinical variables, smoking history, non-white ethnicity,
bleeding history, and heart failure were independent predictors of TTR (Table 3.10), with the
addition of vascular disease as an independent predictor of PINRR (Table 3.12).
When TTR and PINRR were dichotomised to TTR<70% and PINRR<70%, unadjusted
analyses revealed that non-white ethnicity and anaemia significantly predicted poor TTR and
PINRR (TTR<70%). Further factors predicting poor TTR (TTR<70%) were vascular disease
and heart failure (Appendix 4, Table A4.2 and A4.3 for full model). However, after adjusting
for demographic and clinical variables, non-white ethnicity and anaemia remained as
independent predictors of poor TTR (TTR<70%) and PINRR (PINRR<70%), with vascular
disease as an additional factor for predicting poor TTR (Table 3.13).
205
Table 3.9: Unadjusted demographic and clinical predictors of time in therapeutic range (TTR) by the Rosendaal method (linear
regression)
Variables Contribution to Unstandardized Standardized 95% CI for B F Significance p-value

R2 B coefficient Beta
¥
Age at first INR 0.002 0.07 0.05 -0.02 to 0.15 2.15 0.14
Female sex 0.00 0.30 0.01 -1.36 to 1.95 0.12 0.73
Smoking history 0.01 -0.93 -0.04 -2.91 to 1.05 0.86 0.36
Ethnicity (non-white) 0.04 -7.09 -0.21 -9.17 to -5.02 45.10 <0.001
Hypertension 0.001 0.84 0.03 -1.19 to 2.87 0.66 0.42
Stroke/TIA 0.001 1.33 0.04 -0.82 to 3.47 1.48 0.23
Heart failure 0.007 -3.29 -0.09 -5.67 to -0.92 7.42 0.007
Diabetes 0.002 -1.43 -0.04 -3.47 to 0.60 1.90 0.17
Vascular disease 0.004 -2.33 -0.07 -4.55 to -0.11 4.24 0.040
Kidney disease 0.00 0.01 0 -1.70 to 1.71 0 0.10
Anaemia 0.014 -4.42 -0.12 -6.74 to -2.11 14.06 <0.001
Bleeding history 0.008 -4.42 -0.09 -7.45 to -1.39 8.19 0.004
¥ Continuous variable
INR: international Normalised Ratio; TIA: transient ischemic attack; TTR: Time in therapeutic range
206
Table 3.10: Adjusted demographic and clinical predictors of time in therapeutic range (TTR) by the Rosendaal method (linear
regression)
Variables Contribution to Unstandardized Standardized 95% CI for B F Significance p-

2
R B coefficient Beta value
Overall model 0.086 - - - 5.51 -
Age at first INR¥ 0.06 0.04 -0.05 to 0.17 0.27
Female sex -0.25 -0.01 -2.30 to 1.79 0.81
Smoking history -2.49 -0.09 -4.55 to -0.42 0.02
Ethnicity (non-white) -8.09 -0.24 -10.69 to -5.49 <0.001
Hypertension 1.94 0.06 -0.52 to 4.37 0.12
Stroke/TIA 0.91 0.03 -1.56 to 3.38 0.47
Heart failure -3.18 -0.09 -5.81 to -0.54 0.02
Diabetes 1.73 0.05 -0.73 to 4.18 0.17
Vascular disease -1.19 -0.03 -3.87 to 1.49 0.38
Kidney disease -0.30 -0.01 -2.39 to 1.78 0.78
Anaemia -2.62 -0.07 -5.56 to 0.32 0.08
Bleeding history -3.83 -0.08 -7.39 to -0.27 0.04
INR: international Normalised Ratio; TIA: transient ischemic attack TTR: Time in therapeutic range
207
Table 3.11: Unadjusted demographic and clinical predictors of percentage of INRs in range using the PINRR method (linear
regression)
Variables Contribution Unstandardized Standardized 95% CI for B F Significance

2
to R B coefficient Beta p-value
Age at first INR¥ 0.00 0.02 0.01 -0.06 to 0.09 0.15 0.70
Female sex 0.00 0.06 0.00 -1.35 to 1.46 0.01 0.94
Smoking history 0.001 -0.70 -0.03 -2.35 to 0.95 0.69 0.41
Ethnicity (non-white) 0.05 -6.54 -0.23 -8.29 to -4.79 53.75 <0.001
Hypertension 0.001 1.06 0.04 -0.66 to 2.78 1.47 0.23
Stroke/TIA 0.001 0.72 0.03 -1.09 to 2.54 0.61 0.44
Heart failure 0.006 -2.60 -0.08 -4.61 to -0.59 6.42 0.011
Diabetes 0.008 -2.49 -0.09 -4.21 to -0.77 8.06 0.005
Vascular disease 0.011 -3.24 -0.11 -5.11 to -1.36 11.48 0.001
Kidney disease 0.00 -0.15 -0.01 -1.59 to 1.30 0.04 0.84
Anaemia 0.02 -4.12 -0.13 -6.08 to -2.16 17.03 <0.001
Bleeding history 0.02 -5.13 -0.12 -7.69 to -2.57 15.45 <0.001
INR: international Normalised Ratio; TIA: transient ischemic attack; PINRR: Percentage of INRs within range
208
Table 3.12: Adjusted demographic and clinical predictors of percentage of INRs in range using the PINRR method (linear regression)
Variables Contribution Unstandardized Standardized 95% CI for B F Significance

to R2 B coefficient Beta p-value
Overall model 0.10 6.55
¥
Age at first INR 0.01 0.01 -0.08 to 0.10 0.89
Female sex -0.25 -0.01 -1.94 to 1.44 0.77
Smoking history -1.99 -0.09 -3.70 to -0.28 0.02
Ethnicity (non-white) -6.83 -0.24 -8.98 to -4.67 <0.001
Hypertension 2.48 0.09 0.46 to 4.51 0.02
Stroke/TIA 0.91 0.03 -1.14 to 2.95 0.38
Heart failure -2.16 -0.07 -4.34 to 0.02 0.05
Diabetes -0.18 0.01 -2.21 to 1.86 0.86
Vascular disease -2.26 -0.08 -4.48 to- 0.04 0.05
Kidney disease 0.48 0.02 -1.25 to 2.20 0.59
Anaemia -1.77 -0.06 -4.20 to 0.66 0.15
Bleeding history -4.34 -0.11 -7.29 to -1.39 0.004
209
Table 3.13: Logistic regression for significant predictors of TTR<70% and PINRR
<70% (using Rosendaal and PINRR methods)
Adjusted OR p-value Adjusted OR p-value

(95% CI) TTR<70% (95% CI) PINRR<70%
(Rosendaal method) (PINRR method)
Ethnicity (non- white) 2.62 (1.67-4.10) <0.001 3.47 (1.44-8.34) 0.005
Vascular disease 1.81 (1.16-2.83) 0.01 -
Anaemia 1.65 (1.00-2.70) 0.05 6.27 (1.89-20.94) 0.003
3.4.4 Major adverse clinical outcomes

During a median follow-up of 5.2 years, 50 (5.0%) patients had thromboembolic events (46
patients with stroke/TIA, 4 systemic embolisms), 78 (7.9%) experienced bleeding events [18
patients with major bleed and 62 patients with CRNMB], 226 (22.8%) were hospitalised for
cardiovascular reasons, and 23 (2.3%) patients died. Three hundred and twenty-nine patients
(33.2%) presented with ≥1 major adverse clinical event. The overall major adverse clinical
outcomes for entire population and by ethnic group are presented in Table 3.14.
In terms of the number of events, there were 48 (9.3%) stroke/TIA events, 4 (0.7%) systemic
embolism, 91(17.5%) bleeding events [23 (4.4%) major bleed and 68 (13.1%) CRNMB], 353
(68%) CV hospitalisations, and 23 (4.4%) deaths.
There was no significant difference in the rate of major adverse clinical outcomes among the
three ethnic groups except for CV hospitalisations, where the rate was significantly higher
among South Asians compared to Whites (32.3% vs. 21.3%, respectively; p<0.05 for group
difference) and Afro-Caribbeans [32.3% vs. 25.6%, respectively; p<0.05 for group difference]
(Table 3.14).
When events were stratified by TTR (Table 3.15), patients with CV hospitalisations and ≥1
MACE events were more likely to have poor TTR (TTR<70% and TTR<65%) (CV
hospitalisations: 26.5% vs. 18.1%; p=0.002 and 27.3% vs. 19.8%; p=0.008 respectively; and
210
≥1 MACE events: 37.8% vs. 27.4%; p=0.001 and 37.8% vs. 30.1%; p=0.015 respectively).
There were no significant differences in thromboembolic, bleeding events and death when
TTR was dichotomised into TTR<70% and TTR<65% by either the Rosendaal or PINNR
methods. (Table 3.15) CV hospitalisations were also significantly higher when PINRR was
<70% (24.4%; p=0.004) but not with PINRR<65%. (Table 3.16) Hospitalisations due to non-
cardiac causes were significantly higher when TTR<70% or TTR<65% by both the Rosendaal
(Table 3.15) and PINRR methods. (Table 3.16)
211
Table 3.14: Major adverse clinical outcomes among patients receiving warfarin for stroke prevention for AF overall and by ethnic
group
Outcomes, N South-Asian, Afro-Caribbean, p-

Total, N=991 White, N=807
(%) N=102 N=82 value
Stroke/TIA 46 (4.6) 36 (4.5) 5 (4.9) 5 (6.1) 0.79

SE 4 (0.4) 3 (0.4) 0 1 (0.1) 0.41
Stroke/TIA/SE 50 (5.0) 39 (4.8) 5 (4.9) 6 (7.3) 0.62
Bleeding* 78 (7.9) 64 (7.9) 6 (5.9) 8 (9.8) 0.62
Cardiovascular
226 (22.8) 172 (21.3) a 33 (32.3) a 21 (25.6) 0.036
hospitalisation
Death 23 (2.3) 20 (2.5) 2 (2.0) 1 (1.2) 0.75
≥1 MACE 329 (33.2) 258 (32.0) 40 (39.2) 31 (37.8) 0.22
Cardiovascular hospitalisation: a hospitalisation with a cardiovascular cause: i) heart failure, MI, new angina, non-fatal cardiac arrest, ventricular arrhythmia, uncontrolled AF/atrial flutter,
supraventricular arrhythmia, ii) valve surgery, CABG surgery, PTCA surgery, pacemaker/ICD insertion, carotid endarterectomy, peripheral angioplasty/surgery, limb amputation AND as recorded in
patient’s medical documents; DVT – Deep Vein Thrombosis ; Major Bleeding – ISTH Major Bleeding: fatal bleeding and/or symptomatic bleeding in a critical area or organ, such as intracranial,
intraspinal, intraocular, retroperitoneal, intraarticular or pericardial, or intramuscular with compartment syndrome and/or bleeding causing a fall in haemoglobin level of 2 g/dL (1.24 mmol/L) or more,
or leading to transfusion of two or more units of whole blood or red cells; Clinically relevant non-major bleeding (CRNMB): clinically overt bleeding not satisfying the criteria for major bleeding but meet
at least one of the 3 criteria: i) leading to hospitalisation or increased level of care, ii) requiring medical intervention by healthcare professional and iii) prompting face to face evaluation.; PE – Pulmonary
Embolism; SE: systemic embolism; TIA: transient ischemic attack; VTE: venous thromboembolism.* Bleeding ISTH is combination of major bleed ISTH and clinically relevant non-major bleed (CRNMB);
MACE: major adverse clinical events.
a
significant difference between White and South-Asian groups (p<0.05)
212
Table 3.15: Patients experiencing a major adverse clinical event stratified by TTR (<70% vs. ≥70% and <65% and ≥65%)
N (%) TTR<70% TTR≥70% p-value TTR<65% TTR≥65% p-value
CV hospitalisation 146 (26.5) 80 (18.1) 0.002 109 (27.3) 117 (19.8) 0.008
Stroke/TIA 29 (5.3) 17 (3.9) 0.37 18 (4.5) 28 (4.7) 0.98

SE 3 (0.5) 1 (0.2) 0.78 3 (0.8) 1 (0.2) 0.37
Stroke/TIA/SE 32 (5.8) 18 (4.1) 0.27 21 (5.3) 29 (4.9) 0.93
Bleeding* 50 (9.1) 28 (6.3) 0.14 36 (9.0) 42 (7.1) 0.33
Death 16 (2.9) 7 (1.6) 0.25 14 (3.5) 9 (1.5) 0.07
≥1 MACE 208 (37.8) 121 (27.4) 0.001 151 (37.8) 178 (30.1) 0.015
Other hospitalisations 247 (44.9) 155 (35.1) 0.002 179 (44.8) 223 (37.7) 0.032
at least one of the 3 criteria: i) leading to hospitalisation or increased level of care, ii) requiring medical intervention by healthcare professional and iii) prompting face to face evaluation.; PE – Pulmonary
Embolism; SE: systemic embolism; TIA: transient ischemic attack; VTE: venous thromboembolism.*Bleeding ISTH is combination of major bleed ISTH and clinically relevant non-major bleed (CRNMB);
MACE: major adverse clinical events.
213
Table 3.16: Patients experiencing a major adverse clinical event stratified by PINRR (<70% vs. ≥70% and <65% and ≥65%)
N (%) PINRR<70% PINRR≥70% p-value PINRR<65% PINRR≥65% p-value
Stroke/TIA 40 (4.7) 6 (4.3) 1.00 35 (4.8) 11 (4.3) 0.91

SE 3 (0.4) 1 (0.7) 1.00 3 (0.4) 1 (0.4) 1.000
Stroke/TIA/SE 43 (5.1) 7 (5.0) 1.00 38 (5.2) 12 (4.7) 0.90
Bleeding* 71 (8.3) 7 (5.0) 0.23 63 (8.6) 15 (5.9) 0.22
Death 20 (2.4) 3 (2.1) 1.00 19 (2.6) 4 (1.6) 0.49
≥1 MACE 298 (35.0) 31 (22.1) 0.004 259 (35.2) 70 (27.5) 0.029
Other
360 (42.3) 42 (30.0) 0.008 320 (43.5) 82 (32.2) 0.002
hospitalisations
or leading to transfusion of two or more units of whole blood or red cells; Clinically relevant non-major bleeding (CRNMB): clinically overt bleeding not satisfying the criteria for major bleeding and that
led to hospitalisation, physician medical or surgical treatment, or a change in antithrombotic therapy; PE – Pulmonary Embolism; SE: systemic embolism; TIA: transient ischemic attack; VTE: venous
thromboembolism.* Bleeding ISTH is combination of major bleed ISTH and clinically relevant non-major bleed (CRNMB).
214
3.4.4.1.1 Predictors of adverse clinical outcomes
In this exploratory analysis, unadjusted Cox proportional hazard regression analyses revealed
that only prior stroke/TIA [HR 2.40 (95% CI 1.33-4.30); p=0.003] and diabetes [HR 2.01 (95%
CI 1.11-3.65); p=0.021] predicted thromboembolic (stroke/TIA/ systemic embolism) events
(see Appendix 4, Table A4.4). However, only prior stroke/TIA history [HR 2.29 (95% CI 1.12-
4.68); p=0.02] remained as independent predictor of thromboembolic events, after adjusting
for demographic and clinical variables (Table 3.17).
For bleeding events (major bleed and CRNMB), increasing age, TTR and PINRR (as
continuous variables) predicted bleeding events (see Appendix 4, Table A4.5) but after
adjustment, only TTR <70%, [HR 1.78 (95% CI 1.01-3.13); p=0.05] independently predicted
bleeding events (Table 3.17).
In unadjusted Cox proportional hazard regression analyses, non-white ethnicity, heart failure,
diabetes, vascular disease, anaemia, TTR and PINRR (both continuous and categorical
variable) predicted CV hospitalisation (see Appendix 4, Table A4.6). However, after
adjustment, only heart failure [HR 1.46 (95% CI 1.02-2.11); p=0.04], vascular disease [HR
1.62 (95% CI 1.11-2.34); p=0.01] and TTR<70% [HR 1.38 (95% CI 1.00-1.89) p=0.05] were
independent predictors of CV hospitalisations (Table 3.17). Figure 3.9 shows the Kaplan-
Meier curve illustrating the event free rate for CV hospitalisation by TTR <70% and ≥70%. The
rate of CV hospitalisation was significantly higher in patients with poor TTR (6.0/100 pt-yrs)
(Log rank test: 11.90; p=0.001) compared to patients with optimal TTR (TTR≥70%; 3.7/100
pt-yrs).
Only increasing age and anaemia predicted all-cause mortality in the unadjusted Cox
proportional hazard regression model, however after adjustment, none of the factors were
significant predictors of all-cause mortality (see Appendix 4, Table A4.7).
215
When all adverse clinical outcomes were combined as composite events (MACE) in an
unadjusted model, non-white ethnic group, hypertension, stroke/TIA, heart failure, diabetes,
vascular disease, anaemia and both measures of quality of anticoagulation (TTR and PINNR
as continuous and categorical variable) predicted composite events (see Appendix 4, Table
A4.8). However, after adjusting for demographic and clinical variables, only prior stroke/TIA,
vascular disease and TTR<70% predicted ≥1 MACE (Table 3.17). Kaplan-Meier analysis
(Figure 3.10) shows the higher rate of ≥1 MACE in patients with poor TTR (9.1/100 pt-yrs)
(Log rank test: 14.25; p<0.001) compared to patients with TTR≥70% (5.9/100 pt-yrs).
216
Table 3.17: Cox proportional hazard regression analysis for thromboembolic, bleeding events, CV hospitalisations and composite
outcomes of thromboembolic events, major bleed and clinically relevant non-major bleeding, cardiovascular hospitalisation and all-
cause mortality (≥1 MACE)
Multivariate HR (95% CI) Thromboembolic events Bleeding events CV hospitalisations ≥1 MACE

Stroke/TIA 2.29 (1.12-4.68) - - 1.38 (1.03-1.85)
Heart failure - - 1.46 (1.02-2.11) -
Vascular disease - - 1.62 (1.11-2.34) 1.67 (1.21-2.30)
TTR <70% - 1.78 (1.01-3.13) 1.38 (1.00-1.89) 1.45 (1.11-1.89)
217
p=0.001
p<0.001
Cumulative survival probability

TTR<70%
TTR<70%
TTR≥70%
TTR≥70%
Figure 3.9: Impact of TTR on cardiovascular hospitalization Figure 3.10: Impact of TTR on composite endpoints of thromboembolic
events, major and clinically relevant non-major bleeding events,
cardiovascular hospitalisation and all-cause mortality (≥1 MACE)
218
3.4.4.2 Adverse outcomes by age (≥80 years and <80 years)
Only twelve (6%) of the elderly (≥80 years) patients experienced thromboembolic events; 21
(10.2%) had a bleeding event and eight (4.0%) died. The proportion of bleeding (10.2% vs.
7.3% for ≥80 years and <80 years respectively) and fatal (3.9% vs. 1.9% respectively) events
were higher among elderly patients. However, cardiovascular hospitalisations (23.9% vs.
18.5%) and the composite of adverse clinical outcome (≥1MACE) (33.7% vs. 31.2%) were
proportionally higher in the younger age category (Table 3.18).
The Kaplan-Meier curve illustrates the rate of bleeding events which were significantly higher
in the elderly group compared to those aged <80 years (2.4% vs. 1.3%, respectively) (Log
Rank-test: 6.73; p=0.009 Figure 3.11). Univariate Cox regression analysis (see Appendix 4,
Table A4.9 for full model) showed that only age ≥80 years [HR 1.93 (1.16-3.20); p=0.01] was
associated with bleeding risk and this relationship persisted after adjusting for demographic
and clinical variables [≥80 years: HR 1.90 (1.01-3.56); p=0.047] (Table 3.19). History of
stroke/TIA [HR 1.37 (1.02-1.85); p=0.04], vascular disease [HR1.53 (1.10-2.14); p=0.01] and
poor TTR (TTR<70%) [HR 1.47 (1.13-1.91); p=0.004] were associated with an increased risk
of the composite outcomes (≥1MACE) (Table 3.19), however, age ≥80 years was not (see
Appendix 4, Table A4.10 for full model).
219
prevention in AF overall and in patients ≥80 and <80 years
Outcomes, N Age ≥80, Event Age <80, Event P value for

(%) N=205 rate/100 N=786 rate/100 pt- proportions
pt-yrs yrs
≥1 MACE 64 (31.2) 8.4 265 (33.7) 7.4 0.55
Stroke/TIA/SE 12 (5.9) 1.4 38 (4.8) 0.9 0.68
Bleeding* 21 (10.2) 2.4 57 (7.3) 1.3 0.16
Cardiovascular
38 (18.5) 4.7 188 (23.9) 5.0 0.12
hospitalisation‡
Death 8 (3.9) 0.9 15 (1.9) 0.3 0.15
* Bleeding is combination of major bleed according to International Society on Thrombosis and Haemostasis (ISTH)
and clinically relevant non-major bleed (CRNMB).
‡
Cardiovascular hospitalisation: a hospitalisation with a cardiovascular cause: i) heart failure, myocardial infarction,
new angina, non-fatal cardiac arrest, ventricular arrhythmia, uncontrolled atrial fibrillation/atrial flutter, supraventricular
arrhythmia, ii) valve surgery, coronary artery bypass graft surgery (CABG), percutaneous transluminal coronary
angioplasty (PTCA) surgery, pacemaker/ICD insertion, carotid endarterectomy, peripheral angioplasty/surgery, limb
amputation AND as recorded in patient’s medical documents; DVT – Deep Vein Thrombosis; Major Bleeding – ISTH
major Bleeding: fatal bleeding and/or symptomatic bleeding in a critical area or organ, such as intracranial, intraspinal,
intraocular, retroperitoneal, intraarticular or pericardial, or intramuscular with compartment syndrome and/or bleeding
causing a fall in haemoglobin level of 2 g/dL (1.24 mmol/L) or more, or leading to transfusion of two or more units of
whole blood or red cells; Clinically relevant non-major bleeding (CRNMB): clinically overt bleeding not satisfying the
criteria for major bleeding and that led to hospitalisation, physician medical or surgical treatment, or a change in
antithrombotic therapy; PE – pulmonary embolism; SE: systemic embolism; TIA: transient ischemic attack; VTE: venous
thromboembolism.
220
Table 3.19: Cox proportional hazard regression analysis for the impact of age (≥80 years)
on all bleeding events, including major bleeding and clinically relevant non-major
bleeding and ≥1 MACE
Multivariate HR (95% CI) Bleeding events ≥1 MACE

(95% CI) (95% CI)
Age ≥80 years 1.90 (1.01-3.56) 1.00 (0.72-1.39)
Stroke/TIA - 1.37 (1.02-1.85)
Vascular disease - 1.53 (1.10-2.14)
TTR <70% - 1.47 (1.13-1.91)
CI: confidence interval; HR: hazard ratio; TIA: transient ischemic attack; TTR: time in therapeutic range
p=0.009
p=0.009
Age <80 years

Age <80 years
Age ≥80 years
Age ≥80 years
Figure 3.11: Kaplan-Meier curve of bleeding events among patients age ≥80 and <80 years
221
3.4.4.3 Adverse outcomes by kidney disease
After a median (IQR) of 5.2 (3.2-2.7) years of follow up, 326 patients (33.5%) experienced ≥1
MACE. There is no statistically significant difference in TE, bleeding, CV hospitalisation, death
and ≥1 MACE with eGFR ≥90, eGFR 60-89 and eGFR ≤59 ml/min/1.73m2 respectively (Table
3.20).
prevention in AF overall and by different categories of kidney disease
All eGFR≥90 eGFR 60-89 eGFR ≤59 p-value

Outcomes (N=974) ml/min ml/min ml/min
N=133 N=491 N=350
≥1 MACE 326 (33.5) 52 (39.1) 153 (31.2) 121 (34.6) 0.20
Stroke/TIA/SE 50 (5.1) 7 (5.3) 21 (4.3) 22 (6.3) 0.43
Bleeding* 76 (7.8) 11 (8.3) 39 (7.9) 26 (7.4) 0.94
Cardiovascular
224 (23.0) 37 (27.8) 109 (22.2) 78 (22.3) 0.36
hospitalisation
Death 22 (2.3) 3 (2.3) 9 (1.8) 10 (2.9) 0.62
eGFR ≥90ml/min/1.73m2- normal kidney function; eGFR 60-89 ml/min/1.73m2- mild kidney disease; eGFR ≤59 ml/min/1.73m2- mild-
moderate-severe and kidney failure
Cardiovascular hospitalisation: a hospitalisation with a cardiovascular cause: i) heart failure, MI, new angina, non-fatal cardiac arrest,
ventricular arrhythmia, uncontrolled AF/atrial flutter, supraventricular arrhythmia, ii) valve surgery, CABG surgery, PTCA surgery,
pacemaker/ICD insertion, carotid endarterectomy, peripheral angioplasty/surgery, limb amputation AND as recorded in patient’s
medical documents; Major Bleeding – ISTH Major Bleeding: fatal bleeding and/or symptomatic bleeding in a critical area or organ,
such as intracranial, intraspinal, intraocular, retroperitoneal, intraarticular or pericardial, or intramuscular with compartment syndrome
and/or bleeding causing a fall in haemoglobin level of 2 g/dL (1.24 mmol/L) or more, or leading to transfusion of two or more units of
whole blood or red cells; Clinically relevant non-major bleeding (CRNMB): clinically overt bleeding not satisfying the criteria for major
bleeding but meet at least one of the 3 criteria: i) leading to hospitalisation or increased level of care, ii) requiring medical intervention
by healthcare professional and iii) prompting face to face evaluation.; SE: systemic embolism; TIA: transient ischemic attack; VTE:
venous thromboembolism.
* Bleeding ISTH is combination of major bleed ISTH and clinically relevant non-major bleed (CRNMB).
222
3.5 Discussion
This study found that anticoagulation control, evidenced by TTR and PINRR, was significantly
lower in South-Asian and Afro-Caribbean patients compared to White patients, despite similar
INR monitoring intensity. In contrast, no significant differences in TTR was evident when the
cohort was grouped by age (≥80 vs. <80 years) or kidney disease at baseline. Further, non-
white ethnicity was the strongest independent predictor of poor TTR after adjustment for
demographic and clinical variables.
3.5.1 Anticoagulation control in the overall cohort
In this study, the overall mean TTR by both methods (Rosendaal and PINRR) and across all
ethnic groups was below 70%, which reflects sub-optimal anticoagulation control in this
population, according to European guidelines (3). When the PINRR method was used, the
overall percentage of time in the therapeutic range was 55.2%, indicating even poorer
anticoagulation control. In addition, one third of the patients had at least one INR value above
5, with 4.1% reporting an INR> 8.0, indicative of deranged INR control, associated with higher
risk of bleeding (276, 426, 427). These results are in line with the warfarin arms of the recent
four landmark NOAC trials [RELY(132), ROCKET-AF (133), ARISTOTLE (134) and ENGAGE-
AF (135); mean TTR 64, 58, 66 and 68 respectively] which also showed sub-optimal
anticoagulation control despite strict patient inclusion and follow ups. In clinical practice,
obtaining a good TTR can be difficult as anticoagulation control is affected by many factors
(see Section 1.5.1 pages 78-100 for more details). The SAMe-TT2R2 score (199) could be
used to identify common clinical factors that might predict good/poor anticoagulation and aid
physicians to choose the best anticoagulation treatment in this setting.
223
3.5.1.1 Anticoagulation control in different ethnic groups
Despite similar INR monitoring intensity, it is evident that patients from ethnic minority groups
obtained significantly lower quality of anticoagulation control than white patients.
Previous studies by Yong (209) and Golwala et al (210) in the United States have also reported
similar findings where Black patients had poorer anticoagulation control compared to Whites,
Hispanics, Native Americans and other ethnic minority groups however, no studies to date are
available examining anticoagulation control among South-Asian patients. Rao and colleagues
(198) investigated patient-level and site-level factors that might influence the differences in
anticoagulation control among 9572 Black and 88,481 White patients in the United States.
Their findings suggest that greater proportion of the differences may be attributed to non-
modifiable factors such as young age, region and poverty level, distance to anticoagulation
clinic and presence of co-morbid conditions (198). Although it is not certain, it could be
possible that some of these factors (demographics and clinical) might also contribute to ethnic
disparities in the quality of anticoagulation control in the current cohort. Younger patients were
more commonly seen among South-Asian patients and could perhaps reflect hectic lifestyle
resulting in difficulties attending INR appointments. In addition, the presence of vascular
disease, diabetes and anaemia were also more commonly seen among patients in the non-
white ethnic groups suggesting greater burden of illness among them (see section 1.5.1.1.4
for more explanation).
Secondly, although pharmacogenetic factors were not investigated in this study, warfarin
metabolism and dose response might differ between ethnic groups. Studies have shown that
warfarin dosage requirement is lower in Asians and Whites but higher in Blacks partly due to
racial differences in genotype frequencies (428). Patients who inherit one or two copies of
CYP2C9*2 or CYP2C9*3 allele are more sensitive to warfarin compared to normal
224
metabolisers since they impair S-warfarin metabolism by 30-40% and ~80-90% respectively.
They are at a higher risk of bleeding during warfarin initiation thus requiring lower doses of
warfarin (429-432). The frequencies of the CYP2C9 alleles differ between ethnic groups (433-
435). More Caucasians (10-20%) were found to have the *2 allele compared to Asian (1-3%)
or African (0-6%) populations (216) and less *3 allele were seen among the African-Americans
(209, 216). Moreover, African Americans were also found to have additional CYP2C9 alleles
(CYP2C9*5, *6, *8, and *11; the less common ones) which are also associated with reduced
function of the CYP2C9 enzyme and contribute to dose variability (436). Thus warfarin dosage
can be tailor- made accordingly, if needed, as per guidelines (216, 436).
The role of pharmacogenetics testing and dosing of warfarin patients remains a controversial
topic. Two randomized trials examining the impact of genotype dosing towards improving TTR
have been contradictory. The EU-PACT trial (437) (N=455) which included participants from
European (98.6%), African (0.9%) ancestries and others (0.5%) showed that TTR was
significantly higher in the genotype-guided therapy group compared to standard care at 12
weeks (TTR 67.4% vs. 60.3%; p<0.001) (437). In contrast, the COAG trial (438) (N=1015)
which included more African Americans (27%), 67% European American and 6% Hispanic
showed no difference in the mean TTR in patients with genotype-guided dosing compared to
clinical-guided dosing of warfarin (TTR 45.2% vs. 45.5% respectively) at 4 weeks. However,
TTR improvement by genotype-guided dosing can only be seen among European Americans
(2.8%; p=0.15) but not among African Americans. Instead worsening TTR was seen among
the latter (-8.3%; p=0.01). The differences in the results of these two trials could be attributed
to heterogeneity in racial diversity, CY2C9*2 and *3, VKORC1 frequency and warfarin
indication among the participants. It was also demonstrated (439) that warfarin dosing
variability was affected by both clinical and genetic factors in another study (N=1357)
comprising of more African Americans (43%). Nevertheless, in African Americans, clinical
factors (such as age, CKD, body surface area, and amiodarone) account for higher proportion
225
of dose variability than genetic factors. Hence, a pharmacogenetic algorithm for warfarin
dosing that is race-specific; rather than race-adjusted was suggested.
Third, although this was not investigated in this cohort, previous studies have shown that many
AF patients have little knowledge of their condition and lack understanding of the risk and
benefit of anticoagulant therapy, in particular among ethnic minority patients (383, 384). This
may also contribute to poor adherence to medication and may (or may not) result in poor
anticoagulation control. Interventions such as more frequent follow up visits and reviews,
educational interventions and counselling are needed (219). For example, discussions about
specific food types and health supplements that may interact with warfarin (herbal remedies,
vegetarianism, or inconsistent amount of food rich with vitamin K, for example, green leafy
vegetables) or specific cultural differences (for example the impact of fasting or excess alcohol
on the quality of anticoagulation control) that may affect their understanding are required
especially to the ethnic minority groups to ensure that their TTR can be improved in order to
achieve the best outcomes and prevent treatment complications (316). Materials for the
education intervention should be available in different languages and in different media
(booklets, video-clips etc.) so patients who do not speak English but is able to read in their
own language can also benefit. Interpreters may also be required so that information shared
between the healthcare professional and patient during the counselling session can be
effectively understood. Patients should also be encouraged to be actively involved during
sessions and to raise any concerns regarding warfarin treatment so that any barriers to
adherence can be discussed and overcome. Recently, the TREAT intervention (a one-off
educational behavioral session) delivered by a health psychologist demonstrated a significant
improvement of TTR compared to patients receiving usual care alone (TTR 76.2% vs. 73.1%,
p=0.035) (219). The design of this behavior-change intervention package (consist of DVD
delivered by ‘expert patient’ narratives and consultant cardiologist, educational booklet, diary
and patient worksheet) was based on theoretical models (Common Sense Model and
Necessity-Concerns Framework), clinical guidelines, relevant literature and AF patient
226
feedback. This intervention not only resulted in improvement of the quality of anticoagulation
control and a greater understanding of AF and its treatment; it also changed patients’ belief
surrounding their treatment necessity and potential harm (293).
3.5.1.2 Anticoagulation control in in elderly patients
The quality of anticoagulation control was similar among those aged ≥80 and <80 years
despite fewer INR visits and shorter follow up among the very elderly patients. Moreover, less
than half (44%) of the elderly patients had optimal TTR (TTR≥70%).
These results are consistent with the data obtained from the BAFTA (250) and WASPO (251)
trials, with mean TTR comparable to the current elderly cohort (mean TTR 67% and 69%
respectively vs. 67% in current elderly cohort). Two other Italian studies (247, 252) reported
slightly higher TTR in their cohort (mean TTR 71% in both studies vs. 67% in current elderly
cohort) while TTR was lower in another study by Hylek et al (241) among the elderly (≥80
years; mean TTR 58%). This may be explained by the inclusion of an inception cohort by
Hylek et al (241), whereas the current study included patients throughout the entire period of
treatment (median duration of VKA treatment 5.2 years reflecting long term VKA
management). Low TTR (mean TTR 48%) has also been reported in another inception cohort
study (200) suggesting the difficulties in achieving good control with VKA therapy especially
during the inception period (200).
227
3.5.1.3 Anticoagulation control in different categories of kidney disease
There was no significant difference in TTR when calculated using the Rosendaal method
among AF patients with different categories of kidney disease at baseline. However, when the
PINRR method was utilised, a higher percentage of INRs within range was seen among AF
patients in with eGFR 60-89 ml/min/1.73m2 and eGFR ≤59 ml/min/1.73m2 compared to
patients with normal kidney function (eGFR ≥90 ml/min/1.73m2).
Despite the challenges faced in managing AF patients with CKD, many studies have shown
the benefits of VKA therapy in AF and CKD patients in reducing TE, bleeding, cardiovascular
hospitalisation and all-cause death (263-265, 269, 440). As mentioned in the literature review,
eight (263-270) studies reported TTR data and seven (263-265, 267-270) have shown a
significant trend of worsening TTR as the kidney disease worsened (see section 1.5.1.1.5,
pages 81-82 for more information). For example, in the SPORTIF III and IV trials, the mean
TTR among AF patients with eGFR < 60 ml/min was significantly lower than those with eGFR
≥60 ml/min (mean TTR 66.6% vs. 69.6%; p<0.001 respectively) (264). In a retrospective
analysis of the Swedish health registers of 307,351 AF patients comprising 13,435 patients
with renal disease (diagnosis based on ICD-10 codes), the mean TTR for those with renal
disease was also significantly lower than those without renal disease (mean TTR 66.7% vs.
74.6%; p<0.001 respectively) (265).
In contrast, this study did not show a similar trend, instead a non-significant higher TTR was
observed in patients with eGFR 60-89 ml/min/1.73m2 and eGFR ≤59 ml/min/1.73m2 compared
to patients with eGFR ≥90 ml/min/1.73m2 and this trend was statistically significant when the
PINRR method was utilised. Also, sub-therapeutic INRs (INR <2.0) and the presence of at
least one INR >8.0 was significantly more prevalent among AF patients with normal kidney
function. Similarly, when investigating the influence of kidney disease towards anticoagulation
control in linear and logistic regression analyses, the presence of kidney disease (eGFR ≤59
228
ml/min/1.73m2) at baseline did not influence sub-optimal anticoagulation control during follow
up. This finding is also similar to a study among 724 non-dialysis dependent chronic kidney
disease on VKA therapy (266). TTR was significantly higher (75.1%) in patients with moderate
kidney disease (eGFR 30-60 ml/min) compared to patients without CKD (eGFR>60 ml/min:
TTR 67.0%; p<0.01). There was also a non-significant trend towards higher TTR in the severe
CKD group (eGFR <30 ml/min: TTR 70.3%; p=0.41) compared to those without kidney
disease. Similar to the current cohort, renal function was assessed at the start of VKA therapy
and TTR was calculated throughout the entire treatment period, although the categorisations
of kidney disease were different compared to the current cohort. Despite that, one common
finding seen was that the quality of VKA therapy seemed to be better in patients with moderate
kidney disease than with normal kidney function.
There are several potential explanations for these contradictory findings. First of all, perhaps
in this cohort, AF patients with concomitant kidney disease are well managed by a dedicated
anticoagulant service thus resulting in similar or better anticoagulation control compared to
patients without kidney disease. It could also be that the presence of concomitant ‘kidney
disease’ serves as a ‘flag’ to the anticoagulant services so that extra care and attention is
given throughout the entire monitoring period. This is because patients with concomitant
kidney disease are considered to be a ‘vulnerable’ group of patients and are at higher risk of
adverse events including thromboembolism and bleeding. Indeed in the SPORTIF III and IV
trials (264), the presence of CKD was significantly associated with an increased risk of stroke
and high TTR (TTR ≥70%) was significantly associated with a reduction in the risk of stroke
[HR 0.63 (95% CI 0.41-0.98)], major bleeding [HR 0.58 (0.42-0.80)], and mortality [HR 0.63
(0.47-0.84)]. Likewise, a study (266) in The Netherlands of 724 AF patients with VKA therapy
and CKD also showed that patients with eGFR <30ml/min were at increased risk of major
bleeding and stroke/TIA compared to those with eGFR 30-60 ml/min [HR 1.86 (95% CI 1.08-
3.21) and HR 3.93 (95% CI 1.71-9.00) respectively] and this was mediated when the
anticoagulation control was sub-optimal. Lastly, better anticoagulation control among patients
229
with eGFR 60-89 and eGFR ≤59 may be driven by the significantly higher proportion of white
patients with eGFR 60-89 and eGFR ≤59. Results from the main analysis have shown that a
greater proportion of optimised TTR was seen among white patients compared to non-white
patients. Furthermore, logistic regression analysis has shown that non-white ethnicity is an
independent predictor of poor TTR. So, this might indirectly influence the TTR results among
the CKD patients.
3.5.2 Predictors of anticoagulation control in the whole cohort
This study demonstrates that slightly more than half of the cohort experienced difficulties
achieving optimal quality of anticoagulation and therefore further investigation on the
predictors of poor anticoagulation control was conducted. Interestingly, non-white ethnicity
emerged as the strongest predictor (in both univariate and multivariate linear and logistic
regression analyses) of poor quality of anticoagulation control when TTR was calculated via
both the Rosendaal and the PINRR methods. Possible explanations towards this finding have
been described in section 3.5.1.2.
Other significant patient factors related to anticoagulation control evident from this cohort is
smoking history, although information on smoking status is only available for 72.4% of
patients. It is an independent predictor of poor TTR and PINRR on linear regression analysis
and was more prevalent among the Whites compared to South-Asian and Afro-Caribbean
patients. This is consistent with three other studies demonstrating smoking as a predictor of
poor TTR (199, 205, 208). The relationship of how smoking can influence anticoagulation
control is unclear but may reflect less interest in maintaining good health that may translate
into poorer adherence to OACs, thus resulting in poor TTR (199).
Clinical factors such as comorbid diseases were shown to have an impact in the quality of
anticoagulation control. In linear regression analyses, heart failure and bleeding history
negatively predicted both TTR and PINRR, with vascular disease also predicting PINRR.
230
Meanwhile, in logistic regression analyses, anaemia was an independent predictor of poor
anticoagulation control for both TTR<70% and PINRR<70% followed by vascular disease for
TTR<70%.
Other studies have also shown an association of poor anticoagulation control with a variety of
comorbid conditions such as heart failure (197-202, 253-255), diabetes (197-200, 202, 205,
256), kidney disease (198, 199, 201, 213, 257), liver disease (198, 199, 254, 257), lung
disease (199, 201, 204, 205), coronary artery disease (199, 201), peripheral vascular disease
(199, 201), stroke (199, 204) and previous bleeding (213). The exact mechanism of this
relationship is unclear but perhaps this reflects greater illness burden and complexity including
more medications prescribed for each of the conditions thus increasing the potential of drug
interaction with warfarin and nonadherence leading to poorer anticoagulation control (201).
3.5.3 Adverse clinical outcomes in the whole cohort
At least 30% of the patients experienced ≥1 MACE and there was no significant difference in
terms of the rate of thromboembolic, bleeding events and mortality across the three ethnic
groups except for CV hospitalizations, where it was highest among the South-Asians. In this
cohort of patients, Afro-Caribbeans had the highest risk of stroke. Although underpowered,
there was no significant difference in terms of TE events across different ethnic groups
although proportionally, TE and bleeding events were highest among Afro-Caribbeans.
Previous epidemiological studies have shown that black and Hispanic individuals have a two-
fold higher annual risk of stroke compared to Whites (441, 442). However, this was not evident
in the present study, a finding consistent with the ORBIT-AF registry, that showed no
difference in stroke or all-cause mortality among white, black and Hispanic participants where
anticoagulation use was high (210). In a subgroup analysis of the AFFIRM trial, there was no
231
difference in overall survival at 5 year follow up among White, Black and Hispanic participants
(443).
Independent predictors of the composite outcome were prior stoke/TIA, vascular disease and
poor TTR. Other studies have shown that poor TTR is related to thromboembolic and bleeding
events (119, 444) and in this cohort poor TTR (<70%) independently predicted composites of
TE, bleeding events, CV hospitalizations and all-cause mortality.
3.5.3.1 Adverse clinical outcomes in elderly
Exploratory analyses of the elderly showed no significant differences in the composite
endpoints (≥1 MACE) between the elderly (age ≥80 years) and those aged <80 years.
However, age ≥80 years was significantly associated with higher bleeding risk even after
adjustment for demographics and clinical variables.
Previous studies have reported conflicting results regarding the increased risk of bleeding
among elderly patients on OAC therapy. The absolute rate of major bleeding was 2.5 vs. 0.9
per patient years among ≥80 vs. <80 year old AF patients, respectively receiving warfarin
therapy in one Italian study (247). Conversely, a 5-fold increase in incidence rate of bleeding
was reported in those aged ≥80 years compared to <80 years (13.1 vs. 4.8 per 100 patients
years respectively) in another study (241). Age ≥80 years was associated with increased risk
of bleeding events in both studies (241, 247). The difference in bleeding rate between these
studies might be explained by the higher proportion of patients experiencing CAD (35% vs.
20%) who were prescribed concomitant aspirin therapy (40% vs. 3.5%) in the latter (241)
compared to the former (247) respectively; both of which are factors known to increase risk of
bleeding.
In the current cohort, when investigating the factors associated with bleeding events during
the entire period of warfarin exposure, age ≥80 years was the only significant factor associated
with bleeding events, similar to previous studies (241, 247). Indeed, very close attention needs
232
to be paid to the very elderly patients who are on OAC therapy to prevent bleeding
complications. Various bleeding scores are available to assess bleeding risk in AF patients (3,
176). These scores can be used to guide physicians to ‘flag up’ factors that may predispose
patients to bleeding events. Any modifiable risk factors for bleeding, such as uncontrolled
hypertension in the ‘H’ component of the HAS-BLED score (180) should be addressed by
controlling patient’s blood pressure. The risk of bleeding is not static thus needs to be
evaluated periodically (3, 242).
3.5.3.2 Adverse clinical outcomes in different categories of kidney disease
There were no significant differences in thromboembolic, bleeding, CV hospitalisations, all-
cause mortality and ≥1 MACE according to kidney disease though this analysis was purely
exploratory. This result must be interpreted with caution as this investigation was not powered
to detect any significant difference in any of the adverse clinical outcomes.
Nevertheless, other studies have shown increased risk of thromboembolism, bleeding and
mortality in patients with concomitant AF and CKD (264, 445-448). One Swedish AF cohort
study from the health registers (265) reported a higher annual rate of stroke (3.9% vs. 2.9%
respectively), any bleeding (9.8% vs. 4.1% respectively) and mortality (36.0% vs. 11.5%
respectively) in patients with renal failure (definition obtained from the ICD-10 codes N17-19
or by local Swedish procedure codes for haemodialysis, peritoneal dialysis or renal
transplantation) compared to those without renal failure (265). In this Swedish cohort, renal
failure independently predicted intracranial bleeding [adjusted HR 1.27 (95% CI 1.09-1.49)].
However, despite the high risk of bleeding, the use of warfarin compared to no warfarin therapy
was beneficial in renal failure patients in the composite endpoint of ischaemic stroke,
intracranial bleed or death [adjusted HR 0.76 (95% CI 0.72-0.80)] (265). Similarly, in the
SPAF-III trial, (440) use of warfarin reduced the risk of ischemic stroke and systemic embolism
by 76% (95% CI 42%-90%; p<0.001) among high risk AF patients with CKD stage 3 (eGFR
30-59 ml/min) compared to the combination of low dose warfarin and aspirin (440).
233
3.5.4 Strengths and limitations
This is the first study to assess anticoagulation control and adverse clinical outcomes in
different ethnic groups in the UK. Other studies looking at the differences in anticoagulation
control between ethnic minority groups were conducted in the United States comparing
Whites, African-American, Hispanic and Native Americans (198, 210). In addition, two further
ancillary analyses were undertaken among elderly and patients with different categories of
kidney disease to give insights into the quality of warfarin control in these two sub-groups,
managed by one anticoagulation clinic in this Trust.
Furthermore, two methods of calculating TTR were utilised, with both methods correlating with
each other and demonstrating similar results. Researchers recommend that ≥2 VKA control
measures are reported per study as the quality of anticoagulation control can vary depending
on the method of TTR reported (347, 449, 450). Further, TTR was also calculated using a
large number of INR results [mean (SD) 58.7 (25.5)] for a median of 5.2 (3.2-7.0) years of
follow up reflecting the long-term quality of anticoagulation control in this centre.
The cohort comprised 991 patients but this was only 43 % of the available cohort and thus it
is possible that the results are not representative of the whole (N=2478) cohort. Nonetheless,
the proportion of patients from each ethnic group included was representative of the total
cohort in this Trust. Furthermore, Afro-Caribbean and South-Asian patients constituted about
10% each from the whole population included in the study; similar to the ethnic composition
of studies of anticoagulation control by Yong et al and Golwala et al (8.3% and 5% Blacks
respectively) (198, 210). However, in general, AF is more prevalent among Whites (8.0%)
than Asians (3.9%) and Blacks (3.8%)(35).
In addition, the retrospective review from medical records means that some information was
not readily available, including the patient’s ethnic group, medical history and medication
234
history, and thus a small number (3.2%) of patients had to be excluded. Also, recording of
adverse clinical events were based on the events occurring in this Trust, thus any events
occurring outside this Trust were not captured thus might lead to underestimation of the events
of interest.
Finally, this study assessed anticoagulant control by looking at objective measures available
on clinical databases. Other variables that could influence anticoagulation control such as
distance to anticoagulation clinic, education level, history of employment, quality of life of AF
patients (414), magnitude of drug and food interaction, CYP2C9 or VKORC1 genotype (329)
were not taken into account in the current analyses.
3.5.5 Clinical implications
The most clinically relevant finding of this study is that achieving optimal quality of
anticoagulation control with warfarin is more challenging among non-whites but that being very
elderly or having impaired renal function does not independently predict TTR in a setting with
a well-managed anticoagulant clinic. This work highlights the importance of good TTR and
that there is considerable room for improvement given that less than half the cohort achieved
optimal TTR, especially Afro-Caribbean and South-Asian patients. This suggests the need to
focus on individual reasons for poor INR control and to develop strategies to improve
anticoagulation control where required. For example, more frequent follow up via phone call
or face-to-face could be arranged so that more attention can be given to these patients. During
the follow ups, factors that can influence anticoagulation control and the importance of being
adherent to the anticoagulant should be emphasized. In addition, patients should also be
encouraged to inform healthcare professionals if they encounter any problems with VKA
therapy so that appropriate remedial action can be taken, and the patient should be reviewed.
235
3.5.6 Future research
Further prospective, multicenter, observational studies with larger sample sizes (>1000
patients) especially within ethnic minority groups are required to confirm these findings.
Perhaps this study can be extended into a multi-national registry including other ethnically
diverse countries for example in South-East Asian countries like Malaysia, Thailand, Vietnam,
Myanmar, Cambodia and Laos which still uses VKA as the OAC of choice for stroke prevention
in AF and in other thrombotic diseases (VTE/PE). It would be of interest to investigate TTR,
determinants of TTR (including ethnicity) and its impact on predicting TE and bleeding events
in Asian countries. Regional comparison of TTR and adverse clinical outcomes could be
undertaken within each country where data in this area is lacking.
Asian populations (451, 452) still rely heavily on herbal preparations for treating medical
ailments. For example, ‘tongkat ali’ (Eurycoma longifolia), originating from Malaysia and
Indonesia is used as an alternative for testosterone replacement therapy or treatment of
impotence (451). Meanwhile, Withania somnifera, an Indian ginseng was shown to have
potential for cancer-related fatigue and improvement of quality of life in a non-randomised
comparative trial of 100 patients (N= 50 chemotherapy + WH vs. N=50 chemotherapy alone)
with breast cancer (451, 453). Hence, it would be interesting to investigate the impact of herbal
products on INR control as VKA-herb interaction have major safety concerns and more data
is required in this field.
3.6 Conclusions
Ethnic disparities in the quality of anticoagulation control are evident but not among the very
elderly and patients with different categories of kidney disease. South-Asians and Afro-
Caribbeans had poorer INR control compared to Whites, despite similar intensity INR-
monitoring. After adjustment, non-white ethnicity and anaemia remained the strongest
independent predictor of poor TTR and PINRR. Meanwhile, CV hospitalisations were more
236
prevalent among the South-Asians. Closer attention needs to be given to patients from non-
white ethnic groups to understand the reasons of poor anticoagulation control so that effective
strategies can be developed and implemented by healthcare providers to improve outcomes.
237
Chapter 4. Anticoagulation control in operated valvular heart
disease patients with and without atrial fibrillation receiving vitamin
K antagonist
4.1 Abstract
Introduction: Good quality anticoagulation control among patients with operated valvular
heart disease (VHD) is needed to reduce ischaemic complications. There is limited evidence
on factors affecting anticoagulation control among this patient population.
Objective: To investigate the quality of VKA control (TTR), predictors of anticoagulation
control and the prevalence of adverse clinical outcomes [thromboembolic (stroke/TIA and
systemic embolism), bleeding events, cardiovascular hospitalisation and all-cause mortality
and ≥1 composite endpoints (MACE)] in operated VHD patients at one acute Trust in the West
Midlands, United Kingdom. Exploratory analyses investigated the relationship between INR
control and adverse clinical outcomes.
Methods: Retrospective data collection from the electronic medical record database were
undertaken to collect all demographics and clinical information. The Rosendaal and
percentage INRs in range (PINRR) methods were used to calculate TTR among 456 operated
VHD patients of whom 164 (36%) with AF and 292 (64%) without AF. Patient’s demographics,
comorbidities and other clinical data were used as predictors of TTR and were examined by
logistic regression analysis. Chi-squared tests were utilised to explore the relationship
between INR control and adverse clinical outcomes.
Results: The mean (SD) age was 51 (14.7), 64.5% were male, 96.1% had a mechanical
prosthesis and 64% had aortic valve replacement. Operated VHD patients with AF had lower
mean TTR and PINRR [mean (SD) TTR 55.7% (14.2) vs. 60.1% (14.6); p=0.002 respectively;
mean PINRR 47.4% (13.5) vs. 51.6% (13.7); p=0.002 respectively], lower proportions with
238
optimal anticoagulation control (TTR ≥70%) (14.0% vs. 25.7%; p=0.004) and higher
proportions with sub-therapeutic INRs (28.4% vs. 23.4%; p<0.001) despite a similar number
of INR tests compared to operated VHD patients without AF. Independent factors predicting
poor TTR after adjustment for demographic and clinical variables were: female, the presence
of AF at baseline, anaemia/bleeding history and HAS-BLED score. Significantly higher
proportions of patients with operated VHD and AF died [all-cause mortality (20.7% vs. 5.8%;
p<0.001)]. Similarly, more deaths (13.1% vs. 4.1%; p=0.011) and ≥1 MACE (42.7% vs. 27.6%;
p=0.006) were seen in patients with TTR <70% compared to TTR≥70% respectively.
Conclusion: Operated VHD patients with AF at baseline have poorer anticoagulation control
compared to those without AF at baseline. The presence of concomitant AF,
anaemia/bleeding history, as well as female gender, independently predicted poor TTR and
the rate of all-cause mortality was significantly higher among operated VHD patients with AF.
These findings suggest closer INR monitoring among operated VHD patients especially those
with AF to improve anticoagulation control and prevent adverse clinical outcomes.
239
4.2 Introduction
VKAs are the only anticoagulant of choice in patients undergoing heart valve replacement,
especially with mechanical prosthesis (358). The target INR for AF is 2.0-3.0, (349, 351, 358)
whereas the INR targets for patients with VHD post-surgery varies depending on factors such
as patient risk factors, (example: mitral/tricuspid valve replacement, previous TE, AF, mitral
stenosis and LVEF <35%) type of valve, and the thrombogenicity of the prosthesis (349, 351,
358). The 2017 ESC guidelines on the management of valvular heart disease (358)
recommend a median INR value be maintained in place of a range to prevent extreme values
within the target range. They also recommend a higher median INR value for patients with ≥1
risk factor than those without any of these risk factors (358)(More details in section 1.7.2.1
page 116). The newer types of valve used more commonly now, such as Carbomedics, St
Jude or Medtronic have low valve thrombogenicity with limited data on the rate of valves
thrombogenicity as they are also influenced by patient related risk factor and study design
(454); however one review reported HR 1.06 (0.05-0.56) for valve thrombosis among St Jude
and Carbomedics valves (455). Patients with risk factors receiving newer types of valves are
recommended to achieve a median target INR of 3.0 compared to those without risk factors,
where the target INR is lower at 2.5 (349, 351, 358).
4.2.1 Study objectives

To date, only five (377-381) studies, conducted between 2002 and 2018, have investigated
anticoagulation control after valve replacement; two (380, 381) used anticoagulation variability
while the others (377-380) used TTR. Therefore, the objective of this study was to investigate
anticoagulation control measured using TTR (Rosendaal method) and the PINRR method
among operated VHD patients, comparing patients with and without AF. Second, to investigate
the predictors for poor anticoagulation control, and finally to investigate the prevalence of
adverse clinical outcomes including stroke/TIA, bleeding, CV hospitalisations, death and the
240
composites of ≥1 MACE. Exploratory analyses investigated the relationship between INR
control and adverse clinical outcomes.
4.3 Methods
4.3.1 Study design

This is a single centre, retrospective analysis of patients with VHD receiving VKA therapy after
valve replacement therapy at one acute Trust in the West Midlands, United Kingdom (SWBH
NHS Trust). Data was collected from 1st November 2017 to 31st March 2018. VHD patients
receiving VKA therapy were identified from the DAWN AC® anticoagulation management
software (described in section 3.3.1, pages 174-175).
This study was considered as service evaluation by the SWBH Research and Development
department and therefore did not require REC approval. However, local R&D approval was
obtained (see email confirmation from SWBH R&D Department, Appendix 5).
4.3.1.1 Patient selection
A list of patients with VHD receiving VKA therapy (N=604) was generated from the DAWN AC
management software. However, 148 patients were excluded due to: i) VHD but without
surgical intervention [N=38; mitral stenosis (N=22), aortic stenosis (N=2), mitral regurgitation
(N=2), mitral valve repair (N=4), valvuloplasty (N= 6), and valvulotomy (N= 2)]; ii) incomplete
INR results (N=3) and iii) incomplete medical information (N=107). The final cohort consists of
456 VHD patients who had surgical intervention of the affected valve(s) and were prescribed
VKA therapy post-surgery. They were stratified into those with and without AF (Figure 4.1).
4.3.1.2 Procedure
All baseline characteristics and clinical information including medical history, medication, and
laboratory tests were collected from the point that VKA was initiated after surgical replacement
241
of the valves (i.e., mechanical and tissue valve repair). Information on outcomes i.e., INR
results and adverse clinical outcomes occurring after this point were collected using a
proforma (see Appendix 4, Table A4.1).
242
VHD patients on VKA therapy selected from
SWBH DAWN AC® Database (N=604)
Patient’s baseline demographic characteristics, medical and

medication history recorded from CDA and information on INRs
from DAWN AC® databases
Patients excluded N=148:

-VHD without surgical intervention, N=38
• Mitral stenosis, N=22; aortic stenosis,
N=2; mitral regurgitation, N=2; mitral
valve repair, N=4; valvuloplasty, N=6;
valvulotomy; N=2
-Insufficient INR result, N=3
-Insufficient medical information, N=107
Final cohort, N=456

Operated VHD with AF (N=164) and without AF
(N=292)
Outcomes
Measures of Predictors of
anticoagulation anticoagulation Adverse clinical
control control outcomes
Time to therapeutic Thromboembolic and

Percentage of INR
range (TTR) by bleeding events, CV
within range (PINRR)
Rosendaal hospitalization, death
CDA: clinical data archive; CV: cardiovascular; INR: international normalised ratio; TTR: time in therapeutic range;
PINRR: percentage of INRs in range; SWBH: Sandwell and West Birmingham Hospitals; VHD: valvular heart
disease; VKA: vitamin K antagonist
243
Dependent/outcome variable
4.3.1.2.1 Time in therapeutic range, TTR
INR values for VKA therapy were collected from CDA and DAWN databases from SWBH NHS
Trust for patients with at least three INR values in a year starting from February 2009 until 31
January 2018. The year 2009 is the period where INR readings were consistently available in
the hospital databases. Prior to this, another system was utilised and is no longer currently
active to allow complete INR data collection resulting in wide gaps between the dates of each
INR reading. The quality of anticoagulation control was calculated using the Rosendaal and
the PINRR methods (421)(see section 2.3 for definition and description). TTR and PINRR
were calculated based on each patient’s individual target INR range determined by the
surgeon; thus, INR ranges differ between patients. TTR and PINRR were further dichotomised
into TTR ≥70% and <70% and PINRR ≥70% and <70%, with TTR and PINRR ≥70% reflecting
optimal anticoagulation control based on the ESC guideline (3). The proportions of sub-
therapeutic INRs (INRs below the target range), supra-therapeutic (INRs above target range)
and patients with at least one INR >5.0 or >8.0 were also collected. The follow-up period was
defined as the duration of VKA monitoring i.e., from the start date of INR collection until 31st
January 2018.
4.3.1.2.2 Definition of Atrial fibrillation
The diagnosis of AF at baseline was obtained directly from the CDA. This was defined as the
presence of AF as part of the concomitant diseases at the time of surgery or was diagnosed
after the surgery (post-operative). Types of AF including paroxysmal, persistent, long standing
and permanent were recorded. If this information was not available, an assumption was made
based on the length of time since AF diagnosis and the pattern of ECG recordings available,
with confirmation of AF from a medical doctor, according to ESC AF guidelines (3).
244
4.3.1.3 Predictors: Patient demographics and clinical factors
Patient’s age was calculated based on the date of their first valve surgery (ranging from 1972-
2017). Other demographic information such as gender, ethnicity, information regarding
smoking status and alcohol intake, and other comorbidities, medication history and laboratory
parameters were obtained as near to the date (or within one month) of VKA initiation after the
first valve surgery, from the CDA. Information on smoking status was available for 372 patients
(82%); data on alcohol intake was only available for 353 patients (77%). This information was
used to calculate the individual HAS-BLED and SAMe-TT2R2 scores. Assumptions were also
made for chronic kidney disease, liver disease and anaemia based on the laboratory results
(see page 181 for details). The calculation of the CHA2DS2-VASc, HAS-BLED and SAMe-
TT2R2 scores were made on the basis of baseline information (see section 3.3.3, page 177
for more details).
4.3.1.4 Adverse clinical outcome
Information on adverse clinical outcomes were collected from the CDA covering the same time
frame to that of the INR collection i.e., from the point/date where INR was consistently
available in the system until 31 January 2018. Adverse clinical outcomes of interest were
stroke/TIA/systemic embolism, bleeding (combination of major bleed and CRNMB), CV
hospitalisation, death and a composite (≥ 1) of these MACE events. Definitions of each
outcome are described in section 2.3.3. In this study, the cause of death was specified as CV
death when specific information was available. Where cause of death was unavailable, death
was classified as all-cause mortality.
245
After performing normality tests, by the histogram plot method and the Kolmogorov-Smirnov
test where a bell-shaped distribution in the former and p-values >0.05 in the latter were
indicative of normally distributed data, all normally distributed data were expressed as mean
(SD), and non-normally distributed data as median (IQR). Demographic and clinical
characteristics of patients with categorical data were compared with chi-square or Fisher’s
exact test when appropriate and are reported as counts and percentage. Independent t-tests
were used to compare the means of continuous data for normally distributed data; the Mann-
Whitney tests were used for data that was not normally distributed. Univariate and multivariate
logistic regression analyses were performed to investigate the predictors of poor TTR
(TTR<70%). The relationship between TTR and adverse clinical outcomes were investigated
(exploratory) using the chi-squared test and are reported as counts and percentage. A Log-
Rank test was performed for AF categories and TTR categories and Kaplan-Meier Curves
were used to report the differences in survival and ≥1 MACE between the subgroups. All
analyses were conducted using SPSS version 23.0 (406), with p-values <0.05 considered
statistically significant.
246
4.4 Results
4.4.1 Baseline characteristics

Among the 456 patients with operated VHD, only 164 (36.0%) had AF at baseline. The overall
mean (SD) age was 51 (14.7), the majority were male (64.5%), of white ethnicity (65.2%), with
a mechanical prosthesis (96.1%), and the most common operation was aortic valve
replacement (64%) (Table 4.1). Patients with operated VHD with AF were significantly older
[mean (SD) age 56.6 (13.3); p<0.001], more likely to be female (48.2%; p<0.001), to receive
a tissue prosthesis (8.5%; p<0.001), to have had the mitral valve (41.5%; p<0.001) or both
mitral and aortic (20.7%; p<0.001) valves replaced. In addition, patients with operated VHD
and AF were also more likely to have concomitant heart failure (21.3%; p<0.001), hypertension
(72.0%; p=0.007), and pulmonary disease (25.6%; p=0.014) and were likely to be prescribed
diuretics (70.1%; p<0.001), amiodarone (22.6%; p<0.001) and digoxin (36.0%; p<0.001), and
had higher mean (SD) CHA2DS2-VASc [2.6 (1.5); p<0.001] and HAS-BLED scores [1.8 (1.1);
p=0.014] compared to patients with operated VHD without AF (Table 4.1).
247
Table 4.1: Baseline characteristics of patients with operated valvular heart disease, with and without AF
Total, AF No AF p-value
N (%) N=456 N=164 N=292
Age at implantation Mean age (SD) 51.1 (14.7) 56.6 (13.3) 48.0 (15.0) <0.001
≤64 382 (83.8) 117 (71.3) 265 (90.8)
Age groups 65-74 59 (12.9) 35 (21.3) 24 (8.2) <0.001
≥75 15 (3.3) 12 (7.3) 3 (1.0)
Female 162 (35.5) 79 (48.2) 83 (28.4)
Sex <0.001
Male 294 (64.5) 85 (51.8) 209 (71.6)
White 296 (65.2) 114 (69.9) 182 (62.5)
Ethnic groups‡ (N=454) South-Asian 120 (26.4) 35 (29.2) 85 (29.2) 0.20
Afro-Caribbean 38 (8.4) 14 (8.6) 24 (8.2)
Alcohol >14unit/day 32 (9.1) 9 (6.8) 23 (10.4)
Alcohol intake 0.26
(N=353)
Smoking/ex-smoker 83 (22.3) 27 (19.1) 56 (24.2)
Smoking status 0.25
(N=372)
Mitral 110 (24.1) 68 (41.5) 42 (14.4) <0.001
Site(s) of prosthesis Aortic 292 (64.0) 62 (37.8) 230 (78.8) <0.001
Both mitral and aortic 54 (11.8) 34 (20.7) 20 (6.8) <0.001
Types of valve Mechanical valve 438 (96.1) 150 (91.5) 288 (98.6) <0.001
replacement Tissue valve 18 (3.9) 14 (8.5) 4 (1.4) <0.001
Heart failure 53 (11.6) 35 (21.3) 18 (6.2) <0.001
Past medical history Hypertension 291 (63.8) 118 (72.0) 173 (59.2) 0.007
Diabetes 71 (15.6) 32 (19.5) 39 (13.4) 0.08
248
Table 4.1 continued
N=456 N=164 N=292
Stroke/TIA 66 (14.5) 30 (18.3) 36 (12.3) 0.08
Past medical history Vascular disease* 118 (25.9) 35 (21.3) 83 (28.4) 0.10
Lung disease# 89 (19.5) 42 (25.6) 47 (16.1) 0.014
Kidney disease† 17 (3.7) 7 (4.3) 10 (3.4) 0.65
Anaemia/previous 189 (41.4) 70 (42.7) 119 (40.8) 0.69
bleeding
Beta-blocker 177 (38.8) 62 (37.8) 115 (39.4) 0.74
ACEI/ARB 247 (54.2) 94 (57.3) 153 (52.4) 0.31
Diuretics 233 (51.1) 115 (70.1) 118 (40.4) <0.001
Current medications Amiodarone 50 (11.0) 37 (22.6) 13 (4.5) <0.001
Concurrent antiplatelet 79 (17.3) 21 (12.8) 58 (19.9) 0.06
Digoxin 69 (15.1) 59 (36.0) 10 (3.4) <0.001
Calcium channel blocker 54 (11.8) 20 (12.2) 34 (11.6) 0.86
CHA2DS2-VASc score Mean (SD) 2.0 (1.4) 2.6 (1.5) 1.7 (1.3) <0.001
CHA2DS2-VASc score Low risk 102 (22.4) 21 (12.8) 81 (27.7)
categories Intermediate 134 (29.4) 45 (27.4) 89 (30.5) <0.001
High risk 220 (48.2) 98 (59.8) 122 (41.8)
HAS-BLED score Mean 1.6 (1.2) 1.8 (1.1) 1.5 (1.2) 0.014
HAS-BLED score Low risk (0-2) 359 (78.7) 127 (77.4) 232 (79.5)
0.61
categories High risk (≥3) 97 (21.3) 37 (22.6) 60 (20.5)
249
Table 4.1 continued
N=456 N=164 N=292
SAMe-TT2R2 score Mean 2.7 (1.4) 2.7 (1.4) 2.7 (1.4) 0.53
SAMe-TT2R2 score 0-2 200 (43.9) 71 (43.3) 129 (44.2)
0.86
categories >2 256 (56.1) 93 (56.7) 163 (55.8)
ACEI/ARB: angiotensin converting enzyme inhibitor/ angiotensin receptor blockade; AF: atrial fibrillation; CHA2DS2-VASc score - Congestive heart failure/left ventricular dysfunction, Hypertension,
Age ≥75years [2 points], Diabetes, Stroke [2 points], Vascular disease, Age 65–74 years, and Sex category (female). Total scores range between 0-9; low risk CHA2DS2-VASc score: 0 male; 1 female,
intermediate: 1male, ≥2 female, high risk CHA2DS2-VASc score: ≥2 male; ≥3 female; TIA: transient ischemic attack; eGFR: estimated glomerular filtration rate, ml/min/1.73 m2; HAS-BLED score –
uncontrolled Hypertension: systolic ≥160 mmHg, Abnormal renal/liver function, Stroke, Bleeding history or predisposition, Labile INR ratio/TTR <60, Drugs/alcohol concomitantly. Total scores range
between 0-9; low risk of bleeding range between 0-2 and high risk of bleeding ≥3; SAMe-TT2R2 score – Sex female, Age<60, Medical history (more than two comorbidities), Treatment (interacting
drug, e.g. Amiodarone), Tobacco use (doubled) and Race (non-white, doubled). Total scores ranged from 0-8; probable good response to VKA therapy range between 0-2 and probable poor response
to VKA therapy ranged from ≥3; SD: standard deviation
†
* Vascular disease: prior myocardial infarction, peripheral artery disease or aortic plaque; # Lung disease: obstructive and restrictive diagnosed lung conditions; eGFR <60ml/min or as noted in
medical notes; ‡:2 missing information on ethnicity
250
4.4.2 Quality of anticoagulation control of patients with operated valvular
heart disease, with and without AF
As shown in Table 4.2, higher INR target ranges [INR 3.0-4.0; 41.4%] were used more often
to maintain effectiveness and safety of VKA therapy in the overall population of patients with
operated VHD. The overall mean (SD) TTR and PINRR for the cohort was 58.5 (14.6) and
50.1 (13.8) respectively; only 98 patients (21.5%) achieved the optimal TTR target (TTR≥70%)
during a median (IQR) of 6.2 (3.3-8.5) years of follow up.
Operated VHD patients with AF had a significantly higher INR target range (3.5, 49.4%;
p=0.03), lower mean TTR and PINRR [mean (SD) TTR 55.7 (14.2) vs. 60.1 (14.6); p=0.002
respectively; mean PINRR 47.4 (13.5) vs. 51.6 (13.7); p=0.002 respectively] (Figure 4.2),
lower proportions with optimal anticoagulation control (TTR ≥70%) (14.0% vs. 25.7%;
p=0.004) and higher proportions with sub-therapeutic INRs (28.4% vs. 23.4%; p<0.001)
(Figure 4.3) despite a similar number of INR tests compared to operated VHD patients without
AF. There was no significant difference in INRs above the therapeutic range or the proportions
of patients with one or more INR >5.0 or >8.0 (Table 4.2 and Figure 4.4).
In exploratory analyses (Table 4.3), with measures of anticoagulation control stratified
according to the different target INR ranges, patients with a higher TTR target (INR 3.5) had
a significantly lower mean (SD) TTR and PINRR [mean (SD) TTR 51.7% (11.8); p<0.001;
mean PINRR 42.6% (10.9); p<0.001 respectively], higher mean (SD) number of INR test
[122.3 (59.6); p<0.001], higher sub-therapeutic [30.1 (10.2); p<0.001] and supra-therapeutic
[26.8 (7.7); p=0.001] INRs and a significantly longer duration of VKA treatment compared to
those with target ranges of 2.5 and 3.0. In contrast, a significantly higher proportion of patients
with ≥1 INR >5.0 (95.8%; p<0.001) or >8.0 (22.2%; p<0.001) was evident among those with a
target INR of 3.0 (Table 4.3).
251
Table 4.2: Measures of anticoagulation control of patients with operated valvular heart disease, with and without AF
Total, AF No AF F- X2 value p-value

Measures of anticoagulation control, N (%)
N=456 N=164 N=292 value
Median target INR 2.5‡ 110 (24.1) 33 (20.1) 77 (26.4) -
3.0 157 (34.4) 50 (30.5) 107 (36.6) - 6.76 0.034
3.5 189 (41.4) 81 (49.4) 108 (37.0) -
Mean (SD) TTR Rosendaal* 58.5 (14.6) 55.7 (14.2) 60.1 (14.6) 0.09 - 0.002
TTR<70% 358 (78.5) 141 (86.0) 217 (74.3) -
8.46 0.004
TTR≥70% 98 (21.5) 23 (14.0) 75 (25.7) -
TTR<65% 310 (68.0) 126 (76.8) 184 (63.0) - 0.002
9.21
TTR≥65% 146 (32.0) 38 (23.2) 108 (37.0) -
Mean (SD) PINRR* 50.1 (13.8) 47.4 (13.5) 51.6 (13.7) 0.60 - 0.002
PINRR<70% 417 (91.4) 154 (93.9) 263 (90.1) - 0.16
1.97
PINRR ≥70% 39 (8.6) 10 (6.1) 29 (9.9) -
PINRR <65% 398 (87.3) 150 (91.5) 248 (84.9) - 0.05
4.04
PINRR ≥65% 58 (12.7) 14 (8.5) 44 (15.1)
Mean (SD) number of INR tests 96.2 (55.3) 100.7 (58.8) 93.7 (53.1) 0.60 - 0.19
Mean (SD) percentage INRs below the range 25.2 (12.1) 28.4 (12.5) 23.4 (11.6) 0.85 - <0.001
Mean (SD) percentage above the range 24.9 (9.5) 24.1 (8.6) 25.3 (9.9) 0.64 - 0.22
INR>5 312 (68.4) 118 (72.0) 194 (66.4) 1.48 - 0.22
INR>8 64 (14.0) 26 (15.9) 38 (13.1) - 0.70 0.40
Median (IQR) years of follow-up 6.24 (3.3-8.5) 5.7 (3.7-8.5) 5.7 (3.1-8.5) - - 0.87
AF: atrial fibrillation; INR: international normalized ratio; IQR: interquartile range; PINRR: percentage of INRs within range; SD: standard deviation; TTR: time in therapeutic range; *TTR and PINRR
were calculated based on the INR ranges obtained from the anticoagulation clinic; †Median target INR ranges for each patient were different depending on indication and type of surgery and valve
used which was set by the operating surgeon.
252
30%
26%
25%
22%
20%
%TTR/PINRR
15% 14% TTR≥70

PINRR≥70
10%
10% 9%
6%
5%
0%
All AF Without AF
Figure 4.2: Percentage of patients with optimal TTR/PINRR among operated valvular heart
disease patients with and without AF
253
120%
100%
25% 24% 25%
80%
% INRs
25% 23% % INRs above the range

60% 28%
% INRs below the range
% INRs within the range
40%
50% 52%
47%
20%
0%
All AF Without AF
Figure 4.3: Percentage of INRs within range, below the range and above the range among operated
valvular heart disease, with and without AF
80%
72%
70% 68%
66%
60%
50%
% patients
40% INRs >5

INRs>8
30%
20% 16%
14% 13%
10%
0%
All AF Without AF
Figure 4.4: Percentage of patients with INRs>5.0 and INRs >8.0 among operated valvular heart
disease, with and without AF
254
Table 4.3: Measures of anticoagulation control of patients according to different target INRs
Median INR Median INR Median INR X2 value p-value

Measures of anticoagulation control, N (%) 2.5 3.0 3.5
N=110 N=157 N=189
Mean (SD) TTR Rosendaal* 68.4 (14.4) 59.8 (13.3)a 51.7 (11.8)b, c - <0.001
a b, c
TTR<70% 53 (48.2) 124 (79.0) 181 (95.8)
a
93.3 <0.001
TTR≥70% 57 (51.8) 33 (21.0) 8 (4.2) b, c
TTR<65% 37 (33.6) 100 (63.7) a 173 (91.5) b, c <0.001
a b, c
109.1
TTR≥65% 73 (66.4) 57 (36.3) 16 (8.5)
Mean (SD) PINRR* 61.4 (12.5) 51.1 (11.8)a 42.6 (10.9)b, c - <0.001
a b
PINRR<70% 79 (71.8) 187 (98.9) 151 (96.2) <0.001
a b
72.3
PINRR ≥70% 31 (28.2) 2 (1.1) 6 (3.8)
PINRR <65% 67 (60.9) 145 (92.4) a 186 (98.4) b, c <0.001
a b, c
93.7
PINRR ≥65% 43 (39.1) 12 (7.6) 3 (1.6)
Mean (SD) number of INR tests 71.2 (41.7) 82.3 (44.1) 122.3 (59.6)b,c <0.001
Mean (SD) percentage INRs below the range 15.1 (10.1) 26.4 (11.4) a 30.1 (10.2) b, c - <0.001
Mean (SD) percentage above the range 24.0 (11.5) 23.1 (9.4) 26.8 (7.7)b, c - 0.001
a b, c
INR>5 35 (31.8) 181 (95.8) 96 (61.1) 137.5 <0.001
a b, c
INR>8 3 (2.7) 42 (22.2) 19 (12.1) 22.6 <0.001
Median (IQR) years of follow-up 5.2 (2.0-8.3) 5.7 (3.0-8.5) 6.5 (5.0-8.6)b, c - <0.001
AF: atrial fibrillation; INR: international normalized ratio; IQR: interquartile range; PINRR: percentage of INRs within range; SD: standard deviation; TTR: time in therapeutic range; *TTR and PINRR
were calculated based on the INR ranges obtained from the anticoagulation clinic; †INR ranges for each patient were different depending on indication and type of surgery and this was set by the
operating surgeon; a: significant difference between median target INR 2.5 to 3.0; b: significant difference between median target INR 2.5 to 3.5; c: significant difference between median target INR
3.0 to 3.5
255
4.4.3 Predictors of poor anticoagulation control, TTR <70%
Table 4.4 presents the results obtained from univariate logistic regression analyses
investigating the predictors of poor anticoagulation control. Being female [OR 2.2 (95% CI
1.32-3.73)], having an operated mitral valve [OR 1.8 (95% CI 1.02-3.28)] or both mitral and
aortic valves [OR 3.0 (95% CI 1.14-7.62)], AF [2.12 (95% CI 1.27-3.54)], anaemia/bleeding
history [OR 1.8 (95% CI 1.12-2.92)], digoxin [OR 5.2 (95% CI 1.85-14.69), increasing
CHA2DS2-VASc [OR 1.2 (95% CI 1.02-1.43) and HAS-BLED [OR 2.7 (95% CI 2.01-3.48)]
scores predicted poor TTR among patients with operated VHD with and without AF.
Models 1-6 in Table 4.5 present the independent factors predicting poor TTR after adjustment
for demographic and clinical variables. Being female, the presence of AF at baseline, and
anaemia/bleeding history, were consistently present in 4 of the 6 models predicting poor TTR.
The HAS-BLED score, which also contains anaemia/bleeding history, also predicted poor TTR
in 2 of the 6 models (models 4 and 6).
256
Table 4.4: Demographics and clinical characteristics associated with predictors of poor TTR (<70%), in univariate analysis among
patients with operated valvular heart disease, with and without AF
N (%) Odds ratio (95% CI) p-value
Age at implantation Age 1.00 (0.99-1.02) 0.36

Sex Female 2.22 (1.32-3.73) 0.003
White (ref) - -
Ethnicity‡
Non-White 1.22 (0.76-1.96) 0.42
Alcohol >14unit/day (N=353) 2.41 (0.71-8.15) 0.16
Social history
Smoking/ex-smoker (N=372) 0.72 (0.40-1.30) 0.28
Mitral only 1.83 (1.02-3.28) 0.044
Sites of prosthesis Aortic only 0.41 (0.24-0.69) 0.001
Mitral and aortic 2.95 (1.14-7.62) 0.025
Types of valve Mechanical valve 0.72 (0.21-2.55) 0.61
replacement Tissue valve 1.39 (0.39-4.88) 0.61
Atrial fibrillation 2.12 (1.27-3.54) 0.004
Heart failure 1.20 (0.58-2.49) 0.62
Hypertension 0.97 (0.61-1.55) 0.91

Past medical history Diabetes 1.26 (0.66-2.42) 0.48
Stroke/TIA 1.14 (0.59-2.18) 0.70
Vascular disease* 1.10 (0.66-1.84) 0.72
Lung disease# 1.44 (0.79-2.64) 0.24
Kidney disease† 0.89 (0.28-2.78) 0.84
Anaemia/bleeding history 1.81 (1.12-2.92) 0.015
257
Table 4.4 continued
Odds ratio (95% CI) p-value
Beta-blocker 0.77 (0.49-1.20) 0.25
ACEI/ARB 0.66 (0.42-1.04) 0.07
Diuretics 1.11 (0.71-1.74) 0.64
Current medications Amiodarone 1.50 (0.68-3.30) 0.32
Concurrent antiplatelet 1.21 (0.65-2.22) 0.55
Digoxin 5.21 (1.85-14.69) 0.002
Calcium channel blocker 1.08 (0.53-2.18) 0.83
CHA2DS2-VASc score Mean (SD) 1.21 (1.02-1.43) 0.028
HAS-BLED score Mean 2.65 (2.01-3.48) <0.001
SAMe-TT2R2 score Mean 1.11 (0.94-1.30) 0.21
ACEI/ARB: angiotensin converting enzyme inhibitor/ angiotensin receptor blockade; AF: atrial fibrillation; CI: confidence interval; CHA2 DS2-VASc score - Congestive heart failure/left ventricular
dysfunction, Hypertension, Age ≥75years [2 points], Diabetes, Stroke [2 points], Vascular disease, Age 65–74 years, and Sex category (female). Total scores range between 0-9; low risk CHA2DS2-
VASc score: 0 male; 1 female, intermediate: 1male, ≥2 female, high risk CHA2DS2-VASc score: ≥2 male; ≥3 female; TIA: transient ischemic attack; eGFR: estimated glomerular filtration rate,
ml/min/1.73 m2; HAS-BLED score – uncontrolled Hypertension: systolic ≥160 mmHg, Abnormal renal/liver function, Stroke, Bleeding history or predisposition, Labile INR ratio/TTR <60, Drugs/alcohol
concomitantly. Total scores range between 0-9; low risk of bleeding range between 0-2 and high risk of bleeding ≥3; SAMe-TT2R2 score – Sex female, Age<60, Medical history (more than two
comorbidities), Treatment (interacting drug, e.g. Amiodarone), Tobacco use (doubled) and Race (non-white, doubled). Total scores ranged from 0-8; probable good response to VKA therapy range
between 0-2 and probable poor response to VKA therapy ranged from ≥3; SD: standard deviation
†
* Vascular disease: prior myocardial infarction, peripheral artery disease or aortic plaque; # Lung disease: obstructive and restrictive diagnosed lung conditions; eGFR <60ml/min or as noted in
medical notes; ‡ 2 missing information on ethnicity
258
Table 4.5: Models of predictors of poor TTR (<70%) in the overall cohort of patients with operated valvular heart disease
Predictors Model 1α Model 2† Model 3‡ Model 4¥ Model 5§ Model 6¶

(OR 95% CI)
Age (continuous) 1.00 (0.98-1.02); 1.00 (0.98-1.02);
- -
p=0.98 p=0.96 1.12 (0.94-1.34); 1.12 (0.93-1.34);
Female sex 1.93 (1.13-3.30); 2.05 (1.21-3.50); p=0.21 ‡ 2.28 (1.29-4.02); p=0.23§ 2.51 (1.42-4.44);
p=0.016 p=0.008 p=0.004 p=0.002
Site of replacement 2.06 (0.77-5.48); 1.15 (0.30-4.35); 2.45 (0.93-6.44); 2.02 (0.73-5.58); 1.16 (0.31-4.36); 1.99 (0.50-7.90);
(2 valves vs. 1 valve)* p=0.15 p=0.84† p=0.07 p=0.17 p=0.83§ p=0.33¶
Atrial fibrillation 1.75 (1.01-3.03); 1.89 (1.10-3.27); 1.74 (1.01-3.00); 1.38 (0.78-2.43); 1.94 (1.13-3.33); 1.51 (0.86-2.65);
p=0.045 p=0.022 p=0.047 p=0.26 p=0.016 p=0.16
Anaemia/bleeding 1.84 (1.13-3.00); 1.86 (1.14-3.03); 1.72 (1.06-2.80); 2.60 (1.98-3.43); 1.75 (1.08-2.84); 2.65 (2.01-3.49);
history p=0.014 p=0.012 p=0.028 p=<0.001¥ p=0.024 p=<0.001¶
*2 valves: aortic AND mitral valve vs. 1 valve: aortic OR mitral valve
α
Model 1 includes age, female, site or replacement (2 vs. 1 valve), AF, anaemia/bleeding history
†
Model 2 includes age; female, type of valve (mechanical vs. tissue), AF, anaemia/bleeding history
‡
Model 3 includes CHA2DS2-VASc score, site or replacement (2 vs. 1 valve), AF, anaemia/bleeding history
¥
Model 4 includes HAS-BLED score, female, site or replacement (2 vs. 1 valve), AF
§
Model 5 includes CHA2DS2-VASc score, type of valve (mechanical vs. tissue), AF, anaemia/bleeding history
¶
Model 6 includes female, type of valve (mechanical vs. tissue), AF and HAS-BLED score
259
4.4.4 Adverse clinical outcome
Table 4.6 shows the proportions of patients with adverse clinical outcomes. Overall there were
31 TE events, 113 bleeding events, 123 CV hospitalisations, 51 deaths and 316 experiences
≥1 MACE. There were no significant differences in TE, bleeding, CV hospitalisation and ≥1
MACE between those with and without AF. However, significantly higher proportions of
patients with operated VHD and AF died [all-cause mortality (20.7% vs. 5.8%; p<0.001); CV
mortality (7.3% vs. none; p<0.001) and non-CV mortality (13.4% vs. 5.8%; p=0.009)]
compared to those operated without AF. In survival analyses, operated VHD patients with AF
had a significantly higher risk of all-cause mortality compared to those without AF (Log-Rank:
21.570; p<0.001; Figure 4.5)
Table 4.7 compares the adverse clinical outcomes by TTR ≥70% and <70% and TTR≥65%
and TTR <65%. Higher proportions of patients died (13.1% vs. 4.1%; p=0.011) and
experienced ≥1 MACE (42.7% vs. 27.6%; p=0.006) when their TTR was <70% compared to
those with TTR≥70%. In survival analyses, patients with TTR <70% had a significantly higher
risk of all-cause mortality (Log-Rank: 5.845, p=0.016; Figure 4.6) and ≥1 MACE. (Log-Rank:
7.541, p=0.006; Figure 4.7) A similar pattern emerged when TTR was stratified as <65%; a
significantly higher proportion of patients with TTR <65% died (14.2% vs. 4.8%; p=0.003) or
experienced ≥1 MACE (44.5% vs. 28.8%; p=0.001).
260
Table 4.6: Adverse clinical outcome among patients with operated valvular heart disease, with and without AF
Total, Event AF Event No AF Event

Outcomes, N (%) N=456 rate/100 pt N=164 rate/100 pt N=292 rate/100 pt p-value*
yrs yrs yrs
Stroke/TIA/SE 25 (5.5) 1.0 8 (4.9) 0.9 17 (5.8) 1.1 0.67
Bleeding* 85 (18.6) 3.6 30 (18.3) 3.6 55 (18.8) 3.6 0.89
CV hospitalisation 78 (17.1) 3.4 31 (18.9) 3.8 47 (16.1) 3.2 0.45
All-cause death 51 (11.2) 1.9 34 (20.7) 3.6 17 (5.8) 1.0 <0.001
CV death 12 (2.6) 0.5 12 (7.3) 1.3 0 - <0.001
Non-CV death 39 (8.6) 1.5 22 (13.4) 2.3 17 (5.8) 1.0 0.009
†
≥1 MACE 180 (39.5) 8.7 75 (45.7) 10.1 105 (36.0) 7.8 0.051
*p-value for proportion; Cardiovascular hospitalisation: a hospitalisation with a cardiovascular cause: i) heart failure, MI, new angina, non-fatal cardiac arrest, ventricular arrhythmia, uncontrolled
AF/atrial flutter, supraventricular arrhythmia, ii) valve surgery, CABG surgery, PTCA surgery, pacemaker/ICD insertion, carotid endarterectomy, peripheral angioplasty/surgery, limb amputation AND
as recorded in patient’s medical documents; DVT – Deep Vein Thrombosis ; * Bleeding ISTH is combination of major bleed ISTH and clinically relevant non-major bleed (CRNMB); Major Bleeding –
ISTH Major Bleeding: fatal bleeding and/or symptomatic bleeding in a critical area or organ, such as intracranial, intraspinal, intraocular, retroperitoneal, intraarticular or pericardial, or intramuscular
with compartment syndrome and/or bleeding causing a fall in haemoglobin level of 2 g/dL (1.24 mmol/L) or more, or leading to transfusion of two or more units of whole blood or red cells; Clinically
relevant non-major bleeding (CRNMB): clinically overt bleeding not satisfying the criteria for major bleeding but meet at least one of the 3 criteria: i) leading to hospitalisation or increased level of care,
ii) requiring medical intervention by healthcare professional and iii) prompting face to face evaluation; SE: systemic embolism; TIA: transient ischemic attack; VTE: venous thromboembolism.
†
≥1MACE: major adverse clinical event defined a composite of TE, bleeding, CV hospitalisation and all-cause death
261
p<0.001
No AF
AF
Number No AF 292 253 214 188 140 107

at risk
AF 164 143 128 103 77 55
Figure 4.5: Kaplan-Meier curves among operated VHD patients

stratified by the presence of AF for all-cause mortality
262
Table 4.7: Adverse clinical outcome vs. TTR among patients with operated valvular heart disease, with and without AF
TTR<70% TTR≥70% TTR<65% TTR≥65%

N (%) p-value p-value
N=358 N=98 N=310 N=146
Stroke/TIA/SE 23 (6.4) 2 (2.0) 0.13 20 (6.5) 5 (3.4) 0.19
Bleeding* 72 (20.1) 13 (13.3) 0.12 65 (21.0) 20 (13.7) 0.06
All-cause death 47 (13.1) 4 (4.1) 0.011 44 (14.2) 7 (4.8) 0.003
≥1 MACE† 153 (42.7) 27 (27.6) 0.006 138 (44.5) 42 (28.8) 0.001
patient’s medical documents; DVT – Deep Vein Thrombosis; Major Bleeding – ISTH Major Bleeding: fatal bleeding and/or symptomatic bleeding in a critical area or organ, such as intracranial,
at least one of the 3 criteria: i) leading to hospitalisation or increased level of care, ii) requiring medical intervention by healthcare professional and iii) prompting face to face evaluation; SE: systemic
embolism; TIA: transient ischemic attack; VTE: venous thromboembolism.
* Bleeding ISTH is combination of major bleed ISTH and clinically relevant non-major bleed (CRNMB); †≥1 MACE: major adverse clinical event defined a composite TE, bleeding, CV hospitalisation
and all cause death
263
p=0.016
p=0.006
TTR <70%
TTR ≥70%
TTR <70%
TTR ⩾70%
Number at risk Number at risk
TTR <70% 358 317 272 233 171 123 TTR <70% 358 275 207 160 108 67
TTR⩾70% 98 84 70 58 46 39 TTR⩾70% 98 78 64 47 36 27
Figure 4.6: Kaplan-Meier curves among operated VHD Figure 4.7: Kaplan-Meier curves among operated VHD patients
patients stratified by categories of TTR (TTR <70% vs. stratified by categories of TTR (TTR <70% vs. TTR≥70%) for
TTR≥70%) for all-cause mortality composites of thromboembolic, bleeding event, cardiovascular
hospitalisation and all-cause mortality (≥1 MACE)
264
4.5 Discussion
This study has three main findings. First, the quality of anticoagulation control was significantly
lower in operated VHD patients with AF at baseline compared to those without AF, using both
the Rosendaal and PINRR methods. Second, females, the presence of AF, and
anaemia/bleeding history significantly predicted poorer anticoagulation control in the overall
cohort. Third, the rate of death was significantly higher in those with operated VHD with AF
compared to operated VHD patients without AF. To date, this is the first study assessing the
quality of anticoagulation control among operated VHD patients stratified by the presence of
AF at baseline.
Mechanical heart valves are more thrombogenic but more long-lasting compared to tissue
valves (377). Due to this reason, patients with mechanical valve prosthesis require lifelong
anticoagulation therapy with a VKA compared to those with tissue valves (without other
indication for OAC therapy; for example, AF) who only require anticoagulation therapy for at
least the first 3 months following surgery(349, 351, 358). In this cohort of operated VHD
patients, 36.0% of the population had concomitant AF, a proportion consistent with two
recently published studies (377, 379). In the present cohort, AF patients were also significantly
older and more likely to have additional stroke risk factors such as hypertension and heart
failure and a greater proportion with AF received mitral or both mitral and aortic valve
replacement, compared to patients without AF. Moreover, operated VHD patients with AF
have significantly higher mean CHA2DS2-VASc and HAS-BLED scores than those without AF
at baseline, indicating their higher risk of TE and bleeding events.
In terms of anticoagulation management, the majority of the operated VHD patients with AF
had a higher target INR (i.e., INR 3.5) compared to those without AF. In the recent ESC
guidelines (358), a target INR of 3.5 is recommended for those patients with ≥1 risk factor (for
example AF) and with medium prosthesis thrombogenicity. In this operated VHD cohort, there
is insufficient information about the type of valve used thus it was not possible to ascertain the
265
valves’ thrombogenicity and explain the reasons behind the high target range. One possibility
is that patients who had their valves replaced from as early as the 1970’s were perhaps treated
based on recommendations from previous guidelines which suggested a target range of 3.0-
4.5 for all patients with prosthetic heart valves (regardless of type) (381). When examining
anticoagulation control among patients with different target INRs in an exploratory analysis,
TTR and PINRR were significantly lower among patients with a high target INR (INR 3.5)
compared to those with target INRs of 2.5 and 3.0, respectively. Furthermore, other markers
of poor anticoagulation control (sub-therapeutic INRs, supra-therapeutic INRs, INRs >5.0 and
INRS >8.0) were also significantly more prevalent among those with higher INR targets (INR
3.5) compared to those with lower INR targets (INR 2.5 and INR 3.0). The PLECTRUM study
(377), a retrospective observational multi-centre study among patients with mechanical heart
valves also investigated TTR according to different INR targets and showed consistent results
with the current study; lower median TTR among patients with a higher INR target (INR 3.5)
[TTR 71.5% vs. 58.6% vs. 46%; p=0.0001 for INR targets of 2.5, 3.0 and 3.5 respectively]
(377). Similarly, in the PLECTRUM cohort and the present cohort, TTR was better when the
intended INR target was kept at 2.5. These findings may imply the difficulties in achieving
INRs within the therapeutic range when a more intense anticoagulation regimen is adopted.
Overall in the present study, the mean (SD) TTR is 58.5% (14.6) and less than a quarter of
the cohort achieved optimal TTR (TTR ≥70%), reflecting poor anticoagulation control among
operated VHD patients. There is a paucity of literature on the quality of anticoagulation control
among operated VHD patients, especially those with AF. Only four studies are available
assessing TTR among VHD patients (377-380). The Swedish groups (379) examined TTR
among 534 patients (379) and 4687 patients (380) with mechanical heart valves and reported
a mean TTR of 71.3% (379) and 72.5% (380) respectively; higher than the mean TTR in the
present study. In contrast, two recent studies (377, 378), conducted in Italy (N=2357) (377)
and Denmark (N=659) (378), reported a median (IQR) TTR of 60% (47-74%) and 54.9% (39.0-
72.9%), respectively; comparable to the TTR in the current study. The findings of the Italian
266
and Danish studies (377, 378) and the current study show sub-optimal quality of
anticoagulation control among operated VHD patients. In contrast, the two Swedish study
(379, 380) showed optimal anticoagulation control among operated VHD patients although
this could be explained by the fact that generally, Sweden (379) is known to have excellent
anticoagulation management resulting in better TTR compared to other countries (206, 456).
This again reinforces one important message; the difficulties in maintaining INR levels at the
therapeutic range among anticoagulated operated VHD patients. This is more worrying in
patients with concomitant AF, as AF patients with VHD carry an even higher risk of TE
complications (5-62%)(457) than patients with NVAF (0-18%) (457).
In logistic regression analyses, after adjusting for demographics and clinical variables, being
female, the presence of AF and anaemia/bleeding history consistently predicted poor TTR in
four of the six models. In addition, the HAS-BLED (which also consist of anaemia/bleeding
history) score also predicted poor TTR (<70%) in two of six models. The finding that being
female predicts poor TTR is consistent with other non-valvular AF studies (197-203) and is
difficult to explain but could be influenced by several factors. The mean age of the overall VHD
population was 51 years which is working age. It could be that working women have more
hectic lifestyles with household, work and family responsibilities that makes them prone to
being non-adherent to medication in general including towards anticoagulation therapy thus
leading to poor anticoagulation control. One large American study (458) evaluating medication
use and adherence among 16.0 million women and 13.5 million men showed that women
were more likely to be non-adherent to their diabetic (35.4% vs. 32.5%; p<0.0001) and
antihypertensive (25.8% vs. 24.8; p<0.0001) medications compared to men respectively and
also speculated due to more complex medications regime, more side effects and more
responsibilities resulting in self neglect compared to men (458). Furthermore, in this study, the
majority of the operated VHD females also had AF at baseline which is also a predictor of poor
TTR. Operated VHD patients with AF are older and had multiple comorbidities with complex
disease management which might contribute to the lower quality of anticoagulation control
267
(197, 199, 201, 253). Lastly, history of anaemia/bleeding among operated VHD patients was
also an independent predictor of poor TTR consistent with another study among non-valvular
AF patients (213). It may be that these patients were managed more cautiously in terms of
dosing of VKA. Although information on the dosage of VKAs used was not available, perhaps
a lower dosage was used in this group of patients due to the fear of bleeding complications
thus leading to the risk of suboptimal anticoagulation control in this population. No other
studies have investigated the predictors of TTR specifically among operated VHD patients so
comparison with other studies regarding the predictors of poor TTR among operated VHD
patients could not be undertaken. However, Poli et al (377) has investigated predictors of TE
among mechanical heart valves patients and showed that AF, history of TE and prosthesis at
mitral position were associated with TE complications (377).
During a median follow up of 6.2 years, at least two-fifths of all operated VHD patients had ≥1
MACE. The rate of TE was 1.0/100 pt-yrs, which is comparable with that reported by
Cannegieter et al (362) (1.0 /100 pt-yrs) but slightly higher than the rate reported by Poli et al
(377) (0.67/100 pt-yrs), although the latter acknowledged the low rate of TE events in their
cohort despite an overall suboptimal TTR compared to other studies (362). Perhaps the higher
TE rate in the current cohort, compared to Poli et al (377), is driven by the higher proportion
of patients with history of stroke/TIA prior to valve surgery in the current study compared to
Poli et al (377) (14.5% vs. 8.3%).
The rate of bleeding events was 3.7/100 pt-yrs in the current study, higher than Cannegieter
et al (362) and Poli et al (377), with 1.4/100 pt-yrs (362) and 1.0/100 patient-years,
respectively. (377) The higher bleeding rate in the present cohort might be influenced by the
higher target INR rate used (target INR 3.5), although further analyses on predictors of
bleeding events (and all the other events) was not undertaken due to lack of power for these
analyses.
268
Additionally, in this cohort, 11% of the patients died which was also higher than that reported
from Poli et al (377) (7.4% deaths); this might be explained by differences in the demographic
and clinical characteristics of the cohorts. There were more males, patients from ethnic
minority groups, smokers/ex-smokers and a higher disease burden (stroke/TIA, diabetes,
vascular disease and anaemia) in the present study which could potentially contribute to the
differences in the mortality rate.
The proportions of operated VHD patients who had a TE, bleeding event, CV hospitalisation
and ≥1 MACE was similar among those with and without AF at baseline. Nevertheless, all-
cause mortality (including CV and non-CV related death) was significantly higher among those
with AF compared to those without AF indicating that in this cohort, patients with operated
VHD and AF have a worse prognosis than those without AF.
When investigating the impact of TTR on adverse clinical outcomes, significantly higher
proportions of deaths (all-cause mortality) and patients with ≥1 MACE were seen in the
suboptimal TTR category (TTR<70%). This suggests an increased risk of adverse clinical
outcomes among operated VHD patients in this cohort when TTR is not optimised. In contrast,
two other studies (377, 379) investigating the impact of TTR on TE events among mechanical
heart valve patients reported contrasting results. The PLECTRUM (377) study and the
Swedish registry (379) showed no relationship between poor TTR and TE events but poor
TTR (<61.6%) was associated with bleeding events in the Swedish study (adjusted OR 2.9;
p=0.011)(379). However, another study in 2002 (381) showed that mortality was significantly
increased in patients with high anticoagulation variability (381). These studies vary in some
aspects compared to the present study, especially the target INR used among the patients
(high INR target vs. normal INR target vs. patient-specific target), the method of calculating
the quality of anticoagulation control (TTR vs. anticoagulation variability), the study design
(prospective vs. retrospective), the settings (high TTR setting vs. normal TTR setting), the
year of study (2002 vs. 2018) and the sample size included, which could affect the results.
269
More studies regarding TTR and adverse clinical outcomes among operated VHD patients are
needed to confirm these findings.
4.5.1 Strengths and Limitations
This is the first study investigating anticoagulation control in the UK among operated VHD
patients stratified by the presence of AF at baseline (obtained from the post-operative notes).
Although it is limited by the relatively small sample size, it provides some insights on
anticoagulation control among operated VHD patients, with and without AF. Studies
investigating anticoagulation control among VHD patients are lacking thus the information
gained from this study adds to the limited current literature. In addition, anticoagulation control
was assessed for 6.2 years reflecting long term anticoagulation control among VHD patients.
This study is limited by its retrospective, single centre design and the small number of operated
VHD patients included, so caution must be applied as the findings might not be transferable
to other settings. There is no information on the proportion of pregnancies, the doses of VKAs
and type of valve inserted in the patients; if patients were offered PSM or home monitoring
service, distance to anticoagulation clinic, level of education, drugs and food interaction and
genetic information which could impact the quality of anticoagulation control. Additionally, this
study is not powered for adverse clinical outcomes so analyses pertaining to outcomes were
exploratory in nature.
4.5.2 Clinical implications
The findings of this study suggest that operated VHD patients with AF at baseline need closer
attention and a more robust support system than those without AF. For example, more
frequent follow up for closer INR monitoring should be arranged if patients INRs are not within
the therapeutic range. During these follow ups, knowledge regarding the need for
anticoagulation and the importance of keeping INR within the therapeutic range, the
importance of being adherent to medication and avoiding or minimising food and drug
270
interactions should be reinforced from time to time. Perhaps barriers to medication adherence
and persistence should be identified and addressed to improve adherence if this is identified
as a cause of poor TTR.
In addition, patient self-monitoring (PSM) of INR could be offered to patients who have
difficulties in coming for frequent INR visits provided that appropriate training was given in
advance and patients’ suitability for PSM has been assessed by the anticoagulant experts.
One meta-analysis (459) of eleven trials comparing self-monitoring (self-testing) or self-
management (self-testing and self-dosage) versus usual care (dosed by physician or
anticoagulation clinic) among anticoagulated patients with VKA therapy for AF, mechanical
heart valve and others showed that TE was significantly reduced in patients who were in the
PSM group compared to control [HR 0.51; 95% CI 0.31-0.85)] but there was no difference
between groups on bleeding and mortality (459). Also, patients with a mechanical heart valve
benefited from the PSM with a significant reduction in thrombotic events [HR 0.52; 95% CI
0.35-0.77)] (459).
There may be a role for pharmacogenetic testing in VHD patients with difficulties achieving
therapeutic INRs or if they are complicated with multiple strokes or bleeding events after
receiving VKA therapy. This could potentially investigate any polymorphism to the CYP2C9 or
the VKORCI enzymes which are responsible for warfarin metabolism thus affecting their
response to therapy. As NOAC is contraindicated in operated VHD patients with mechanical
heart prosthesis, every effort should be made to ensure treatment with VKA is optimal to
prevent serious adverse clinical event.
4.5.3 Future work
Future studies with more focus on the quality of anticoagulation in operated VHD patients with
AF is recommended as less attention has been paid to this population despite them having
higher risk of TE complications than NVAF patients. As more studies are utilising INR
variability as another method of measuring quality of anticoagulation control, perhaps future
271
work could include this method alongside with the TTR via the Rosendaal and PINRR methods
and investigate the correlation between these methods. While TTR measures anticoagulation
intensity, INR variability measures anticoagulation stability and both methods have been
shown to predict warfarin related adverse outcomes (460). However, to date, insufficient
evidence exist regarding which method is better in predicting adverse events, thus this would
be an area for future research.
Guidelines on antithrombotic management in operated VHD were based on observational
studies with small sample sizes and expert opinions. For example, there is as yet no
agreement about the optimal level of anticoagulation intensity (INR targets) in different patient
populations in order to have the net-clinical benefit of avoiding stroke and bleeding
complications. Thus, future work with larger sample sizes (>1000) and prospective study
designs is needed to provide greater insights into this population so that better management
strategies can be provided to patients.
4.6 Conclusion
Operated VHD patients with AF at baseline have poorer anticoagulation control compared to
those without AF at baseline. The presence of concomitant AF, anaemia/bleeding history, as
well as female gender, independently predicted poor TTR. The rate of all-cause mortality was
significantly higher among operated VHD patients with AF. These findings suggest closer INR
monitoring among operated VHD patients, especially those with AF is warranted, to improve
anticoagulation and prevent adverse clinical outcomes.
272
Chapter 5. General discussion and conclusions
Overall, this thesis included prospective (TREAT-2 study) and retrospective (study 2 and 3)
studies primarily based on one acute Trust in the West Midlands.
Study 1 (TREAT-2 study) included newly anticoagulated AF patients (warfarin or NOACs)
whereas studies 2 and 3 comprised AF and operated VHD patients on long term VKA therapy
for the prevention of thromboembolic complications. The TREAT-2 study examined self-
reported assessment of psychological measures, knowledge, beliefs and quality of life in AF
patients new to OAC therapy. In contrast, studies 2 and 3 investigated objective measures of
anticoagulation control (TTR) in AF patients in a multi ethnic population, inclusive of the elderly
and patients with different categories of chronic kidney disease and operated VHD patients
with and without AF. In addition, the prevalence of adverse clinical outcomes was also
explored in studies 2 and 3. These studies were conducted in separate cohorts in order to
achieve the objectives stated in section 1.8. However, for Study 1, due to a change in clinical
practice regarding the prescription of a NOAC instead of warfarin since 2016, there were
insufficient patients initiated on warfarin therapy within the Trust. Therefore, the comparison
of the impact of the TREAT-2 intervention on TTR among warfarin-treated patients could not
be examined. The main findings of the studies are summarised below:
273
Study 1 (TREAT-2) main findings:
• Newly anticoagulated AF patients did not appear to be depressed or anxious, had poor
knowledge of AF and its treatment and poor quality of life. Despite this, they had a
positive perception regarding their medication
• These findings remained unchanged during follow up at six months. However, more
patients were aware of the consequences of AF and for some, AF symptoms had
improved over time
Study 2 (TTR vs. ethnicity) main findings:
• Differences in the quality of anticoagulation control were evident amongst different
ethnic groups but not within elderly populations and patients with different categories
of kidney disease
• Despite similar intensity INR-monitoring, South-Asians and Afro-Caribbeans had
poorer INR control compared to Whites
• Non-white ethnicity and anaemia remained the strongest independent predictor of poor
TTR and PINRR after adjustment of demographic and clinical factors
• CV hospitalisations were more prevalent amongst the South-Asians compared to Afro-
Caribbeans and Whites
Study 3 (TTR in VHD) main findings:
• Poorer anticoagulation control was seen in operated VHD patients with AF compared
to those without AF
• Independent predictors of poor TTR included AF, anaemia/bleeding history and female
gender
• Operated VHD patients with AF had higher rates of death compared to those without
AF
274
Clinical implications
Results from the TREAT-2 study highlighted some positive aspects and identified areas for
continued development for newly anticoagulated AF patients. Reassuringly, most patients did
not report significant levels of depression or anxiety and most patients understood the
importance of taking OAC. However, there was room for improvement in terms of increasing
knowledge of AF as a chronic condition and enhancing quality of life.
Findings from Study 2 and 3 suggests that achieving good anticoagulation control is more
challenging and the prevalence of adverse clinical outcomes is more commonly seen in AF
patients from ethnic minority groups (Afro-Caribbeans and South-Asians) and among
operated VHD patients with AF. In Study 2, non-white ethnicity was a significant independent
predictor of poor-quality anticoagulation control, while Study 3 identified AF, anaemia/bleeding
history and female sex as independent predictors of poor TTR among operated VHD patients.
Therefore, in future, more efforts need be made to engage with patients from these
populations so that we can fully explore their behaviours and factors contributing to poor INR
control and develop strategies to optimise TTR and reduce adverse clinical outcomes.
Future research
To date, many countries still use VKA (rather than NOAC) for stroke prevention in AF
especially within Asian countries and therefore the TREAT-2 study could be conducted in
these countries. Malaysia, a multi-ethnic country consists of 67.4% Bumiputra (Malays and
Indigenous Bumiputra), 24.6% Chinese, 7.3% Indians and 0.7% others (461) still uses warfarin
as the main OAC of choice for stroke prevention in AF as well as the treatment of other
conditions like venous thromboembolism, pulmonary embolism, mechanical heart valves
transplantation.
275
A further prospective study/registry in a Malaysian cohort (N>1000 including data from the
warfarin Medication Therapy Adherence Clinic at regional public hospitals) might further help
fill the knowledge gap in understanding the impact of ethnicity on quality of anticoagulation
control and adverse clinical outcomes. Adult AF patients who are newly prescribed with OAC
therapy (warfarin or NOACs) for: i) stroke prevention in AF ii) prevention of TE complications
among operated VHD patients can be included and followed up for 2 years. Demographics
and clinical determinants of TTR among the Malaysian population can be investigated.
Furthermore, it would be of interest to examine the impact of herbal medications/products on
TTR as its usage is extensive among Malaysians due to aggressive promotions by promoters
of herbal medicines. As mentioned in section 1.5.1.2.1 page 91-92, concomitant use of herbal
medicines/products with warfarin therapy has resulted in a major safety concern due to
warfarin-herbal interaction which might potentially increase the risk of thrombosis and bleeding
complications. Thus, future work in this area is needed. Ancillary analyses might also include
validation of the SAMe-TT2R2 score in the Malaysian cohort as the score includes non-white
ethnicity as one of the predictors of anticoagulation control. Exploratory analysis can
additionally investigate TE and bleeding outcomes in relation to TTR cut off values of ≥70%
and <70%. Moreover, the incidence of TE, bleeding and mortality between patients prescribed
with warfarin vs. NOAC could also be investigated at a population level. These analyses can
be investigated separately among AF and operated VHD patients as data on both cohorts are
limited in Malaysia.
Studies focusing on patient knowledge, psychological aspects and quality of life in
anticoagulated patients in Malaysia are lacking. Thus, it would be of essence to investigate
these elements and extend the TREAT-2 study to the Malaysian anticoagulated AF patients
managed by the warfarin Medication Therapy Adherence Clinic (more details can be found
in section 2.5.2, pages 169-170). A similar study (TREATS-AF study) is also being planned
in Thailand with additional aims of investigating the cost-effectiveness of educational-
behavioural intervention, cross cultural adaptation, acceptance and satisfaction of this
276
intervention among Thai AF patients (MRC grant number: MR/ R020892/1 and personal
communication with Dr Lane and Professor Lip).
Some patients prescribed long term warfarin therapy may experience worsening or
improvements of TTR over time as there are likely to be changes in their comorbidities, drug
therapy, anticoagulation management system, etc. that can influence anticoagulation control
as stated in section 1.5.1. One recent Italian study (462) has shown that about 20% of their
NVAF patients (N=1341) showed a worsening in TTR over time (mean follow up 37.7 months)
and this was associated with increased risk of cardiovascular events [HR 2.1 (95% CI 1.06-
4.14); p=0.03]. Hence, another retrospective longitudinal study could also be designed to
examine temporal trends in TTR among NVAF patients receiving long-term warfarin therapy
in a Malaysian cohort. It would be of value to build on this area of interest and determine if
worsening TTR predicts worse CV outcomes as no study to date has investigated this in a
Malaysian population.
The information gained from the proposed future research could then be conveyed to key
stakeholders managing patients on anticoagulation therapy so that appropriate actions can be
taken to improve their services for the benefits of the patients. Apart from that, these results
could add to the current body of knowledge in the field of anticoagulation among Asian patients
as this information is currently limited
Conclusions
Among newly anticoagulated AF patients, improvements are needed in AF knowledge.
Although quality of life was reduced, most patients were not significantly anxious or depressed
and they hold positive beliefs about their medication. Meanwhile, good anticoagulation control
is more difficult to achieve in non-white AF patients and operated VHD patients with AF.
Predictors of poor TTR include non-white ethnicity and anaemia in the former and the
277
presence of AF, anaemia and female sex in the latter. Lastly, the prevalence of CV
hospitalisation was more common in South-Asian patients while mortality rates were higher
among operated VHD patients with AF.
278
Appendices
279
Appendix 1
Table A1.1 Proforma

Date of study entry: __________ The TREAT-2 study
RXK number:____________ DOB:___/___/____ SEX: Male Female
Age: ____________
Weight:_______Height:_______ BMI:_____________ BP: _______ HR:_________
Group: AF date diagnosed: ___/____/___OAC type and start date:_____________

1 AF (warfarin + usual care) Paroxysmal
2 AF (NOAC control) Persistent
3 AF (warfarin + TREAT) Long-standing persistent
4 AF (NOAC+ usual care) Permanent
1. Ethnic origin:____________________
2. Education level
Primary school Secondary school College/university
3. Age of leaving formal education__________________________

4. Marital status
Married Separated Widowed
Single Divorced
5. Past Medical History

Hypertension Diabetes mellitus (DM) Coronary artery
disease/myocardial
infarction/heart
attack
Peripheral arterial disease Gastritis Previous
(PAD) stroke/transient
ischemic attack
(TIA)
Asthma Renal disease, eGFR______ Hepatic disease
COPD DVT/PE Bleeding event
Congestive heart failure EHRA class Thyroid disease
(CHF), NYHA class_____ I(no symptoms) Others:
II(mild symptoms-normal activity x affected)
III(severe symptoms-NA affected)
IV(disabling symptoms)
6. Current smoking status
Yes No Stopped smoking
past 2 yrs
7. Alcohol use:______________
8. Medications:
280
9. HAS-BLEED score: _____________ 10. CHA2DS2-VASc score:_________
Hypertension CHF
Abnormal liver/renal(1/2) Hypertension
Stroke Age ≥75 (2)
Bleeding Diabetes
Labile INR Stroke/TIA/TE (2)
Elderly >65 Vascular Disease
Drugs/alcohol (1/2) Age 65-74
Female
11. SAMe-TT2R2 score

Sex (female)
Age (<60y)
Medical history
Treatment
Tobacco use (within 2 yrs) (2)
Race (not white) (2)
Total score
12. Inclusion criteria

Adult (>18 years old)
Women CHA2DS2-VASc score of ≥2
Men CHA2DS2-VASc score of ≥1
NVAF
OAC-naïve
13. Follow up event and dates

Stroke
TIA
Systemic embolism (PE/DVT)
Major bleeding ISTH
CRNMB bleeding ISTH
CV hospitalisation
Death
14. Pill count NOAC at 6 months

received all boxes not received all boxes pills left in the box________________
15. VKA RESULTS

Anticoagulant used: _____________
Start date: _______________End date: __________
TTR (Rosendaal): _______________PINRR: __________
Days on therapy: _______________Total INR: ___________
16. INR<2: ___________INR>3: ___________INR>5: __________INR>8: ________
17. Lab results
HB
Creatinine
eGFR (ml/min)
ALT/ALP
281
Appendix 2
Baseline and 6 months questionnaire
282
BASELINE QUESTIONNAIRE
TREAT-2: Oral anticoagulation in AF

stratified by SAMe-TT2R2 score
283
This questionnaire asks about your personal knowledge of atrial fibrillation, your beliefs about
medication, how you feel generally and about your quality of life after being diagnosed with atrial
fibrillation.
Please read all the questions or statement carefully. If you are unsure about which response to
give to a question or statement, please choose the one that appears most appropriate. This is
often your first response. After you have answered ALL questions, please return them in the pre-
paid envelope given to you with the questionnaires.
Thank you in advance for your kind cooperation.
284
Study ID:
Today’s date:
Date of birth: Age:
Sex: Male Female
Part 1: Background
For each of the following questions please tick (√) one response.
1. Ethnic origin (please tick (√) only one box within one ethnic group)
White
White British Gypsy or Irish Traveller Other white
Irish
Mixed /multiple ethnic groups
White and Black Caribbean White and Asians White and Black
Others African
Asian/Asian British
Indian Bangladeshi Other Asian
Pakistani Chinese
Black/African/Caribbean/Black British
African Caribbean Other Black
Other ethnic group
Arab Other ethnic group
2. Education level
Primary school Secondary school College/university

What age did you leave full-time education? (please write in age)
_________________
3. Marital status: Are you currently (tick (√) one response only)
Married Separated Widowed

Single Divorced
4.Do you currently smoke?
Yes No
If no, have you stopped smoking in the last 2 years? Yes No
285
PART 2:
This section asks about how you have felt generally over the last 2 weeks. Please read each
of the statements below. Please circle one answer for each statement.
More
Over the last 2 weeks, how often have you been Nearly
than half
bothered by any of the following problems? Not at Several every
the days
all days day
1. Little interest or pleasure in doing things 0 1 2 3
2. Feeling down, depressed, or hopeless 0 1 2 3
3. Trouble falling or staying asleep, or sleeping too much 0 1 2 3
4. Feeling tired or having little energy 0 1 2 3
5. Poor appetite or overeating 0 1 2 3
6. Feeling bad about yourself — or that you are a failure or 0 1 2 3

have let yourself or your family down
7. Trouble concentrating on things, such as reading the 0 1 2 3

newspaper or watching television
8. Moving or speaking so slowly that other people could have

noticed? Or the opposite — being so fidgety or restless 0 1 2 3
that you have been moving around a lot more than usual
9. Thoughts that you would be better off dead or of hurting 0 1 2 3
yourself in some way
If you have experienced any problems, how difficult have these problems made it for you
to do your work, take care of things at home, or get along with other people? Please tick
(√) one of the responses below.
Not difficult Somewhat Very Extremely

at all difficult difficult difficult
286
This section asks about how you have felt generally over the last 2 weeks. Please read each
of the statements below. Please circle one answer for each statement.
More
Over the last 2 weeks, how often have you been Nearly
than half
bothered by any of the following problems? Not at Several every
all days the days day
1. Feeling nervous, anxious or on edge 0 1 2 3
2. Not being able to stop or control worrying 0 1 2 3
3. Worrying too much about different things 0 1 2 3
4. Trouble relaxing 0 1 2 3
5. Being so restless that it is hard to sit still 0 1 2 3
6. Becoming easily annoyed or irritable 0 1 2 3
7. Feeling afraid as if something awful 0 1 2 3

might happen
If you have experienced any problems, how difficult have these problems made it for you
to do your work, take care of things at home, or get along with other people? Please tick
(√) one of the responses below.
Not difficult Somewhat Very Extremely

at all difficult difficult difficult
287
This section asks about your knowledge of atrial fibrillation. Please read each of the
questions below. Tick (√) one answer for each question.
1. What are the trigger factors for atrial fibrillation?

Allergy to grass, animals or house dust
Alcohol, coffee or spicy food
Noise or loud sounds
2. Why is it important to take my medication for atrial fibrillation properly?
Because the doctor wants me to
To prevent severe consequences of the arrhythmia
To prevent the possibility of a heart attack or sudden death
3. If atrial fibrillation is identified without the patient experiencing any complaints, the
patient should immediately visit the hospital.
True
False
Don’t know
4. What is atrial fibrillation?
A heart disease in which the heart is not able to pump a sufficient amount of blood through
the body
A blood disorder causing blood clots in the heart
An electric disorder in the atria of the heart which results in the heart contracting too fast
and irregularly
5. Why is oral anticoagulation medication prescribed in certain patients with atrial
fibrillation?
To prevent the risk of blood clots which can cause a stroke
To make the blood flow more easily through the body
To prevent fluid retention in the body
6. Why should a person using anticoagulation medication be careful with the use of
alcohol?
Alcohol increases the retention of fluid in the body resulting in the blood becoming too thin
Alcohol causes a blockage of the blood vessels which in turn, slows blood flow to the heart
Alcohol influences the effect of the medication and this effects the clotting ability of the
blood
7. Atrial fibrillation is a rare condition.
288
True
False
Don’t know
8. It is particularly risky if a person does not feel his/her atrial fibrillation.
True
False
Don’t know
9. Which statement with regard to physical exercise is true of patients with atrial
fibrillation?
It is important for patients to rest in order to maintain normal heart activity

Patients with chronic atrial fibrillation cannot work fulltime
It is important to exercise normally within personal limitations
10. Which statement is true?
Atrial fibrillation is life endangering because it can result in a heart attack

Atrial fibrillation is completely harmless
Atrial fibrillation is harmless if the right medication is taken
11. What is the function of the anticoagulation clinic?
To monitor blood clotting and the number of tablets taken each day
To determine if the arrhythmia is present
To determine if the patient needs to continue taking oral anticoagulation
289
This section asks about your views on medicines prescribed to you. Please read each of
the statements and indicate the extent to which you agree or disagree by circling one
response.
Strongly Disagree Uncertain Agree Strongly

disagree agree
1. My health, at present, depends on
1 2 3 4 5
my medicines
2. Having to take medicines worries
1 2 3 4 5
me
3. My life would be impossible without
1 2 3 4 5
my medicines
4. Without my medicines I would be
1 2 3 4 5
very ill
5. I sometimes worry about long-term
1 2 3 4 5
effects of my medicines
6. My medicines are a mystery to me
1 2 3 4 5
7. My health in the future will depends
1 2 3 4 5
on my medicines
8. My medicines disrupt my life
1 2 3 4 5
9. I sometimes worry about becoming
1 2 3 4 5
too dependent on my medicines
10. My medicines protect me from
1 2 3 4 5
becoming worse
11. Doctors use too many medicines
1 2 3 4 5
12. People who take medicines should
stop their treatment for a while
1 2 3 4 5
every now and again
13. Most medicines are addictive

1 2 3 4 5
14. Natural remedies are safer than
1 2 3 4 5
medicines
15. Medicines do more harm than good
1 2 3 4 5
16. All medicines are poisons
1 2 3 4 5
17. Doctors place too much trust on
1 2 3 4 5
medicines
18. If doctors had more time with
patients they would prescribe fewer 1 2 3 4 5
medicines.
290
This section refers to how atrial fibrillation affects your quality of life. Please read each of
the statements below. Tick (√) one answer for each statement.
Are you currently in atrial fibrillation? Yes No
If No, when was the last time you were aware of having had an episode of atrial fibrillation?
Please tick (√) one answer, which best describes your situation:
Earlier today
Within the past week

Within the past month
1 month to 1 year ago
More than 1 year ago
I was never aware of having atrial fibrillation
On a scale of 1 to 7, over the past 4 weeks, as a result of your atrial fibrillation, how much were you
bothered by: (Please circle one number which best describes your situation)
Not at all Hardly A little Moderately Quite a Very Extremely

bothered bothered bothered bothered bit bothered bothered
or I did not bothered
have this
symptoms
1. Palpitations: 1 2 3 4 5 6 7
Heart fluttering,
skipping or
racing
2. Irregular heart 1 2 3 4 5 6 7
beat
3. A pause in heart 1 2 3 4 5 6 7
activity
4. Lightheadedness 1 2 3 4 5 6 7
or dizziness
limited limited limited limited bit limited limited
limited
5. Ability to have 1 2 3 4 5 6 7
recreational
pastimes, sports,
and hobbies
6. Ability to have a 1 2 3 4 5 6 7
relationship and
do things with
friends and
family
291
No Hardly A little Moderate Quite a A lot of Extreme
difficulty any difficulty difficulty bit of difficulty difficulty
at all difficulty difficulty
7. Doing any activity 1 2 3 4 5 6 7
because you felt
tired, fatigued, or
low on energy
8. Doing physical 1 2 3 4 5 6 7
activity because of
shortness of breath
9. Exercising 1 2 3 4 5 6 7
10. Walking briskly 1 2 3 4 5 6 7
11. Walking briskly 1 2 3 4 5 6 7

uphill or carrying
groceries or other
items, up a flight of
stairs without
stopping
12. Doing vigorous 1 2 3 4 5 6 7
activities such as
lifting or moving
heavy furniture,
running, or
participating in
strenuous sports
like tennis or
racquetball
bothered bothered bothered bothered bit bothered bothered
bothered
13. Feeling worried or 1 2 3 4 5 6 7
anxious that your
atrial fibrillation can
start anytime
14. Feeling worried that 1 2 3 4 5 6 7
atrial fibrillation may
worsen other medical
conditions in the long
run
292
Not at all Hardly A little Moderately Quite a bit Very Extremely
bothered bothere bothered bothered bothered bothered bothered
d
15. Worrying about 1 2 3 4 5 6 7
the treatment
side effects from
medications
complications or
side effects from
procedures like
catheter
ablation,
surgery, or
pacemakers
therapy
side effects of
blood thinners
such as
nosebleeds,
bleeding gums
when brushing
teeth, heavy
bleeding from
cuts, or bruising.
18. Worrying or 1 2 3 4 5 6 7
feeling anxious
that your
treatment
interferes with
your daily
activities
Extremely Very Somewha Mixed with Somewha Very Extremely
satisfied satisfied t satisfied t dissatisfi dissatisfi
satisfied and dissatisfi ed ed
dissatisfie ed
d
19. How well your 1 2 3 4 5 6 7
current
treatment
controls your
atrial fibrillation?
293
20. The extent to 1 2 3 4 5 6 7
which treatment
has relieved
your symptoms
of atrial
fibrillation?
Thank you for completing the questionnaires. Please return them in the stamped-
addressed envelope provided.
294
Appendix 3
Table A3.1: Baseline psychological measures, knowledge and beliefs about medication of AF patients overall and in Group
1, 2, 3 and 4

Baseline measures Overall, Group 1 Group 2 Group 3 Group 4
N=139 N=9 N=102 N=4 N=24
Median (IQR) score 4.0 (1.0-8.0) 2.0 (0.5-4.0) 4.0 (1.0-8.0) 2.5 (0-16.3) 5.0 (2-11)
Minimal 0-4, (%) 80 (57.6) 8 (88.9) 59 (57.8) 2 (50.0) 11 (45.8)
Mild 5-9, (%) 31 (22.3) 1 (11.1) 24 (23.5) 1 (25.0) 5 (20.8)
Moderate 10-14, (%) 20 (14.4) 0 13 (12.7) 0 7 (29.2)
Moderately severe 5 (3.6) 0 5 (4.9) 0 0
15-19, (%)
Severe depression 3 (2.2) 0 1 (1.0) 1 (25.0) 1 (4.2)
20-27, (%)
Median (IQR) score 1.0 (0-5) 0 1.0 (0-2.5) 3.0 (0.3-20.3) 1.0 (0-7.8)
Minimal 0-4, (%) 100 (71.9) 8 (88.9) 74 (72.5) 2 (50.0) 16 (66.7)
Mild 5-9, (%) 23 (16.5) 0 18 (17.6) 1 (25.0) 4 (16.7)
Moderate 10-14, (%) 8 (5.8) 1 (11.1) 5 (4.9) 1 (25.0) 1 (4.2)
Severe anxiety 15- 8 (5.8) 0 5 (4.9) 0 3 (12.5)
21, (%)
AF knowledge scale (11 items; scores range from 0-11)
Total scores, mean 5.7 (1.7) 6.8 (1.4) 5.5 (1.7) 6.8 (1.0) 6.0 (1.6)
(SD) (min-max: 0-11)
Total scores, 0-100% 52.0 (15.4) 61.6 (12.7) 50.2 (15.5) 61.4 (8.7) 54.5 (14.9)
AF in general correct 24.5 (22.9) 29.6 (33.3) 22.5 (22.6) 41.7 (16.7) 27.8 (23.4)
score, % (3
questions)
295
Baseline measures Overall, Group 1 Group 2 Group 3 Group 4
N=139 N=9 N=102 N=4 N=24
AF symptoms 46.5 (32.0) 55.6 (33.3) 43.8 (32.5) 58.3 (31.9) 52.8 (29.4)
recognition correct
score, % (3
questions)
AF treatment correct 71.8 (20.5) 84.4 (16.7) 70.6 (21.0) 75.0 (25.2) 71.7 (18.6)
score, % (5
questions)
Beliefs about medication (BMQ; 18 items)
General overuse 10.5 (2.9) 10.1 (3.1) 10.5 (2.9) 11.3 (2.5) 10.9 (3.1)
(4-20)
General harm 8.6 (2.9) 6.7 (0.9) 8.6 (2.9) 9.0 (9.5) 9.1 (2.9)
(4-20)
Specific necessity 19.0 (3.0) 19.1 (2.8) 18.9 (3.1) 19.3 (1.7) 19.3 (3.0)
(5-25)
Specific concern 13.3 (3.5) 12.4 (2.7) 13.1 (3.5) 11.8 (3.1) 14.5 (3.7)
(5-25)*
Necessity-concern 5.7 (4.2) 6.7 (4.2) 5.8 (4.2) 7.5 (3.7) 4.9 (4.1)
differential
Symptoms (0-100)* 79.2 (58.3-95.8) 58.3 (41.7-97.9) 83.3 (62.5-96.9) 62.5 (38.5-92.7) 62.5 (50.0-86.5)
Daily activity (0-100) 60.4 (39.6-79.2) 68.8 (31.3-80.2) 60.4 (39.6-81.3) 35.5 (27.6-95.8) 59.4 (37.0-68.2)
Treatment concern 75.0 (52.8-86.1) 77.8 (72.2-84.7) 75.0 (55.6-91.7) 68.1 (50.0-75.7) 65.3 (25.0-83.3)
(0-100)*
Satisfaction (N=111) 75.0 (66.7-83.3) 75.0 (60.4-89.6) 75.0 (66.7-83.3) 58.3 (50.0-) 70.8 (50.0-83.3)
(0-100)*
Overall global score 66.7 (53.7-77.8) 59.3 (50.9-80.1) 66.7 (54.4-83.3) 65.7 (38.7-87.3) 63.9 (39.1-74.8)
(0-100)*
*mean (SD); PHQ-9: Patient Health Questionnaire to measure depression; GAD-7: Generalised Anxiety Disorder to measure anxiety; AF knowledge scale to measure
knowledge of atrial fibrillation; BMQ: beliefs about medication questionnaire; AFEQT: Atrial Fibrillation Effect on Quality of Life Questionnaire
296
Table A3.2: Six month-follow up psychological measures, knowledge and beliefs about medication of AF patients overall
and in Group 1, 2, 3 and 4

Follow up Overall, Group 1 Group 2 Group 3 Group 4
measures N=105 N=7 N=80 N=2 N=16
Median score (IQR) 4.0 (0.0-9.0) 3.0 (0-8.0) 4.0 (0-9.0) 2.0 (0-0) 5.0 (0.7-9.5)
Minimal 0-4, (%) 56 (54.4) 5 (71.4) 44 (55.0) 2 (50.0) 7 (43.8)
Mild 5-9, (%) 25 (24.3) 1 (14.3) 19 (23.8) 1 (25.0) 5 (31.3)
Moderate 10-14, (%) 15 (14.6) 0 12 (15.0) 0 3 (18.8)
Moderately severe 6 (5.8) 1 (14.3) 4 (5.0) 0 1 (6.3)
15-19, (%)
Severe depression 1 (1.0) 0 1 (1.3) 1 (25.0) 0
20-27, (%)
Median score (IQR) 1.0 (0-5.0) 0 (0-5) 1.0(0-5.8) 0 (0-5) 1.0 (0-4.8)
Minimal 0-4, (%) 68 (66.0) 4 (57.1) 52 (65.0) 2 (50.0) 12 (75.0)
Mild 5-9, (%) 25 (24.3) 2 (28.6) 21 (26.3) 1 (25.0) 2 (12.5)
Moderate 10-14, (%) 6 (5.8) 0 5 (6.3) 1 (25.0) 1 (6.3)
Severe anxiety 15- 1.0 (1.0) 1 (14.3) 2 (2.5) 0 1 (6.3)
21, (%)
AF knowledge scale (11 items; scores range from 1-11)
Total scores, mean 5.9 (1.9) 7.0 (1.3) 5.9 (1.8) 7.0 (1.4) 5.4 (2.1)
(SD) (min-max: 0-11)
Total scores, 0-100% 53.9 (16.9) 63.6 (11.7) 53.6 (16.6) 63.6 (12.9) 49.4 (19.3)
AF in general correct 18.4 (22.6) 9.5 (16.3) 18.3 (22.4) 16.7 (23.6) 23.0 (26.4)
score, % (3
questions)
AF symptoms 54.3 (31.4) 71.4 (40.5) 54.2 (31.5) 50.0 (23.6) 47.9 (27.1)
recognition correct
score, % (3
questions)
297
Follow up Overall, Group 1 Group 2 Group 3 Group 4
measures N=105 N=7 N=80 N=2 N=16
AF treatment correct 74.9 (24.2) 91.4 (15.7) 74.5 (22.5) 100 66.3 (31.6)
score, % (5
questions)
BMQ (18 items)
General overuse 10.7 (2.9) 10.7 (1.1) 11.0 (2.9) 11.0 (2.8) 9.3 (3.0)
(4-20)*
General harm 8.2 (2.4) 8.4 (1.4) 8.3 (2.4) 8.5 (0.7) 7.8 (2.8)
(4-20)*
Specific necessity (5- 19.1 (3.1) 18.6 (1.3) 19.0 (3.0) 19.5 (2.1) 19.5 (4.1)
25)*
Specific concern 12.9 (3.8) 14.0 (4.7) 12.8 (3.8) † 13.2 (3.7)
(5-25)*
Necessity-concern 6.1 (4.4) 4.6 (4.3) 6.2 (4.2) 8.5 (2.1) 6.3 (5.3)
differential
Symptoms* 83.3 (64.6-100) 75.0 (50.0-95.8) 87.5 (66.7-100) 79.2 (58.3-.) 68.8 (53.1-79.2)
(0-100)
Daily activity* 54.2 (34.4-77.1) 66.7 (20.8-85.4) 54.2 (35.4-80.7) 46.9 (22.9-.) 49.0 (29.2-69.8)
(0-100)
Treatment concern * 72.2 (58.3-88.9) 72.2 (41.7-83.3) 72.2 (59.0-88.2) 81.9 (63.9-.) 69.4 (35.4-97.2)
(0-100)
Satisfaction* 83.3 (66.7-91.7) 66.7 (66.7-83.3) 83.3 (66.7-91.7) 91.7 (91.7-91.7) 83.3 (68.8-83.3)
(0-100)
Overall global score* 66.7 (49.1-81.9) 77.1 (38.0-83.3) 69.0 (50.9-83.1) 65.7 (56.5-.) 61.6 (43.9-70.0)
(0-100)
*mean (SD); †Patients in Group 3 did not answer; PHQ-9: Patient Health Questionnaire to measure depression; GAD-7: Generalised Anxiety Disorder to measure
anxiety; AF knowledge scale to measure knowledge of atrial fibrillation; BMQ: beliefs about medication questionnaire; AFEQT: Atrial Fibrillation Effect on Quality of
Life Questionnaire
298
Appendix 4
Table A4.1: Project proforma: Ethnic differences in anticoagulation control among atrial fibrillation patients
receiving warfarin for stroke prevention
Demographics
RXK number: ______________ DOB: ____/___/____Ethnic origin: _______________
Patient number: ______________ Age: ____ Sex: M□ F□
Clinical information
AF date diagnosed: _________________ VKA naïve: yes □ no □
- Paroxysmal □
- Persistent □
- Long-standing persistent □
- Permanent □
Past Medical History Details: Date diagnosed
Valve
Valvular heart disease □ Mitral stenosis □ replacement □ Rheumatic □ _____
_____
Stroke □ Ischaemic □ Haemorrhagic □ _____
_____
TIA □ Total_________ _____
History of TE □
_____
Coronary artery disease □ MI □ CABG □ PCI □ _____
_____
Heart failure □ HF-PEF □ HF-REF □ EF ______ % _____
_____
Hypertension □ Essential □ Secondary □ _____
_____
Diabetes □ Type 1 □ Type 2 □ _____
_____
Lung disease □ Obstructive □ Restrictive □ _____
_____
Vascular disease □ PAD □ Aortic Plaque □ MI □ _____
_____
Cardiomyopathy □ Dilated □ Hypertrophic □ _____
_____
Kidney disease □ <60 ml/ml □ <30 ml/min □ Dialysis □ _____
_____
Hypercholesterolemia □ _____
Alcohol yes □ no □ Current smoker □ Ex-smoker □ Never smoked □
Lab results
Hb
Egfr
Creatinine
Alt
Alp
299
Clinical Outcomes Details: Yes □ Total________ No □
Stroke □ Total______ Ischaemic □ Haemorrhagic □ __________
TIA □ Total______ __________

VTE DVT PE
□ First □ Recurrent □ ________
Bleeding □ Major ISTH* □ CR Non-Major ** □ NCR □ ________
Hospitalisation □
- CV cause Total__________ __________
- Bleeding cause □ ________________________ Total__________ __________

□ ________________________
- Stroke/TIA cause □ ________________________ Total__________ __________
Other cause hospitalization □ __________________________________________________________________
Death □ Cause: ______________________________Date: ____/____/____

Baseline medication list Date: Most recent medication list Date:
1. 1.
2. 2.
3. 3.
4. 4.
5. 5.
6. 6.
7. 7.
8. 8.
9. 9.
10. 10.
VKA results
Anticoagulant used:__________________ Start date: ____/____/____ End date: ____/____/____
Average dose: ________
Target INR range: _______
TTR Rosendaal:_________ %INR range:_____
Days on therapy: ____________ Total INRs:_______
INR<2: _________ INR>3: ____________ INR>5:__________ INR>8:___________
*Major Bleeding: fatal bleeding and/or symptomatic bleeding in a critical area or organ, such as intracranial, intraspinal, intraocular, retroperitoneal, intraarticular or pericardial, or intramuscular
with compartment syndrome and/or bleeding causing a fall in haemoglobin level of 2 g/dL (1.24 mmol/L) or more, or leading to transfusion of two or more units of whole blood or red cells;
Clinically relevant non-major bleeding (CRNMB): clinically overt bleeding not satisfying the criteria for major bleeding and that led to hospitalisation, physician medical or surgical treatment,
or a change in antithrombotic therapy
300
INR Recordings Patient number _____________ Hospital number___________
DATE INR DOS DATE INR DOSE DATE INR DOSE

E
1. 35. 69.
2. 36. 70.
3. 37. 71.
4. 38. 72.
5. 39. 73.
6. 40. 74.
7. 41. 75.
8. 42. 76.
9. 43. 77.
10. 44. 78.
11. 45. 79.
12. 46. 80.
13. 47. 81.
14. 48. 82.
15. 49. 83.
16. 50. 84.
17. 51. 85.
18. 52. 86.
19. 53. 87.
20. 54. 88.
21. 55. 89.
22. 56. 90.
23. 57. 91.
24. 58. 92.
25. 59. 93.
26. 60. 94.
27. 61. 95.
28. 62. 96.
29. 63. 97.
30. 64. 98.
31. 65. 99.
32. 66. 100.
301
Table A4.2: Logistic regression for TTR<70% (using Rosendaal method) in relation to demographic and clinical factors
Univariate OR (95% CI) p-value Multivariate OR (95% CI) p-value

Age at first INR¥ 0.99 (0.98-1.0) 0.11 0.99 (0.97-1.0) 0.12
Female sex 1.05 (0.82-1.36) 0.69 1.21 (0.87-1.68) 0.26
Smoking history 1.05 (0.78 -1.42) 0.73 1.27 (0.91-1.77) 0.16
Ethnicity (non- white) 2.44 (1.72-3.47) <0.001 2.62 (1.67-4.10) <0.001
Hypertension 0.85 (0.62-1.16) 0.29 0.83 (0.56-1.27) 0.35
Stroke/TIA 0.69 (0.50-0.96) 0.027 0.82 (0.55-1.22) 0.33
Heart failure 1.60 (1.10-2.33) 0.014 1.45 (0.94-2.24) 0.09
Diabetes 1.10 (0.81-1.50) 0.55 0.76 (0.51-1.13) 0.17
Vascular disease 1.83 (1.28-2.61) 0.001 1.81 (1.16-2.83) 0.01
Kidney disease 1.01 (0.78-1.31) 0.93 0.97 (0.69-1.35) 0.85
Anaemia 1.96 (1.34-2.85) <0.001 1.65 (1.00-2.70) 0.05
Bleeding history 1.51 (0.93- 2.43) 0.09 1.51 (0.83-2.75) 0.17
¥
continuous variable
INR: international Normalised Ratio; TIA: transient ischemic attack TTR: Time in therapeutic range
302
Table A4.3: Logistic regression for PINRR <70% (using PINRR method) in relation to demographic and clinical factors
Univariate OR (95% CI) p-value Multivariate OR (95% CI) p-value

Age at first INR¥ 1.0 (0.98-1.02) 0.79 0.99 (0.97-1.02) 0.58
Female sex 1.17 (0.81-1.68) 0.40 0.98 (0.62-1.56) 0.95
Smoking history 0.79 (0.52-1.21) 0.28 0.94 (0.60-1.49) 0.80
Ethnicity (non- white) 3.34 (1.72-6.49) <0.001 3.47 (1.44-8.34) 0.005
Hypertension 0.90 (0.57-1.41) 0.64 0.80 (0.45-1.43) 0.45
Stroke/TIA 1.21 (0.75-1.98) 0.44 1.17 (0.66-2.08) 0.59
Heart failure 1.30 (0.75-2.26) 0.36 1.31 (0.70-2.45) 0.40
Diabetes 1.55 (0.95-2.54) 0.08 0.98 (0.55-1.78) 0.96
Kidney disease 0.79 (0.55-1.14) 0.21 0.66 (0.42-1.05) 0.08
Anaemia 2.45 (1.26-4.79) 0.009 6.27 (1.89-20.94) 0.003
Bleeding history 1.50 (0.71-3.19) 0.29 1.0 (0.40-2.54) 0.98
¥
continuous variable
303
Table A4.4: Cox proportional hazard regression analysis for the outcome of thromboembolic events (including stroke,
transient ischaemic attack and pulmonary embolism)
Univariate HR (95% CI) p-value Multivariate p-value

Age¥ 0.99 (0.96-1.02) 0.51 0.97 (0.94-1.01) 0.16
Female sex 0.82 (0.46-1.45) 0.50 0.59 (0.28-1.23) 0.16
Smoking history 0.67 (0.33-1.33) 0.17 0.61 (0.28-1.30) 0.20
Ethnicity (non- white) 1.43 (0.73-2.80) 0.29 1.14 (0.49-2.68) 0.76
Hypertension 1.26 (0.61-2.60) 0.53 1.12 (0.44-2.85) 0.81
Stroke/TIA history 2.40 (1.33-4.30) 0.003 2.29 (1.12-4.68) 0.02
Heart failure 0.86 (0.37-2.01) 0.73 0.83 (0.31-2.19) 0.70
Diabetes 2.01 (1.11-3.65) 0.021 1.85 (0.85-4.01) 0.12
Kidney disease 1.54 (0.88-2.69) 0.13 1.42 (0.69-2.93) 0.34
Anaemia 1.79 (0.89-3.58) 0.10 1.24 (0.47-3.30) 0.67
TTR (continuous) 0.99 (0.97-1.01) 0.20 0.98 (0.96-1.01) 0.15
TTR <70% 1.56 (0.88-2.79) 0.13 -
PINRR (continuous) 0.98 (0.96-1.01) 0.19 -
PINRR <70% 1.17 (0.52-2.60) 0.70 -
¥
Continuous variable
304
Table A4.5: Cox- proportional hazard regression analysis for all bleeding event, including major bleed and clinically
relevant non-major bleed
Univariate HR (95% CI) p-value Multivariate HR p-value Multivariate HR p-value

(95% CI) (Model 1)* (95% CI) (Model 2)#
Age¥ 1.03 (1.01-1.06) 0.016 1.02 (0.99-1.06) 0.14 1.02 (0.99-1.06) 0.13
Female sex 0.92 (0.58-1.45) 0.71 0.85 (0.48-1.50) 0.58 0.82 (0.47-1.44) 0.49
Smoking history 0.86 (0.51-1.45) 0.57 0.91 (0.52-1.61) 0.74 0.90 (0.51-1.59) 0.71
Ethnicity (non- white) 1.13 (0.64-2.02) 0.67 0.96 (0.47-1.96) 0.90 0.97 (0.48-1.96) 0.93
Hypertension 1.31 (0.73-2.33) 0.37 1.53 (0.71-3.31) 0.28 1.49 (0.69-3.21) 0.31
Stroke/TIA history 1.10 (0.63-1.93) 0.74 1.09 (0.56-2.12) 0.80 1.10 (0.57-2.15) 0.77
Heart failure 1.38 (0.78-2.47) 0.27 1.68 (0.89-3.19) 0.11 1.71 (0.90-3.22) 0.10
Diabetes 1.18 (0.69-2.02) 0.54 0.95 (0.48-1.86) 0.87 0.93 (0.47-1.83) 0.84
Vascular disease 0.97 (0.51-1.83) 0.92 0.83 (0.38-1.81) 0.63 0.78 (0.36-1.71) 0.53
Kidney disease 0.92 (0.57-1.47) 0.72 0.63 (0.35-1.14) 0.12 0.63 (0.35-1.13) 0.12
Anaemia 1.46 (0.80-2.65) 0.22 1.64 (0.77-3.52) 0.20 1.61 (0.75-3.46) 0.22
Bleeding history 1.48 (0.68-3.23) 0.32 1.05 (0.40-2.79) 0.92 1.04 (0.39-2.75) 0.94
TTR (continuous) 0.98 (0.97-0.998) 0.026 0.98 (0.97-1.00) 0.08 -
TTR <70% 1.52 (0.95-2.42) 0.08 - 1.78 (1.01-3.13) 0.05
PINRR (continuous) 0.97 (0.95-0.99) 0.003 - -
PINRR <70% 2.03 (0.93-4.42) 0.07 - -
¥
continuous variable
*Model 1: excluding PINRR continuous, TTR category and PINRR category, #Model 2: excluding TTR continuous, TTR category, PINRR category
305
Table A4.6: Cox- proportional hazard regression analysis for cardiovascular hospitalisation
Univariate HR p-value Multivariate* HR p-value Multivariate# HR p-value

(95% CI) (95% CI) (Model 1) (95% CI) (Model 2)
¥
Age 0.99 (0.98-1.01) 0.41 0.99 (0.97-1.01) 0.23 0.99 (0.97-1.01) 0.23
Female sex 0.97 (0.74-1.26) 0.81 1.04 (0.76-1.44) 0.80 1.03 (0.75-1.41) 0.87
Smoking history 0.88 (0.66-1.18) 0.40 0.88 (0.64-1.22) 0.45 0.89 (0.64-1.23) 0.48
Hypertension 1.31 (0.93-1.85) 0.12 0.95 (0.65-1.38) 0.77 0.93 (0.64-1.35) 0.68
Heart failure 1.76 (1.27-2.42) 0.001 1.45 (1.00-2.09) 0.05 1.46 (1.02-2.11) 0.04
Diabetes 1.37 (1.01-1.86) 0.042 1.15 (0.80-1.65) 0.45 1.14 (0.79-1.63) 0.49
Vascular disease 2.17 (1.61-2.93) <0.001 1.64 (1.12-2.38) 0.01 1.62 (1.11-2.34) 0.01
Kidney disease 1.05 (0.80-1.38) 0.71 0.93 (0.67-1.28) 0.64 0.94 (0.68-1.29) 0.68
Anaemia 1.64 (1.17-2.30) 0.004 1.26 (0.80-1.96) 0.32 1.23 (0.79-1.96) 0.36
TTR (continuous) 0.98 (0.97-0.99) <0.001 0.99 (0.98-0.99) 0.01 -
TTR <70% 1.61 (1.23-2.12) 0.001 - 1.38 (1.00-1.89) 0.05
PINRR (continuous) 0.97 (0.96-0.98) <0.001 - -
PINRR <70% 2.31 (1.42-3.74) 0.001 - -
*TTR continuous; # TTR categorical
¥
continuous variable
306
Table A4.7: Cox proportional hazard regression analysis for all-cause mortality
Univariate HR (95% CI) p-value Multivariate HR (95% CI) p-value

Age¥ 1.09 (1.03-1.14) 0.001 1.06 (0.99-1.13) 0.07
Female sex 0.66 (0.27-1.61) 0.36 0.86 (0.26-2.86) 0.81
Smoking history 2.13 (0.73-6.24) 0.17 2.87 (0.86-9.52) 0.09
Ethnicity (non- white) 0.78 (0.23-2.63) 0.69 1.19 (0.25- 5.79) 0.83
Hypertension 1.49 (0.51-4.39) 0.47 3.05 (0.39- 24.06) 0.29
Heart failure 0.94 (0.28-3.18) 0.93 0.99 (0.21-4.59) 0.99
Diabetes 0.94 (0.32-2.76) 0.91 0.65 (0.13-3.21) 0.59
Kidney disease 1.30 (0.56-3.02) 0.54 1.43 (0.48-4.21) 0.52
Anaemia 2.86 (1.13-7.27) 0.027 3.39 (0.92-12.52) 0.07
TTR (continuous) 0.98 (0.95-1.01) 0.13 0.98 (0.94-1.01) 0.20
TTR <70% 1.79 (0.73-4.41) 0.21 -
PINRR (continuous) 0.97 (0.93-1.01) 0.13 -
PINRR <70% 1.39 (0.41-4.69) 0.60 -
¥
continuous variable
307
Table A4.8: Cox proportional hazard regression analysis for composite outcome of thromboembolic events, major bleed
and clinically relevant non-major bleeding, cardiovascular hospitalisation and all-cause mortality (incorporating TTR as
continuous and categorical variables separately)
Univariate HR (95% CI) p-value Multivariate* HR p-value Multivariate† HR p-

(95% CI) (Model 1) (95% CI) (Model 2) value
Age¥ 1.0 (0.99-1.01) 0.91 0.99 (0.98-1.00) 0.17 0.99 (0.98-1.00) 0.20
Female sex 0.92 (0.74-1.15) 0.48 0.93 (0.71-1.22) 0.62 0.91 (0.70-1.20) 0.51
Smoking history 0.81 (0.63-1.03) 0.09 0.79 (0.60-1.04) 0.09 0.79 (0.60-1.04) 0.09
Hypertension 1.36 (1.02-1.81) 0.035 1.09 (0.78-1.52) 0.61 1.06 (0.76-1.48) 0.72
Heart failure 1.51 (1.14-1.99) 0.004 1.28 (0.93-1.76) 0.13 1.30 (0.94-1.78) 0.11
Diabetes 1.41 (1.10-1.82) 0.008 1.24 (0.92-1.67) 0.16 1.22 (0.91-1.65) 0.19
Vascular disease 1.93 (1.49-2.50) <0.001 1.69 (1.23-2.33) 0.001 1.67 (1.21-2.30) 0.002
Kidney disease 1.18 (0.95-1.48) 0.14 0.99 (0.76-1.30) 0.96 1.00 (0.76-1.31) 0.99
Anaemia 1.69 (1.27-2.24) <0.001 1.32 (0.91-1.91) 0.14 1.28 (0.88-1.85) 0.20
TTR (continuous) 0.98 (0.98-0.99) <0.001 0.99 (0.98-0.99) 0.001 -
TTR <70% 1.54 (1.23-1.92) <0.001 - 1.45 (1.11-1.89) 0.006
PINRR (continuous) 0.97 (0.96-0.98) <0.001 - -
PINRR <70% 1.99 (1.37-2.88) <0.001 - -
*This model considers TTR as continuous variable; † This model considers TTR as categorical variable
¥
continuous variable
308
Table A4.9: Cox proportional hazard regression analysis for all bleeding events, including major bleeding and clinically
relevant non-major bleeding including age ≥80 years
Univariate Multivariate
HR (95% CI) p-value HR (95% CI) p-value
Age ≥80 years 1.93 (1.16-3.20) 0.01 1.90 (1.01-3.56) 0.047
Female sex 0.92 (0.58-1.45) 0.71 0.82 (0.47-1.44) 0.49
Smoking history 0.86 (0.51-1.45) 0.57 0.91 (0.51-1.61) 0.75
Ethnicity (non- white) 1.13 (0.64-2.02) 0.67 1.01 (0.49-2.05 0.98
Hypertension 1.31 (0.73-2.33) 0.37 1.54 (0.72-3.31) 0.27
Heart failure 1.38 (0.78-2.47) 0.27 1.71 (0.91-3.24) 0.10
Diabetes 1.18 (0.69-2.02) 0.54 0.93 (0.47-1.84) 0.84
Kidney disease 0.92 (0.57-1.47) 0.73 0.65 (0.36-1.17) 0.15
Anaemia 1.46 (0.80-2.65) 0.22 1.67 (0.78-3.61) 0.19
Concomitant antiplatelet therapy 1.47 (0.54-4.03) 0.45 0.92 (0.21-44.06) 0.91
TTR <70% 1.52 (0.95-2.42) 0.08 1.74 (0.99-3.05) 0.055
309
Table A4.10: Cox- proportional hazard regression analysis for composite endpoints of thromboembolic event, bleeding
event, cardiovascular hospitalisation and all-cause mortality including age ≥80 years
Univariate HR (95% CI) p-value Multivariate HR (95% CI) p-value

Age ≥80 1.10 (0.84-1.45) 0.50 1.00 (0.72-1.39) 0.99
Female sex 0.92 (0.74-1.15) 0.48 0.90 (0.69-1.18) 0.45
Smoking history 0.81 (0.63-1.03) 0.09 0.82 (0.62-1.07) 0.14
Ethnicity (non- white) 1.47 (1.13-1.91) 0.004 1.22 (0.89-1.68) 0.22
Hypertension 1.36 (1.02-1.81) 0.04 1.05 (0.75-1.46) 0.78
Heart failure 1.51 (1.14-1.99) 0.004 1.32 (0.96-1.82) 0.09
Diabetes 1.41 (1.10-1.82) 0.008 1.22 (0.91-1.65) 0.19
Vascular disease 1.93 (1.49-2.50) <0.001 1.53 (1.10-2.14) 0.01
Kidney disease 1.18 (0.95-1.48) 0.14 0.96 (0.73-1.25) 0.75
Anaemia 1.69 (1.27-2.24) <0.001 1.27 (0.87-1.83) 0.21
Concomitant antiplatelet therapy 2.50 (1.67-3.73) <0.001 1.36 (0.81-2.30) 0.24
TTR <70% 1.54 (1.22-1.92) <0.001 1.47 (1.13-1.91) 0.004
310
Appendix 5
List of ethical approvals:
Study 1: TREAT-2 study
1. South Birmingham Research Ethics Committee (REC)
2. Health Research Authority (HRA)
3. Confirmation of capacity and capability from Sandwell and West Birmingham
Hospitals Research and Development (R&D) department
4. University Hospitals Birmingham (UHB) Research and Development (R&D)
department
Study 2 and 3
1. Institutional review from Sandwell and West Birmingham Hospitals Research

and Development (R&D) department
311
References
1. Zulkifly H, Lip GYH, Lane DA. Epidemiology of atrial fibrillation. Int J Clin Pract.
2018;72(3):e13070.
2. Rahman F, Kwan GF, Benjamin EJ. Global epidemiology of atrial fibrillation. Nat Rev
Cardiol. 2014;11(11):639-54.
3. Kirchhof P, Benussi S, Kotecha D, Ahlsson A, Atar D, Casadei B, et al. 2016 ESC
Guidelines for the management of atrial fibrillation developed in collaboration with EACTS.
Eur Heart J. 2016;37(38):2893-962.
4. Chugh SS, Havmoeller R, Narayanan K, Singh D, Rienstra M, Benjamin EJ, et al.
Worldwide Epidemiology of Atrial Fibrillation: A Global Burden of Disease 2010 Study.
Circulation. 2013;129(8):837–47.
5. Pistoia F, Sacco S, Tiseo C, Degan D, Ornello R, Carolei A. The Epidemiology of Atrial
Fibrillation and Stroke. Cardiol Clin. 2016;34(2):255-68.
6. Stefansdottir H, Aspelund T, Gudnason V, Arnar DO. Trends in the incidence and
prevalence of atrial fibrillation in Iceland and future projections. EP Europace.
2011;13(8):1110-7.
7. Lane DA, Skjøth F, Lip GYH, Larsen TB, Kotecha D. Temporal Trends in Incidence,
Prevalence, and Mortality of Atrial Fibrillation in Primary Care. J Am Heart Assoc. 2017;6(5).
8. Colilla S, Crow A, Petkun W, Singer DE, Simon T, Liu X. Estimates of Current and
Future Incidence and Prevalence of Atrial Fibrillation in the U.S. Adult Population. Am J
Cardiol. 2013;112(8):1142-7.
9. Shavadia J, Yonga G, Mwanzi S, Jinah A, Moriasi A, Otieno H. Clinical characteristics
and outcomes of atrial fibrillation and flutter at the Aga Khan University Hospital, Nairobi.
Cardiovasc J Afr. 2013;24(2):6-9.
10. Bai Y, Wang YL, Shantsila A, Lip GYH. The Global Burden of Atrial Fibrillation and
Stroke: A Systematic Review of the Clinical Epidemiology of Atrial Fibrillation in Asia. Chest.
2017;152(4):810–20.
11. Kang S-H, Choi E-K, Han K-D, Lee S-R, Lim W-H, Cha M-J, et al. Underweight is a
risk factor for atrial fibrillation: A nationwide population-based study. Int J Cardiol.
2016;215:449-56.
12. Li L-H, Sheng C-S, Hu B-C, Huang Q-F, Zeng W-F, Li G-L, et al. The prevalence,
incidence, management and risks of atrial fibrillation in an elderly Chinese population: a
prospective study. BMC Cardiovasc Disord. 2015;15:31.
330
13. Guo Y, Tian Y, Wang H, Si Q, Wang Y, Lip GYH. Prevalence, incidence, and lifetime
risk of atrial fibrillation in China: new insights into the global burden of atrial fibrillation. Chest.
2015;147(1):109-19.
14. Sano F, Ohira T, Kitamura A, Imano H, Cui R, Kiyama M, et al. Heavy Alcohol
Consumption and Risk of Atrial Fibrillation: The Circulatory Risk in Communities Study
(CIRCS). Circ J. 2014;78(4):955-61.
15. Chuang SY, Wu CC, Hsu PF, Chia-Yu Chen R, Liu WL, Hsu YY, et al. Hyperuricemia
and incident atrial fibrillation in a normotensive elderly population in Taiwan. Nutr Metab
Cardiovasc Dis. 2014;24(9):1020-6.
16. Suzuki S, Sagara K, Otsuka T, Kano H, Matsuno S, Takai H, et al. Usefulness of
Frequent Supraventricular Extrasystoles and a High CHADS2 Score to Predict First-Time
Appearance of Atrial Fibrillation. Am J Cardiol. 2013;111(11):1602-7.
17. Chao T-F, Hung C-L, Chen S-J, Wang K-L, Chen T-J, Lin Y-J, et al. The association
between hyperuricemia, left atrial size and new-onset atrial fibrillation. Int J Cardiol.
2013;168(4):4027-32.
18. Rhee CW, Lee J, Oh S, Choi NK, Park BJ. Use of bisphosphonate and risk of atrial
fibrillation in older women with osteoporosis. Osteoporos Int. 2012;23(1):247-54.
19. Watanabe H, Tanabe N, Yagihara N, Watanabe T, Aizawa Y, Kodama M. Association
Between Lipid Profile and Risk of Atrial FibrillatioN: Niigata Preventive Medicine Study. Circ
J. 2011;75(12):2767-74.
20. Iguchi Y, Kimura K, Shibazaki K, Aoki J, Kobayashi K, Sakai K, et al. Annual Incidence
of Atrial Fibrillation and Related Factors in Adults. Am J Cardiol. 2010;106(8):1129-33.
21. Sliwa K, Carrington MJ, Klug E, Opie L, Lee G, Ball J, et al. Predisposing factors and
incidence of newly diagnosed atrial fibrillation in an urban African community: insights from
the Heart of Soweto Study. Heart. 2010;96(23):1878.
22. Kamel H, Kleindorfer DO, Bhave PD, Cushman M, Levitan EB, Howard G, et al. Rates
of Atrial Fibrillation in Black Versus White Patients With Pacemakers. J Am Heart Assoc.
2016;5(2).
23. Yap KB, Ng TP, Ong HY. Low prevalence of atrial fibrillation in community-dwelling
Chinese aged 55 years or older in Singapore: a population-based study. J Electrocardiol.
2008;41(2):94-8.
24. Phrommintikul A, Detnuntarat P, Prasertwitayakij N, Wongcharoen W. Prevalence of
atrial fibrillation in Thai elderly. J Geriatr Cardiol. 2016;13(3):270-3.
25. Ball J, Thompson DR, Ski CF, Carrington MJ, Gerber T, Stewart S. Estimating the
current and future prevalence of atrial fibrillation in the Australian adult population. Med J Aust.
2015;202(1):32-5.
331
26. Sturm JW, Davis SM, O'Sullivan JG, Vedadhaghi ME, Donnan GA. The Avoid Stroke
as Soon as Possible (ASAP) general practice stroke audit. Med J Aust. 2002;176(7):312-6.
27. Quality and Outcomes Framework-Prevalence, Achievements and Exceptions Report.
2014-2015. Health & Social Care Information Centre.
28. Naccarelli GV, Varker H, Lin J, Schulman KL. Increasing Prevalence of Atrial
Fibrillation and Flutter in the United States. Am J Cardiol. 2009;104(11):1534-9.
29. Marcolino MS, Palhares DMF, Benjamin EJ, Ribeiro AL. Atrial fibrillation: prevalence
in a large database of primary care patients in Brazil. EP Europace. 2015;17(12):1787-90.
30. Sriharibabu M, Himabindu Y, Kabir Z. Rheumatic heart disease in rural south India: A
clinico-observational study. J Cardiovasc Dis Res. 2013;4(1):25-9.
31. White E. Ethnicity and National Identity in England and Wales: 2011. Office for National
Statistics.
32. Nybo MS, Skov J. Patient knowledge of anticoagulant treatment does not correlate
with treatment quality. Public Health. 2016;141(Supplement C):17-22.
33. Lau C-P, Gbadebo TD, Connolly SJ, Van Gelder IC, Capucci A, Gold MR, et al. Ethnic
Differences in Atrial Fibrillation Identified Using Implanted Cardiac Devices. J Cardiovasc
Electrophysiol. 2013;24(4):381-7.
34. Lahiri MK, Fang K, Lamerato L, Khan AM, Schuger CD. Effect of Race on the
Frequency of Postoperative Atrial Fibrillation Following Coronary Artery Bypass Grafting. Am
J Cardiol. 2011;107(3):383-6.
35. Yuh-Jer Shen A, Contreras R, Sobnosky S, Shah AI, Ichiuji AM, Jorgensen MB, et al.
Racial/Ethnic Differences in the Prevalence of Atrial Fibrillation Among Older Adults—A
Cross-Sectional Study. J Natl Med Assoc. 2010;102(10):906-14.
36. Marcus GM, Alonso A, Peralta CA, Lettre G, Vittinghoff E, Lubitz SA, et al. European
ancestry as a risk factor for atrial fibrillation in African Americans. Circulation.
2010;122(20):2009-15.
37. Alonso A, Agarwal SK, Soliman EZ, Ambrose M, Chamberlain AM, Prineas RJ, et al.
Incidence of atrial fibrillation in whites and African-Americans: The Atherosclerosis Risk in
Communities (ARIC) study. Am Heart J. 2009;158(1):111-7.
38. Michael Smith J, Soneson EA, Woods SE, Engel AM, Hiratzka LF. Coronary artery
bypass graft surgery outcomes among African-Americans and Caucasian patients. Int J Surg.
2006;4(4):212-6.
39. Ruo B, Capra AM, Jensvold NG, Go AS. Racial variation in the prevalence of atrial
fibrillation among patients with heart failure. J Am Coll Cardiol. 2004;43(3):429-35.
40. Upshaw CB. Reduced prevalence of atrial fibrillation in black patients compared with
white patients attending an urban hospital: an electrocardiographic study. J Natl Med Assoc.
2002;94(4):204-8.
332
41. Go AS, Hylek EM, Phillips KA, et al. Prevalence of diagnosed atrial fibrillation in adults:
National implications for rhythm management and stroke prevention: the anticoagulation and
risk factors in atrial fibrillation (atria) study. JAMA. 2001;285(18):2370-5.
42. Afzal A, Ananthasubramaniam K, Sharma N, Ai-Malki Q, Ali AS, Jacobsen G, et al.
Racial differences in patients with heart failure. Clin Cardiol. 1999;22(12):791-4.
43. Winkelmayer WC, Patrick AR, Liu J, Brookhart MA, Setoguchi S. The increasing
prevalence of atrial fibrillation among hemodialysis patients. J Am Soc Nephrol.
2011;22(2):349-57.
44. Hernandez MB, Asher CR, Hernandez AV, Novaro GM. African American Race and
Prevalence of Atrial Fibrillation:A Meta-Analysis. Cardiol Res Pract. 2012;2012:275624.
45. Zubaid M, Rashed WA, Alsheikh-Ali AA, AlMahmeed W, Shehab A, Sulaiman K, et al.
Gulf Survey of Atrial Fibrillation Events (Gulf SAFE). Circ Cardiovasc Qual Outcomes.
2011;4(4):477.
46. Maru M. Atrial fibrillation and embolic complications. East Afr Med J. 1997;74(1):3-5.
47. Oldgren J, Healey JS, Ezekowitz M, Commerford P, Avezum A, Pais P, et al. Variations
in Cause and Management of Atrial Fibrillation in a Prospective Registry of 15 400 Emergency
Department Patients in 46 Countries. The RE-LY Atrial Fibrillation Registry. Circulation.
2014;129(15):1568.
48. Marini C, De Santis F, Sacco S, Russo T, Olivieri L, Totaro R, et al. Contribution of
Atrial Fibrillation to Incidence and Outcome of Ischemic Stroke. Stroke. 2005;36(6):1115.
49. Rizos T, Wagner A, Jenetzky E, Ringleb PA, Becker R, Hacke W, et al. Paroxysmal
Atrial Fibrillation Is More Prevalent than Persistent Atrial Fibrillation in Acute Stroke and
Transient Ischemic Attack Patients. Cerebrovasc Dis. 2011;32(3):276-82.
50. Hanchate AD, Schwamm LH, Huang W-J, Hylek E. Comparison of Ischemic Stroke
Outcomes and, Patient and Hospital Characteristics by Race/Ethnicity and Socioeconomic
Status. Stroke. 2013;44(2):469-76.
51. Gattellari M, Goumas C, Aitken R, Worthington JM. Outcomes for patients with
ischaemic stroke and atrial fibrillation: the PRISM study (A Program of Research Informing
Stroke Management). Cerebrovasc Dis. 2011;32(4):370-82.
52. Gao Q, Fu X, Wei JW, Chen X, Huang Y, Wang J, et al. Use of oral anticoagulation
among stroke patients with atrial fibrillation in China: the ChinaQUEST (Quality evaluation of
stroke care and treatment) registry study. Int J Stroke. 2013;8(3):150-4.
53. Thakkar S, Bagarhatta R. Detection of paroxysmal atrial fibrillation or flutter in patients
with acute ischemic stroke or transient ischemic attack by Holter monitoring. Indian Heart J.
2014;66(2):188-92.
54. Mahajan SK, Kashyap R, Sood BR, Jaret P, Mokta J, Kaushik NK, et al. Stroke at
moderate altitude. J Assoc Physicians India. 2004;52:699-702.
333
55. Tagawa M, Takeuchi S, Chinushi M, Saeki M, Taniguchi Y, Nakamura Y, et al.
Evaluating patients with acute ischemic stroke with special reference to newly developed atrial
fibrillation in cerebral embolism. Pacing Clin Electrophysiol. 2007;30(9):1121-8.
56. Sposato LA, Cipriano LE, Saposnik G, Vargas ER, Riccio PM, Hachinski V. Diagnosis
of atrial fibrillation after stroke and transient ischaemic attack: a systematic review and meta-
analysis. Lancet Neurol. 2015;14(4):377-87.
57. Mairesse G, Moran P, Van Gelder I, Elsner C, Rosenqvist M, Mant J, et al. Screening
for atrial fibrillation: a European Heart Rhythm Association (EHRA) consensus document
endorsed by the Heart Rhythm Society (HRS), Asia Pacific Heart Rhythm Society (APHRS),
and Sociedad Latinoamericana de Estimulacion Cardiaca y Electrofisiologia (SOLAECE).
Europace. 2017;0:1-35.
58. Fuster V, Rydén LE, Cannom DS, Crijns HJ, Curtis AB, Ellenbogen KA, et al.
ACC/AHA/ESC 2006 Guidelines for the Management of Patients With Atrial Fibrillation. J Am
Coll Cardiol. 2006;48(4):e149.
59. Cottrell C. Atrial fibrillation part 1: pathophysiology. Practice Nursing. 2012;23(1):16-
21.
60. Di Minno MN, Ambrosino P, Dello Russo A, Casella M, Tremoli E, Tondo C.
Prevalence of left atrial thrombus in patients with non-valvular atrial fibrillation. A systematic
review and meta-analysis of the literature. Thromb Haemost. 2016;115(3):663-77.
61. Christensen LM, Krieger DW, Hojberg S, Pedersen OD, Karlsen FM, Jacobsen MD, et
al. Paroxysmal atrial fibrillation occurs often in cryptogenic ischaemic stroke. Final results from
the SURPRISE study. Eur J Neurol. 2014;21(6):884-9.
62. Glotzer TV, Daoud EG, Wyse DG, Singer DE, Ezekowitz MD, Hilker C, et al. The
relationship between daily atrial tachyarrhythmia burden from implantable device diagnostics
and stroke risk: the TRENDS study. Circ Arrhythm Electrophysiol. 2009;2(5):474-80.
63. Vanassche T, Lauw MN, Eikelboom JW, Healey JS, Hart RG, Alings M, et al. Risk of
ischaemic stroke according to pattern of atrial fibrillation: analysis of 6563 aspirin-treated
patients in ACTIVE-A and AVERROES. Eur Heart J. 2014;36(5):281-8.
64. Steinberg BA, Hellkamp AS, Lokhnygina Y, Patel MR, Breithardt G, Hankey GJ, et al.
Higher risk of death and stroke in patients with persistent vs. paroxysmal atrial fibrillation:
results from the ROCKET-AF Trial. Eur Heart J. 2014;36(5):288-96.
65. Friberg L, Hammar N, Rosenqvist M. Stroke in paroxysmal atrial fibrillation: report from
the Stockholm Cohort of Atrial Fibrillation. Eur Heart J. 2009;31(8):967-75.
66. Hobbs FD, Fitzmaurice DA, Mant J, Murray E, Jowett S, Bryan S, et al. A randomised
controlled trial and cost-effectiveness study of systematic screening (targeted and total
population screening) versus routine practice for the detection of atrial fibrillation in people
aged 65 and over. The SAFE study. Health Technol Assess. 2005;9(40):iii-iv, ix-x, 1-74.
334
67. Lowres N, Neubeck L, Redfern J, Freedman SB. Screening to identify unknown atrial
fibrillation. A systematic review. Thromb Haemost. 2013;110(2):213-22.
68. Cooke G, Doust J, Sanders S. Is pulse palpation helpful in detecting atrial fibrillation?
A systematic review. J Fam Pract. 2006;55(2):130-4.
69. Freedman B, Camm J, Calkins H, Healey JS, Rosenqvist M, Wang J, et al. Screening
for Atrial Fibrillation. Circulation. 2017;135(19):1851-67.
70. Sanna T, Diener HC, Passman RS, Di Lazzaro V, Bernstein RA, Morillo CA, et al.
Cryptogenic stroke and underlying atrial fibrillation. N Engl J Med. 2014;370(26):2478-86.
71. Ott A, Breteler MM, de Bruyne MC, van Harskamp F, Grobbee DE, Hofman A. Atrial
fibrillation and dementia in a population-based study. Stroke. 1997;28(2):316-21.
72. Gladstone DJ, Spring M, Dorian P, Panzov V, Thorpe KE, Hall J, et al. Atrial fibrillation
in patients with cryptogenic stroke. N Engl J Med. 2014;370(26):2467-77.
73. Stollberger C, Chnupa P, Abzieher C, Langer T, Finsterer J, Klem I, et al. Mortality and
rate of stroke or embolism in atrial fibrillation during long-term follow-up in the embolism in left
atrial thrombi (ELAT) study. Clin Cardiol. 2004;27(1):40-6.
74. Cabin HS, Clubb Ks Fau - Hall C, Hall C Fau - Perlmutter RA, Perlmutter Ra Fau -
Feinstein AR, Feinstein AR. Risk for systemic embolization of atrial fibrillation without mitral
stenosis. Am J Cardiol. 1990;1(65 (16)):1112-6.
75. Ganesan AN, Chew DP, Hartshorne T, Selvanayagam JB, Aylward PE, Sanders P, et
al. The impact of atrial fibrillation type on the risk of thromboembolism, mortality, and bleeding:
a systematic review and meta-analysis. Eur Heart J. 2016;37(20):1591-602.
76. Benjamin EJ, Levy D, Vaziri SM, D'Agostino RB, Belanger AJ, Wolf PA. Independent
risk factors for atrial fibrillation in a population-based cohort: The framingham heart study.
JAMA. 1994;271(11):840-4.
77. Schnabel RB, Yin X, Gona P, Larson MG, Beiser AS, McManus DD, et al. 50 year
trends in atrial fibrillation prevalence, incidence, risk factors, and mortality in the Framingham
Heart Study: a cohort study. The Lancet. 2015;386(9989):154-62.
78. Selmer C, Olesen JB, Hansen ML, Lindhardsen J, Olsen AM, Madsen JC, et al. The
spectrum of thyroid disease and risk of new onset atrial fibrillation: a large population cohort
study. BMJ. 2012;345:e7895.
79. Buch P, Friberg J, Scharling H, Lange P, Prescott E. Reduced lung function and risk
of atrial fibrillation in the Copenhagen City Heart Study. Eur Respir J. 2003;21(6):1012-6.
80. Baber U, Howard VJ, Halperin JL, Soliman EZ, Zhang X, McClellan W, et al.
Association of Chronic Kidney Disease With Atrial Fibrillation Among Adults in the United
States: REasons for Geographic and Racial Differences in Stroke (REGARDS) Study. Circ
Arrhythm Electrophysiol. 2011;4(1):26.
335
81. Chamberlain AM, Agarwal SK, Folsom AR, Duval S, Soliman EZ, Ambrose M, et al.
Smoking and incidence of atrial fibrillation: results from the Atherosclerosis Risk in
Communities (ARIC) study. Heart Rhythm. 2011;8(8):1160-6.
82. Larsson SC, Drca N, Wolk A. Alcohol consumption and risk of atrial fibrillation: a
prospective study and dose-response meta-analysis. J Am Coll Cardiol. 2014;64(3):281-9.
83. Al-Khatib SM, Allen LaPointe NM, Chatterjee R, Crowley MJ, Dupre ME, Kong DF, et
al. Rate- and rhythm-control therapies in patients with atrial fibrillation: a systematic review.
Ann Intern Med. 2014;160(11):760-73.
84. Nikolaidou T, Channer KS. Chronic atrial fibrillation: a systematic review of medical
heart rate control management. Postgrad Med J. 2009;85(1004):303-12.
85. Tamariz LJ, Bass EB. Pharmacological rate control of atrial fibrillation. Cardiol Clin.
2004;22(1):35-45.
86. Segal JB, McNamara RL, Miller MR, Kim N, Goodman SN, Powe NR, et al. The
evidence regarding the drugs used for ventricular rate control. J Fam Pract. 2000;49(1):47-59.
87. Chatterjee S, Sardar P, Lichstein E, Mukherjee D, Aikat S. Pharmacologic rate versus
rhythm-control strategies in atrial fibrillation: an updated comprehensive review and meta-
analysis. Pacing Clin Electrophysiol. 2013;36(1):122-33.
88. Roy D, Talajic M, Nattel S, Wyse DG, Dorian P, Lee KL, et al. Rhythm control versus
rate control for atrial fibrillation and heart failure. N Engl J Med. 2008;358(25):2667-77.
89. de Denus S, Sanoski CA, Carlsson J, Opolski G, Spinler SA. Rate vs rhythm control
in patients with atrial fibrillation: a meta-analysis. Arch Intern Med. 2005;165(3):258-62.
90. Lip GYH. The ABC pathway: an integrated approach to improve AF management. Nat
Rev Cardiol. 2017;14(11):627-8.
91. Olesen JB, Fauchier L, Lane DA, Taillandier S, Lip GYH. Risk Factors for Stroke and
Thromboembolism in Relation to Age Among Patients With Atrial Fibrillation. Chest.
2012;141(1):147-53.
92. Echocardiographic predictors of stroke in patients with atrial fibrillation: a prospective
study of 1066 patients from 3 clinical trials. Arch Intern Med. 1998;158(12):1316-20.
93. Patel D, Mc Conkey ND, Sohaney R, Mc Neil A, Jedrzejczyk A, Armaganijan L. A
systematic review of depression and anxiety in patients with atrial fibrillation: the mind-heart
link. Cardiovasc Psychiatry Neurol. 2013;2013.
94. Polikandrioti M, Koutelekos I, Vasilopoulos G, Gerogianni G, Gourni M, Zyga S, et al.
Anxiety and Depression in Patients with Permanent Atrial Fibrillation: Prevalence and
Associated Factors. Cardiol Res Pract. 2018;2018.
95. Dąbrowski R, Smolis-Bąk E, Kowalik I, Kazimierska B, Wójcicka M, Szwed H. Quality
of life and depression in patients with different patterns of atrial fibrillation. Kardiol Pol.
2010;68(10):1133-9.
336
96. Thrall G, Lip GYH, Carroll D, Lane D. Depression, Anxiety, and Quality of Life in
Patients With Atrial Fibrillation. Chest. 2007;132(4):1259-64.
97. Frasure-Smith N, Lespérance F, Habra M, Talajic M, Khairy P, Dorian P, et al. Elevated
depression symptoms predict long-term cardiovascular mortality in patients with atrial
fibrillation and heart failure. Circulation. 2009;120(2):134-40.
98. Ong L, Irvine J, Nolan R, Cribbie R, Harris L, Newman D, et al. Gender differences and
quality of life in atrial fibrillation: the mediating role of depression. J Psychosom Res.
2006;61(6):769-74.
99. Lane DA, Langman CM, Lip GYH, Nouwen A. Illness perceptions, affective response,
and health-related quality of life in patients with atrial fibrillation. J Psychosom Res.
2009;66(3):203-10.
100. Wożakowska-Kapłon B, Opolski G, Kosior D, Jaskulska-Niedziela E, Maroszyńska-
Dmoch E, Włosowicz M. Cognitive disorders in elderly patients with permanent atrial
fibrillation. Kardiol Pol. 2009;67(5):487-93.
101. Chatap G, Giraud K, Vincent J-P. Atrial fibrillation in the elderly. Drugs Aging.
2002;19(11):819-46.
102. Prasun MA. Providing best practice in the management of atrial fibrillation in the United
States. J Cardiovasc Nurs. 2012;27(5):445-56.
103. McCabe PJ. What patients want and need to know about atrial fibrillation. J Multidiscip
Healthc. 2011;4:413-9.
104. Trovato G, Pace P, Cangemi E, Martines G, Trovato F, Catalano D. Gender, lifestyles,
illness perception and stress in stable atrial fibrillation. La Clinica Terapeutica.
2012;163(4):281-6.
105. Tailachidis P, Tsimtsiou Z, Galanis P, Theodorou M, Kouvelas D, Athanasakis K. The
Atrial Fibrillation Effect on QualiTy-of-Life (AFEQT) questionnaire: cultural adaptation and
validation of the Greek version. Hippokratia. 2016;20(4):264-7.
106. von Eisenhart Rothe A, Hutt F, Baumert J, Breithardt G, Goette A, Kirchhof P, et al.
Depressed mood amplifies heart-related symptoms in persistent and paroxysmal atrial
fibrillation patients: a longitudinal analysis--data from the German Competence Network on
Atrial Fibrillation. Europace. 2015;17(9):1354-62.
107. Thompson TS, Barksdale DJ, Sears SF, Mounsey JP, Pursell I, Gehi AK. The effect of
anxiety and depression on symptoms attributed to atrial fibrillation. Pacing Clin Electrophysiol.
2014;37(4):439-46.
108. McCabe PJ. Psychological distress in patients diagnosed with atrial fibrillation: the
state of the science. J Cardiovasc Nurs. 2010;25(1):40-51.
337
109. de Queiroz Almeida G, de ACB Noblat L, Passos LCS, do Nascimento HF. Quality of
life analysis of patients in chronic use of oral anticoagulant: an observational study. Health
Qual Life Outcomes. 2011;9(1):91.
110. Ynsaurriaga FA, Peinado RP, Ormaetxe Merodio JM. Atrial fibrillation and quality of
life related to disease and treatment: focus on anticoagulation. Future Cardiol.
2014;10(3):381-93.
111. Diana M-P, Consuelo R-A, Paz A-M, Pia L-J. Analysis of General and Oral Quality of
Life and Satisfaction with Treatment among Anticoagulated Patients. Int J Dent Hyg.
2015;1(5):11-7.
112. Corbi IS, Dantas RA, Pelegrino FM, Carvalho AR. Health related quality of life of
patients undergoing oral anticoagulation therapy. Rev Lat Am Enfermagem. 2011;19(4):865-
73.
113. Casais P, Meschengieser S, Sanchez-Luceros A, Lazzari M. Patients’ perceptions
regarding oral anticoagulation therapy and its effect on quality of life. Curr Med Res Opin.
2005;21(7):1085-90.
114. Esra Yildiz PhD RN, Nuray Dayapoglu PhD RN. The Satisfaction Levels of Patients
Using Anticoagulants. Int J Car Sci. 2017;10(1):568.
115. Matalqah LM, Radaideh K. Health-Related Quality of Life among Atrial fibrillation
Patients Undergoing Anticoagulation Therapy. Epidemiol Biostat Public Health. 2018;15(1).
116. Balcı KG, Balcı MM, Canpolat U, Şen F, Akboğa MK, Süleymanoğlu M, et al.
Comparison of health-related quality of life among patients using novel oral anticoagulants or
warfarin for non-valvular atrial fibrillation. Anatol J Cardiol. 2016;16(7):474.
117. Hart RG, Pearce LA, Aguilar MI. Meta-analysis: Antithrombotic therapy to prevent
stroke in patients who have nonvalvular atrial fibrillation. Ann Intern Med. 2007;146(12):857-
67.
118. Wallentin L, Lopes RD, Hanna M, Thomas L, Hellkamp A, Nepal S, et al. Efficacy and
safety of apixaban compared with warfarin at different levels of predicted international
normalized ratio control for stroke prevention in atrial fibrillation. Circulation.
2013;127(22):2166-76.
119. Gallagher AM, Setakis E, Plumb JM, Clemens A, van Staa T-P. Risks of stroke and
mortality associated with suboptimal anticoagulation in atrial fibrillation patients. Thromb
Haemost. 2011;106(5):968.
120. Wan Y, Heneghan C, Perera R, Roberts N, Hollowell J, Glasziou P, et al.
Anticoagulation Control and Prediction of Adverse Events in Patients With Atrial Fibrillation.
Circ Cardiovasc Qual Outcomes. 2008;1(2):84-91.
121. Lip GYH, Al-Saady N, Jin J, Sun M, Melino M, Winters SM, et al. Anticoagulation
Control in Warfarin-Treated Patients Undergoing Cardioversion of Atrial Fibrillation (from the
338
Edoxaban Versus Enoxaparin–Warfarin in Patients Undergoing Cardioversion of Atrial
Fibrillation Trial). Am J Cardiol. 2017;120(5):792-6.
122. Morgan CL, McEwan P, Tukiendorf A, Robinson PA, Clemens A, Plumb JM. Warfarin
treatment in patients with atrial fibrillation: Observing outcomes associated with varying levels
of INR control. Thromb Res. 2009;124(1):37-41.
123. Petersen P, Godtfredsen J, Boysen G, Andersen E, Andersen Br. Placebo-controlled,
randomised trial of warfarin and aspirin for prevention of thromboembolic complications in
chronic atrial fibrillation: the Copenhagen AFASAK study. Lancet. 1989;333(8631):175-9.
124. Peterson P, Boysen G. Prevention of stroke in atrial fibrillation. N Engl J Med.
1990;323(482).
125. The Effect of Low-Dose Warfarin on the Risk of Stroke in Patients with Nonrheumatic
Atrial Fibrillation. N Engl J Med. 1990;323(22):1505-11.
126. Stroke prevention in atrial fibrillation study: final results. Circulation. 1991;84(2):527-
39.
127. Connolly SJ, Laupacis A, Gent M, Roberts RS, Cairns JA, Joyner C. Canadian atrial
fibrillation anticoaguiation (CAFA) study. J Am Coll Cardiol. 1991;18(2):349-55.
128. Ezekowitz MD, Bridgers SL, James KE, Carliner NH, Colling CL, Gornick CC, et al.
Warfarin in the prevention of stroke associated with nonrheumatic atrial fibrillation. N Engl J
Med. 1992;327(20):1406-12.
129. Secondary prevention in non-rheumatic atrial fibrillation after transient ischaemic
attack or minor stroke. EAFT (European Atrial Fibrillation Trial) Study Group. Lancet.
1993;342(8882):1255-62.
130. Ruff CT, Giugliano RP, Braunwald E, Hoffman EB, Deenadayalu N, Ezekowitz MD, et
al. Comparison of the efficacy and safety of new oral anticoagulants with warfarin in patients
with atrial fibrillation: a meta-analysis of randomised trials. Lancet. 2014;383(9921):955-62.
131. De Caterina R, Husted S, Wallentin L, Andreotti F, Arnesen H, Bachmann F, et al.
Vitamin K antagonists in heart disease: Current status and perspectives (Section III). Position
Paper of the ESC Working Group on Thrombosis – Task Force on Anticoagulants in Heart
Disease. Thromb Haemost. 2013;110(12):1087-107.
132. Connolly SJ, Ezekowitz MD, Yusuf S, Eikelboom J, Oldgren J, Parekh A, et al.
Dabigatran versus Warfarin in Patients with Atrial Fibrillation. N Engl J Med.
2009;361(12):1139-51.
133. Patel MR, Mahaffey KW, Garg J, Pan G, Singer DE, Hacke W, et al. Rivaroxaban
versus Warfarin in Nonvalvular Atrial Fibrillation. N Engl J Med. 2011;365(10):883-91.
134. Granger CB, Alexander JH, McMurray JJV, Lopes RD, Hylek EM, Hanna M, et al.
Apixaban versus Warfarin in Patients with Atrial Fibrillation. N Engl J Med. 2011;365(11):981-
92.
339
135. Giugliano RP, Ruff CT, Braunwald E, Murphy SA, Wiviott SD, Halperin JL, et al.
Edoxaban versus Warfarin in Patients with Atrial Fibrillation. N Engl J Med.
2013;369(22):2093-104.
136. Ntaios G, Papavasileiou V, Makaritsis K, Vemmos K, Michel P, Lip GY. Real-world
setting comparison of nonvitamin-K antagonist oral anticoagulants versus vitamin-K
antagonists for stroke prevention in atrial fibrillation: a systematic review and meta-analysis.
Stroke. 2017;48(9):2494-503.
137. Stroke Risk in Atrial Fibrillation Working Group. Independent predictors of stroke in
patients with atrial fibrillation: a systematic review. Neurology. 2007;69(6):546-54.
138. Pisters R, Lane DA, Marin F, Camm AJ, Lip GY. Stroke and thromboembolism in atrial
fibrillation. Circ J. 2012;76(10):2289-304.
139. Wagstaff A, Overvad TF, Lip G, Lane D. Is female sex a risk factor for stroke and
thromboembolism in patients with atrial fibrillation? A systematic review and meta-analysis.
QJM. 2014;107(12):955-67.
140. Mikkelsen A, Lindhardsen J, Lip G, Gislason G, Torp-Pedersen C, Olesen J. Female
sex as a risk factor for stroke in atrial fibrillation: a nationwide cohort study. J Thromb Haemost.
2012;10(9):1745-51.
141. Lip GH, Lane DA. Stroke prevention in atrial fibrillation: A systematic review. JAMA.
2015;313(19):1950-62.
142. Nielsen PB, Skjøth F, Overvad TF, Larsen TB, Lip GYH. Female Sex Is a Risk Modifier
Rather Than a Risk Factor for Stroke in Atrial Fibrillation. Circulation. 2018;137(8):832.
143. Lehto M, Snapinn S, Dickstein K, Swedberg K, Nieminen MS, investigators O.
Prognostic risk of atrial fibrillation in acute myocardial infarction complicated by left ventricular
dysfunction: the OPTIMAAL experience. Eur Heart J. 2004;26(4):350-6.
144. Olesen JB, Lip GYH, Hansen ML, Hansen PR, Tolstrup JS, Lindhardsen J, et al.
Validation of risk stratification schemes for predicting stroke and thromboembolism in patients
with atrial fibrillation: nationwide cohort study. BMJ. 2011;342:d124.
145. Banerjee A, Taillandier S, Olesen Jonas B, Lane Deirdre A, Lallemand B, Lip Gregory
YH, et al. Ejection fraction and outcomes in patients with atrial fibrillation and heart failure: the
Loire Valley Atrial Fibrillation Project. Eur J Heart Fail. 2014;14(3):295-301.
146. Lip GYH, Nieuwlaat R, Pisters R, Lane DA, Crijns HJGM. Refining clinical risk
stratification for predicting stroke and thromboembolism in atrial fibrillation using a novel risk
factor-based approach: The euro heart survey on atrial fibrillation. Chest. 2010;137(2):263-72.
147. Lip GYH, Frison L, Halperin JL, Lane DA. Identifying Patients at High Risk for Stroke
Despite Anticoagulation. Stroke. 2010;41(12):2731.
340
148. Van Staa TP, Setakis E, Di Tanna GL, Lane DA, Lip GYH. A comparison of risk
stratification schemes for stroke in 79 884 atrial fibrillation patients in general practice. J
Thromb Haemost. 2011;9(1):39-48.
149. Poli D, Lip Gregory YH, Antonucci E, Grifoni E, Lane D. Stroke Risk Stratification in a
“Real-World” Elderly Anticoagulated Atrial Fibrillation Population. J Cardiovasc Electrophysiol.
2011;22(1):25-30.
150. Siu C-W, Lip GY, Lam K-F, Tse H-F. Risk of stroke and intracranial hemorrhage in
9727 Chinese with atrial fibrillation in Hong Kong. Heart Rhythm. 2014;11(8):1401-8.
151. Guo Y, Apostolakis S, Blann AD, Wang H, Zhao X, Zhang Y, et al. Validation of
contemporary stroke and bleeding risk stratification scores in non-anticoagulated Chinese
patients with atrial fibrillation. Int J Cardiol. 2013;168(2):904-9.
152. Risk factors for stroke and efficacy of antithrombotic therapy in atrial fibrillation:
analysis of pooled data from five randomized controlled trials. Arch Intern Med.
1994;154:1449-57.
153. Hart RG, Pearce LA, McBride R, Rothbart RM, Asinger RW. Factors Associated With
Ischemic Stroke During Aspirin Therapy in Atrial Fibrillation: analysis of 2012 participants in
the SPAF I-III clinical trials. Stroke. 1999;30(6):1223.
154. Gage BF, Waterman AD, Shannon W, Boechler M, Rich MW, Radford MJ. Validation
of clinical classification schemes for predicting stroke: results from the National Registry of
Atrial Fibrillation. JAMA. 2001;285(22):2864-70.
155. Wang TJ, Massaro JM, Levy D, Vasan RS, Wolf PA, D'Agostino RB, et al. A risk score
for predicting stroke or death in individuals with new-onset atrial fibrillation in the community:
the Framingham Heart Study. JAMA. 2003;290(8):1049-56.
156. van Walraven C, Hart RG, Wells GA, Petersen P, Koudstaal PJ, Gullov AL, et al. A
clinical prediction rule to identify patients with atrial fibrillation and a low risk for stroke while
taking aspirin. Arch Intern Med. 2003;163(8):936-43.
157. Rietbrock S, Heeley E, Plumb J, van Staa T. Chronic atrial fibrillation: Incidence,
prevalence, and prediction of stroke using the Congestive heart failure, Hypertension, Age>
75, Diabetes mellitus, and prior Stroke or transient ischemic attack (CHADS2) risk stratification
scheme. Am Heart J. 2008;156(1):57-64.
158. Hippisley-Cox J, Coupland C, Brindle P. Derivation and validation of QStroke score for
predicting risk of ischaemic stroke in primary care and comparison with other risk scores: a
prospective open cohort study. BMJ. 2013;346:f2573.
159. Singer DE, Chang Y, Borowsky LH, Fang MC, Pomernacki NK, Udaltsova N, et al. A
new risk scheme to predict ischemic stroke and other thromboembolism in atrial fibrillation:
the ATRIA study stroke risk score. J Am Heart Assoc. 2013;2(3):e000250.
341
160. Japanese Circulation Society Joint Working Group. Guidelines for pharmacotherapy
of atrial fibrillation (JCS 2013). Circ J. 2014;78(8):1997-2021.
161. Verma A, Cairns JA, Mitchell LB, Macle L, Stiell IG, Gladstone D, et al. 2014 Focused
Update of the Canadian Cardiovascular Society Guidelines for the Management of Atrial
Fibrillation. Can J Cardiol. 2014;30(10):1114-30.
162. Hijazi Z, Oldgren J, Andersson U, Connolly SJ, Ezekowitz MD, Hohnloser SH, et al.
Cardiac Biomarkers are Associated with an Increased Risk of Stroke and Death in Patients
with Atrial Fibrillation: A RELY Substudy. Circulation. 2012;125(13):1605-16.
163. Conway DSG, Pearce LA, Chin BSP, Hart RG, Lip GYH. Prognostic Value of Plasma
von Willebrand Factor and Soluble P-Selectin as Indices of Endothelial Damage and Platelet
Activation in 994 Patients With Nonvalvular Atrial Fibrillation. Circulation. 2003;107(25):3141.
164. Lip GYH, Lane D, Van Walraven C, Hart RG. Additive Role of Plasma von Willebrand
Factor Levels to Clinical Factors for Risk Stratification of Patients With Atrial Fibrillation.
Stroke. 2006;37(9):2294.
165. Gladstone DJ, Bui E, Fang J, Laupacis A, Lindsay MP, Tu JV, et al. Potentially
Preventable Strokes in High-Risk Patients With Atrial Fibrillation Who Are Not Adequately
Anticoagulated. Stroke. 2009;40(1):235.
166. Ogilvie IM, Newton N, Welner SA, Cowell W, Lip GYH. Underuse of Oral
Anticoagulants in Atrial Fibrillation: A Systematic Review. Am J Med. 2010;123(7):638-45.e4.
167. Zhu W-G, Xiong Q-M, Hong K. Meta-analysis of CHADS2 versus CHA2DS2-VASc for
predicting stroke and thromboembolism in atrial fibrillation patients independent of
anticoagulation. Tex Heart Inst J. 2015;42(1):6-15.
168. Olesen JB, Torp-Pedersen C, Hansen ML, Lip GY. The value of the CHA2DS2-VASc
score for refining stroke risk stratification in patients with atrial fibrillation with a CHADS2 score
0–1: a nationwide cohort study. Thromb Haemost. 2012;108(06):1172-9.
169. Karthikeyan G, Eikelboom JW. The CHADS2 score for stroke risk stratification in atrial
fibrillation–friend or foe? Thromb Haemost. 2010;104(01):45-8.
170. January CT, Wann LS, Alpert JS, Calkins H, Cigarroa JE, Conti JB, et al. 2014
AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: a report of the
American College of Cardiology/American Heart Association Task Force on Practice
Guidelines and the Heart Rhythm Society. J Am Coll Cardiol. 2014;64(21):e1-e76.
171. Chiang CE, Okumura K, Zhang S, Chao TF, Siu CW, Wei Lim T, et al. 2017 consensus
of the Asia Pacific Heart Rhythm Society on stroke prevention in atrial fibrillation. J Arrhythm.
2017;33(4):345-67.
172. Macle L, Cairns J, Leblanc K, Tsang T, Skanes A, Cox JL, et al. 2016 Focused Update
of the Canadian Cardiovascular Society Guidelines for the Management of Atrial Fibrillation.
Can J Cardiol. 2016;32(10):1170-85.
342
173. Camm AJ, Lip GYH, De Caterina R, Savelieva I, Atar D, Hohnloser SH, et al. 2012
focused update of the ESC Guidelines for the management of atrial fibrillation: An update of
the 2010 ESC guideline for the management of atrial fibrillation. Eur Heart J.
2012;33(21):2719-47.
174. Ogawa S, Aonuma K, Tse HF, Huang D, Huang JL, Kalman J, et al. The APHRS's
2013 statement on antithrombotic therapy of patients with nonvalvular atrial fibrillation. J
Arrhythm. 2013;29(3):190-200.
175. National Institute for Clinical Excellence (2006) Atrial Fibrillation: Management. NICE
Guideline
176. Zulkifly H, Lip GYH, Lane DA. Bleeding Risk Scores in Atrial Fibrillation and Venous
Thromboembolism. Am J Cardiol. 2017;120(7):1139-45.
177. Hijazi Z, Oldgren J, Lindback J, Alexander JH, Connolly SJ, Eikelboom JW, et al. The
novel biomarker-based ABC (age, biomarkers, clinical history)-bleeding risk score for patients
with atrial fibrillation: a derivation and validation study. Lancet. 2016;387(10035):2302-11.
178. Brien ECO, Simon DN, Singer DE, Thomas LE, Hylek EM, Gersh B, et al. The ORBIT
Bleeding Score : A Simple Score to Assess Major Bleeding Risk in Atrial Fibrillation The
ORBIT-AF Registry. Eur Heart J. 2015;36(46):3258-64.
179. Fang MC, Chang Y, Borowsky LH, Pomernacki NK, Udaltsova N, Singer DE. A New
Risk Scheme to Predict Warfarin-Associated Hemorrhage: The ATRIA (Anticoagulation and
Risk Factors in Atrial Fibrillation) Study. J Am Coll Cardiol. 2011;58(4):395-401.
180. Pisters R, Lane DA, Nieuwlaat R, de Vos CB, Crijns HJGM, Lip GYH. A novel user-
friendly score (HAS-BLED) to assess 1-year risk of major bleeding in patients with atrial
fibrillation: the Euro Heart Survey. Chest. 2010;138(5):1093-100.
181. Gage BF, Yan Y, Milligan PE, Waterman AD, Culverhouse R, Rich MW, et al. Clinical
classification schemes for predicting hemorrhage: Results from the National Registry of Atrial
Fibrillation (NRAF). Am Heart J. 2006;151(3):713-9.
182. Shireman TI, Mahnken JD, Howard PA, Kresowik TF, Hou Q, Ellerbeck EF.
DEvelopment of a contemporary bleeding risk model for elderly warfarin recipients*. Chest.
2006;130(5):1390-6.
183. Sanders GD, Lowenstern A, Borre E, Chatterjee R, Goode A, Sharan L, et al. Stroke
Prevention in Patients With Atrial Fibrillation: A Systematic Review Update. Comparative
Effectiveness Review No. 214. (Prepared by the Duke Evidence-based Practice Center under
Contract No. 290-2015-00004-I for AHRQ and PCORI.) 2018.
184. Lowenstern A, Al-Khatib SM, Sharan L, Chatterjee R, Allen LaPointe NM, Shah B, et
al. Interventions for Preventing Thromboembolic Events in Patients With Atrial Fibrillation: A
Systematic Review. Ann Intern Med. 2018;169(11):774-87.
343
185. Borre ED, Goode A, Raitz G, Shah B, Lowenstern A, Chatterjee R, et al. Predicting
Thromboembolic and Bleeding Event Risk in Patients with Non-Valvular Atrial Fibrillation: A
Systematic Review. Thromb Haemost. 2018;118(12):2171-87.
186. Donzé J, Rodondi N, Waeber G, Monney P. Scores to Predict Major Bleeding Risk
During Oral Anticoagulation Therapy : A Prospective Validation Study. Am J Med.
2012;125(11):1095-102.
187. Landefeld CS, Goldman OL. Major bleeding in outpatients treated with warfarin:
incidence and prediction by factors known at the start of outpatient therapy. Am J Med.
1989;87(2):144-52.
188. Beyth RJ, Quinn LM, Landefeld CS. Prospective evaluation of an index for predicting
the risk of major bleeding in outpatients treated with warfarin ∗. Am J Med. 1998;105(2):91-9.
189. Zhu W, He W, Guo L, Wang X, Hong K. The HAS-BLED Score for Predicting Major
Bleeding Risk in Anticoagulated Patients With Atrial Fibrillation: A Systematic Review and
Meta-analysis. Clin Cardiol. 2015;38(9):555-61.
190. Senoo K, Proietti M, Lane DA, Lip GY. Evaluation of the HAS-BLED, ATRIA, and
ORBIT Bleeding Risk Scores in Patients with Atrial Fibrillation Taking Warfarin. Am J Med.
2016;129(6):600-7.
191. Proietti M, Senoo K, Lane DA, Lip GY. Major Bleeding in Patients with Non-Valvular
Atrial Fibrillation: Impact of Time in Therapeutic Range on Contemporary Bleeding Risk
Scores. Sci Rep. 2016;6:24376.
192. Apostolakis S, Lane DA, Guo Y, Buller H, Lip GYH. Performance of the HEMORR 2
HAGES , ATRIA , and HAS-BLED Bleeding Risk – Prediction Scores in Patients With Atrial
Fibrillation Undergoing Anticoagulation The AMADEUS ( Evaluating the Use of SR34006
Compared to Warfarin or Acenocoumarol in Patients With Atrial Fibrillation) Study. JAC.
2012;60(9):861-7.
193. Senoo K, Lip GY. Predictive abilities of the HAS-BLED and ORBIT bleeding risk scores
in non-warfarin anticoagulated atrial fibrillation patients: An ancillary analysis from the
AMADEUS trial. Int J Cardiol. 2016;221:379-82.
194. Gorman EW, Perkel D, Dennis D, Yates J, Heidel RE, Wortham D. Validation of the
HAS-BLED tool in atrial fibrillation patients receiving rivaroxaban. J Atr Fibrillation.
2016;9(2):16-8.
195. White HD, Gruber M, Feyzi J, et al. Comparison of outcomes among patients
randomized to warfarin therapy according to anticoagulant control: Results from sportif iii and
v. Arch Intern Med. 2007;167(3):239-45.
196. Hylek EM, Go AS, Chang Y, Jensvold NG, Henault LE, Selby JV, et al. Effect of
Intensity of Oral Anticoagulation on Stroke Severity and Mortality in Atrial Fibrillation. N Engl
J Med. 2003;349(11):1019-26.
344
197. Nelson WW, Desai S, Damaraju CV, Lu L, Fields LE, Wildgoose P, et al. International
Normalized Ratio Stability in Warfarin-Experienced Patients with Nonvalvular Atrial Fibrillation.
Am J Cardiovasc Drugs. 2015;15(3):205-11.
198. Rao SR, Reisman JI, Kressin NR, Berlowitz DR, Ash AS, Ozonoff A, et al. Explaining
Racial Disparities in Anticoagulation Control. Am J Med Qual. 2014;30(3):214-22.
199. Apostolakis S, Sullivan RM, Olshansky B, Lip GYH. Factors affecting quality of
anticoagulation control among patients with atrial fibrillation on warfarin: The same-tt2r2 score.
Chest. 2013;144(5):1555-63.
200. Rose AJ, Hylek EM, Ozonoff A, Ash AS, Reisman JI, Berlowitz DR. Patient
characteristics associated with oral anticoagulation control: results of the Veterans AffaiRs
Study to Improve Anticoagulation (VARIA). J Thromb Haemost. 2010;8(10):2182-91.
201. Rose AJ, Ozonoff A, Henault LE, Hylek EM. Warfarin for atrial fibrillation in community-
based practise. J Thromb Haemost. 2008;6(10):1647-54.
202. Melamed OC, Horowitz G, Elhayany A, Vinker S. Quality of anticoagulation control
among patients with atrial fibrillation. Am J Manag Care. 2011;17(3):232-7.
203. Boulanger L, Kim J, Friedman M, Hauch O, Foster T, Menzin J. Patterns of use of
antithrombotic therapy and quality of anticoagulation among patients with non-valvular atrial
fibrillation in clinical practice. Int J Clin Pract. 2006;60(3):258-64.
204. Williams BA, Evans MA, Honushefsky AM, Berger PB. Clinical Prediction Model for
Time in Therapeutic Range While on Warfarin in Newly Diagnosed Atrial Fibrillation. J Am
Heart Assoc. 2017;6(10):e006669.
205. Macedo AF, Bell J, McCarron C, Conroy R, Richardson J, Scowcroft A, et al.
Determinants of oral anticoagulation control in new warfarin patients: analysis using data from
Clinical Practice Research Datalink. Thromb Res. 2015;136(2):250-60.
206. Wieloch M, Själander A, Frykman V, Rosenqvist M, Eriksson N, Svensson PJ.
Anticoagulation control in Sweden: reports of time in therapeutic range, major bleeding, and
thrombo-embolic complications from the national quality registry AuriculA. Eur Heart J.
2011;32(18):2282-9.
207. Dlott JS, George RA, Huang X, Odeh M, Kaufman HW, Ansell J, et al. National
assessment of warfarin anticoagulation therapy for stroke prevention in atrial fibrillation.
Circulation. 2014;129(13):1407-14.
208. White RD, Riggs KW, Ege EJ, Petroski GF, Koerber SM, Flaker G. The effect of the
amiodarone–warfarin interaction on anticoagulation quality in a single, high-quality
anticoagulation center. Blood Coagul Fibrinolysis. 2016;27(2).
209. Yong C, Azarbal F, Abnousi F, Heidenreich PA, Schmitt S, Fan J, et al. Racial
Differences in Quality of Anticoagulation Therapy for Atrial Fibrillation (from the TREAT-AF
Study). Am J Cardiol. 2016;117(1):61-8.
345
210. Golwala H, Jackson LR, 2nd, Simon DN, Piccini JP, Gersh B, Go AS, et al.
Racial/ethnic differences in atrial fibrillation symptoms, treatment patterns, and outcomes:
Insights from Outcomes Registry for Better Informed Treatment for Atrial Fibrillation Registry.
Am Heart J. 2016;174:29-36.
211. Okumura K, Komatsu T, Yamashita T, Okuyama Y, Harada M, Konta Y, et al. Time in
the therapeutic range during warfarin therapy in Japanese patients with non-valvular atrial
fibrillation. - A multicenter study of its status and infuential factors. Circ J. 2011;75(9):2087-94.
212. Witt DM, Delate T, Clark NP, Martell C, Tran T, Crowther MA, et al. Outcomes and
predictors of very stable INR control during chronic anticoagulation therapy. Blood.
2009;114(5):952.
213. Lin KJ, Singer DE, Glynn RJ, Blackley S, Zhou L, Liu J, et al. Prediction Score for
Anticoagulation Control Quality Among Older Adults. J Am Heart Assoc. 2017;6(10):e006814.
214. Horne BD, Lenzini PA, Wadelius M, Jorgensen AL, Kimmel SE, Ridker PM, et al.
Pharmacogenetic Warfarin Dose Refinements Remain Significantly Influenced by Genetic
Factors after One Week of Therapy. Thromb Haemost. 2012;107(2):232-40.
215. Ferder NS, Eby CS, Deych E, Harris JK, Ridker PM, Milligan PE, et al. Ability of
VKORC1 and CYP2C9 to predict therapeutic warfarin dose during the initial weeks of therapy.
J Thromb Haemost. 2010;8(1):95-100.
216. Dean L. Warfarin Therapy and the Genotypes CYP2C9 and VKORC1. 2012 Mar 8
[Updated 2016 Jun 8]. In: Pratt V, McLeod H, Dean L, et al., editors. Medical Genetics
Summaries [Internet]. Bethesda (MD): National Center for Biotechnology Information (US);
2012-.
217. Skov J, Bladbjerg E-M, Leppin A, Jespersen J. The influence of VKORC1 and CYP2C9
gene sequence variants on the stability of maintenance phase warfarin treatment. Thromb
Res. 2013;131(2):125-9.
218. Polovina M, Djikic D, Vlajkovic A, Vilotijevic M. Patients’ knowledge and perspectives
on vitamin K antagonists for stroke prevention in atrial fibrillation: implications for treatment
quality. Anatolian J Cardiol. 2017;18(3):239-40.
219. Clarkesmith DE, Pattison HM, Lip GYH, Lane DA. Educational intervention improves
anticoagulation control in atrial fibrillation patients: the TREAT randomised trial. PLoS One.
2013;8(9):e74037.
220. Wang Y, Kong MC, Lee LH, Ng HJ, Ko Y. Knowledge, satisfaction, and concerns
regarding warfarin therapy and their association with warfarin adherence and anticoagulation
control. Thromb Res. 2014;133(4):550-4.
221. Tay KH, Lip GYH, Lane DA. Anticoagulation variability: is it the physician, patient or
hospital? J Intern Med. 2009;265(3):303-6.
346
222. van Walraven C, Jennings A, Oake N, Fergusson D, Forster AJ. Effect of Study Setting
on Anticoagulation Control. Chest. 2006;129(5):1155-66.
223. Baker WL, Cios DA, Sander SD, Coleman CI. Meta-analysis to assess the quality of
warfarin control in atrial fibrillation patients in the United States. J Manag Care Pharm.
2009;15(3):244-52.
224. Jurcuţ R, Militaru S, Geavlete O, Drăgotoiu N, Sipoş S, Roşulescu R, et al. Predictive
factors for obtaining a correct therapeutic range using antivitamin K anticoagulants: a tertiary
center experience of patient adherence to anticoagulant therapy. Patient Prefer Adherence.
2015;9:1271-8.
225. Ababneh MA, Al-Azzam SI, Alzoubi KH, Rababa’h AM. Adherence in outpatients
taking warfarin and its effect on anticoagulation control in Jordan. Int J Clin Pharm.
2016;38(4):816-21.
226. Obamiro KO, Chalmers L, Bereznicki LRE. A Summary of the Literature Evaluating
Adherence and Persistence with Oral Anticoagulants in Atrial Fibrillation. Am J Cardiovasc
Drugs. 2016;16(5):349-63.
227. Yao X, Abraham NS, Alexander GC, Crown W, Montori VM, Sangaralingham LR, et
al. Effect of Adherence to Oral Anticoagulants on Risk of Stroke and Major Bleeding Among
Patients With Atrial Fibrillation. J Am Heart Assoc. 2016;5(2).
228. Manzoor BS, Cheng WH, Lee JC, Uppuluri EM, Nutescu EA. Quality of Pharmacist-
Managed Anticoagulation Therapy in Long-Term Ambulatory Settings: A Systematic Review.
Ann Pharmacother. 2017:1060028017721241.
229. Zhou S, Sheng XY, Xiang Q, Wang ZN, Zhou Y, Cui YM. Comparing the effectiveness
of pharmacist-managed warfarin anticoagulation with other models: a systematic review and
meta-analysis. J Clin Pharm Ther. 2016;41(6):602-11.
230. Heu MK, Welborn T, Nagykaldi Z. Clinical Question: In adult patients on warfarin, does-
home-self-testing of prothrombin time and/or international normalized ratio provide the same
outcomes compared to testing by a home health nurse or in a clinical setting? J Okla State
Med Assoc. 2016;109(3):99-100.
231. Heneghan C, Ward A, Perera R, Bankhead C, Fuller A, Stevens R, et al. Self-
monitoring of oral anticoagulation: systematic review and meta-analysis of individual patient
data. The Lancet. 2012;379(9813):322-34.
232. Bloomfield HE, Krause A, Greer N, et al. Meta-analysis: Effect of patient self-testing
and self-management of long-term anticoagulation on major clinical outcomes. Ann Intern
Med. 2011;154(7):472-82.
233. Cressman AM, Macdonald EM, Yao Z, Austin PC, Gomes T, Paterson JM, et al.
Socioeconomic status and risk of hemorrhage during warfarin therapy for atrial fibrillation: A
population-based study. Am Heart J. 2015;170(1):133-40.
347
234. Violi F, Lip GY, Pignatelli P, Pastori D. Interaction Between Dietary Vitamin K Intake
and Anticoagulation by Vitamin K Antagonists: Is It Really True?: A Systematic Review.
Medicine (Baltimore). 2016;95(10):e2895.
235. Ge B, Zhang Z, Zuo Z. Updates on the Clinical Evidenced Herb-Warfarin Interactions.
Evid Based Complement Alternat Med. 2014;2014:18.
236. Lai YF, Cheen MH, Lim SH, Yeo FH, Nah SC, Kong MC, et al. The effects of fasting
in Muslim patients taking warfarin. J Thromb Haemost. 2014;12(3):349-54.
237. Awiwi MO, Yagli ZA, Elbir F, Aglar AA, Guler E, Vural U. The effects of Ramadan
fasting on patients with prosthetic heart valve taking warfarin for anticoagulation. J Saudi Heart
Assoc. 2017;29(1):1-6.
238. Gomberg-Maitland M, Wenger NK, Feyzi J, Lengyel M, Volgman AS, Petersen P, et
al. Anticoagulation in women with non-valvular atrial fibrillation in the stroke prevention using
an oral thrombin inhibitor (SPORTIF) trials. Eur Heart J. 2006;27(16):1947-53.
239. Pancholy SB, Sharma PS, Pancholy DS, Patel TM, Callans DJ, Marchlinski FE. Meta-
Analysis of Gender Differences in Residual Stroke Risk and Major Bleeding in Patients With
Nonvalvular Atrial Fibrillation Treated With Oral Anticoagulants. Am J Cardiol.
2014;113(3):485-90.
240. Sonuga BO, Hellenberg DA, Cupido CS, Jaeger C. Profile and anticoagulation
outcomes of patients on warfarin therapy in an urban hospital in Cape Town, South Africa. Afr
J Prim Health Care Fam Med. 2016;8(1):1032.
241. Hylek EM, Evans-Molina C, Shea C, Henault LE, Regan S. Major Hemorrhage and
Tolerability of Warfarin in the First Year of Therapy Among Elderly Patients With Atrial
Fibrillation. Circulation. 2007;115(21):2689.
242. Lip GYH, Lane DA. Stroke prevention with oral anticoagulation therapy in patients with
atrial fibrillation. Circ J. 2013;77(6):1380-8.
243. Pengo V, Cucchini U, Denas G, Davidson BL, Marzot F, Jose SP, et al. Lower versus
standard intensity oral anticoagulant therapy (OAT) in elderly warfarin-experienced patients
with non-valvular atrial fibrillation. Thromb Haemost. 2010;103(02):442-9.
244. Pengo V, Legnani C, Noventa F, Palareti G. Oral anticoagulant therapy in patients with
nonrheumatic atrial fibrillation and risk of bleeding. A Multicenter Inception Cohort Study.
Thromb Haemost. 2001;85(3):418-22.
245. Wolff A, Shantsila E, Lip GYH, Lane DA. Impact of advanced age on management and
prognosis in atrial fibrillation: insights from a population-based study in general practice. Age
Ageing. 2015;44(5):874-8.
246. Pugh D, Pugh J, Mead GE. Attitudes of physicians regarding anticoagulation for atrial
fibrillation: a systematic review. Age Ageing. 2011;40(6):675-83.
348
247. Poli D, Antonucci E, Grifoni E, Abbate R, Gensini GF, Prisco D. Bleeding Risk During
Oral Anticoagulation in Atrial Fibrillation Patients Older Than 80 Years. J Am Coll Cardiol.
2009;54(11):999.
248. Fang MC, Chang Y, Hylek EM, Rosand J, Greenberg SM, Go AS, et al. Advanced age,
anticoagulation intensity, and risk for intracranial hemorrhage among patients taking warfarin
for atrial fibrillation. Ann Intern Med. 2004;141(10):745-52.
249. Lip GY, Clementy N, Pericart L, Banerjee A, Fauchier L. Stroke and major bleeding
risk in elderly patients aged >/=75 years with atrial fibrillation: the Loire Valley atrial fibrillation
project. Stroke. 2015;46(1):143-50.
250. Mant J, Hobbs FDR, Fletcher K, Roalfe A, Fitzmaurice D, Lip GYH, et al. Warfarin
versus aspirin for stroke prevention in an elderly community population with atrial fibrillation
(the Birmingham Atrial Fibrillation Treatment of the Aged Study, BAFTA): a randomised
controlled trial. The Lancet. 2007;370(9586):493-503.
251. Rash A, Downes T, Portner R, Yeo WW, Morgan N, Channer KS. A randomised
controlled trial of warfarin versus aspirin for stroke prevention in octogenarians with atrial
fibrillation (WASPO). Age Ageing. 2007;36(2):151-6.
252. Palareti G, Hirsh J, Legnani C, et al. Oral anticoagulation treatment in the elderly: A
nested, prospective, case-control study. Arch Intern Med. 2000;160(4):470-8.
253. Kose E, Arai S, An T, Kikkawa A, Aoyama T, Matsumoto Y, et al. Analysis of factors
affecting time in therapeutic range control after warfarin administration. Pharmazie.
2015;70(7):494-8.
254. Kim EJ, Ozonoff A, Hylek EM, Berlowitz DR, Ash AS, Miller DR, et al. Predicting
outcomes among patients with atrial fibrillation and heart failure receiving anticoagulation with
warfarin. Thromb Haemost. 2015;114(1):70-7.
255. Tomita H, Kadokami T, Momii H, Kawamura N, Yoshida M, Inou T, et al. Patient
Factors against Stable Control of Warfarin Therapy for Japanese Non-valvular Atrial
Fibrillation Patients. Thromb Res. 2013;132(5):537-42.
256. Pignatelli P, Pastori D, Vicario T, Bucci T, Del Ben M, Russo R, et al. Relationship
between Mediterranean diet and time in therapeutic range in atrial fibrillation patients taking
vitamin K antagonists. EP Europace. 2015;17(8):1223-8.
257. Efird LM, Mishkin DS, Berlowitz DR, Ash AS, Hylek EM, Ozonoff A, et al. Stratifying
the Risks of Oral Anticoagulation in Patients With Liver Disease. Circ Cardiovasc Qual
Outcomes. 2014;7(3):461-7.
258. Kooistra HAM, Veeger NJGM, Khorsand N, Kluin-Nelemans HC, Meijer K, Piersma-
Wichers M. Long-term quality of VKA treatment and clinical outcome after extreme
overanticoagulation in 14,777 AF and VTE patients. Thromb Haemost. 2015;113(4):881-90.
349
259. Paradise HT, Berlowitz DR, Ozonoff A, Miller DR, Hylek EM, Ash AS, et al. Outcomes
of Anticoagulation Therapy in Patients with Mental Health Conditions. J Gen Intern Med.
2014;29(6):855-61.
260. Sood MM, Komenda P, Sood AR, Rigatto C, Bueti J. The intersection of risk and
benefit: is warfarin anticoagulation suitable for atrial fibrillation in patients on hemodialysis?
Chest. 2009;136(4):1128-33.
261. Holden RM, Booth SL. Vascular calcification in chronic kidney disease: the role of
vitamin K. Nat Clin Pract Nephrol. 2007;3(10):522.
262. Elliott MJ, Zimmerman D, Holden RM. Warfarin anticoagulation in hemodialysis
patients: a systematic review of bleeding rates. Am J Kidney Dis. 2007;50(3):433-40.
263. Szummer K, Gasparini A, Eliasson S, Ärnlöv J, Qureshi AR, Bárány P, et al. Time in
Therapeutic Range and Outcomes After Warfarin Initiation in Newly Diagnosed Atrial
Fibrillation Patients With Renal Dysfunction. J Am Heart Assoc. 2017;6(3).
264. Proietti M, Lane DA, Lip GYH. Chronic Kidney Disease, Time in Therapeutic Range
and Adverse Clinical Outcomes in Anticoagulated Patients with Non-valvular Atrial Fibrillation:
Observations from the SPORTIF Trials. EBioMedicine. 2016;8:309-16.
265. Friberg L, Benson L, Lip GYH. Balancing stroke and bleeding risks in patients with
atrial fibrillation and renal failure: the Swedish Atrial Fibrillation Cohort study. Eur Heart J.
2015;36(5):297-306.
266. Kooiman J, van Rein N, Spaans B, van Beers KAJ, Bank JR, van de Peppel WR, et
al. Efficacy and safety of vitamin K-antagonists (VKA) for atrial fibrillation in non-dialysis
dependent chronic kidney disease. PLoS One. 2014;9(5):e94420.
267. Kleinow ME, Garwood CL, Clemente JL, Whittaker P. Effect of chronic kidney disease
on warfarin management in a pharmacist-managed anticoagulation clinic. J Manag Care
Pharm. 2011;17(7):523-30.
268. Limdi NA, Beasley TM, Baird MF, Goldstein JA, McGwin G, Arnett DK, et al. Kidney
function influences warfarin responsiveness and hemorrhagic complications. J Am Soc
Nephrol. 2009;20(4):912-21.
269. Esteve-Pastor MA, Rivera-Caravaca JM, Roldán-Rabadán I, Roldán V, Muñiz J, Raña-
Míguez P, et al. Relation of renal dysfunction to quality of anticoagulation control in patients
with atrial fibrillation: the FANTASIIA registry. Thromb Haemost. 2018;118(02):279-87.
270. Lobos-Bejarano JM, Castellanos Rodriguez A, Barrios V, Escobar C, Polo-Garcia J,
Del Castillo-Rodriguez JC, et al. Influence of renal function on anticoagulation control in
patients with non-valvular atrial fibrillation taking vitamin K antagonists. Int J Clin Pract.
2017;71(9).
350
271. Wittkowsky AK, Devine EB. Frequency and Causes of Overanticoagulation and
Underanticoagulation in Patients Treated with Warfarin. Pharmacotherapy. 2004;24(10):1311-
6.
272. McMahan DA, Smith DM, Carey MA, Zhou XH. Risk of major hemorrhage for
outpatients treated with warfarin. J Gen Intern Med. 1998;13(5):311-6.
273. Gong IY, Schwarz UI, Crown N, Dresser GK, Lazo-Langner A, Zou G, et al. Clinical
and genetic determinants of warfarin pharmacokinetics and pharmacodynamics during
treatment initiation. PLoS One. 2011;6(11):e27808.
274. Dreisbach AW, Japa S, Gebrekal AB, Mowry SE, Lertora JJ, Kamath BL, et al.
Cytochrome P4502C9 activity in end-stage renal disease. Clin Pharmacol Ther.
2003;73(5):475-7.
275. Limdi NA, Limdi MA, Cavallari L, Anderson AM, Crowley MR, Baird MF, et al. Warfarin
Dosing in Patients With Impaired Kidney Function. Am J Kidney Dis. 2010;56(5):823-31.
276. Ageno W, Gallus AS, Wittkowsky A, Crowther M, Hylek EM, Palareti G. Oral
Anticoagulant Therapy: Antithrombotic Therapy and Prevention of Thrombosis, 9th ed:
American College of Chest Physicians Evidence-Based Clinical Practice Guidelines. Chest.
2012;141(2 Suppl):e44S-e88S.
277. Joint Formulary Comittee. British National Formulary. 74, editor. London: BMJ Group
and Pharmaceutical Press; 2017-2018.
278. Angoulvant D, Villejoubert O, Bejan-Angoulvant T, Ivanes F, Saint Etienne C, Lip GYH,
et al. Effect of active smoking on comparative efficacy of antithrombotic therapy in patients
with atrial fibrillation: The loire valley atrial fibrillation project. Chest. 2015;148(2):491-8.
279. Nathisuwan S, Dilokthornsakul P, Chaiyakunapruk N, Morarai T, Yodting T,
Piriyachananusorn N. Assessing evidence of interaction between smoking and warfarin: a
systematic review and meta-analysis. Chest. 2011;139(5):1130-9.
280. Lip GY, Waldo AL, Ip J, Martin DT, Bersohn MM, Choucair WK, et al. Determinants of
Time in Therapeutic Range in Patients Receiving Oral Anticoagulants (A Substudy of
IMPACT). Am J Cardiol. 2016;118(11):1680-4.
281. Martin DT, Bersohn MM, Waldo AL, Wathen MS, Choucair WK, Lip GY, et al.
Randomized trial of atrial arrhythmia monitoring to guide anticoagulation in patients with
implanted defibrillator and cardiac resynchronization devices. Eur Heart J. 2015;36(26):1660-
8.
282. Holbrook AM, Pereira JA, Labiris R, et al. Systematic overview of warfarin and its drug
and food interactions. Arch Intern Med. 2005;165(10):1095-106.
283. Lane DA, Barker RV, Lip G. Best Practice for Atrial Fibrillation Patient Education. Curr
Pharm Des. 2015;21(5):533-43.
351
284. Lane DA, Ponsford J, Shelley A, Sirpal A, Lip GYH. Patient knowledge and perceptions
of atrial fibrillation and anticoagulant therapy: Effects of an educational intervention
programme: The West Birmingham Atrial Fibrillation Project. Int J Cardiol. 2006;110(3):354-8.
285. Lip GY, Kamath S, Jafri M, Mohammed A, Bareford D. Ethnic differences in patient
perceptions of atrial fibrillation and anticoagulation therapy: The West Birmingham atrial
fibrillation project. Stroke. 2002;33.
286. McCabe PJ, Schad S, Hampton A, Holland DE. Knowledge and self-management
behaviors of patients with recently detected atrial fibrillation. Heart Lung. 2008;37(2):79-90.
287. Lip GYH, Agnelli G, Thach AA, Knight E, Rost D, Tangelder MJD. Oral anticoagulation
in atrial fibrillation: A pan-European patient survey. Eur J Intern Med. 2007;18(3):202-8.
288. Dantas GC, Thompson BV, Manson JA, Tracy CS, Upshur REG. Patients'
perspectives on taking warfarin: qualitative study in family practice. BMC Fam Pract.
2004;5(1):15.
289. Nadar S, Begum N, Kaur B, Sandhu S, Lip G. Patients' understanding of anticoagulant
therapy in a multiethnic population. J R Soc Med. 2003;96.
290. Fuller R, Dudley N, Blacktop J. Avoidance hierarchies and preferences for
anticoagulation--semi-qualitative analysis of older patients' views about stroke prevention and
the use of warfarin. Age Ageing. 2004;33(6):608-11.
291. Fuller R, Dudley N, Blacktop J. Risk communication and older people—understanding
of probability and risk information by medical inpatients aged 75 years and older. Age Ageing.
2001;30(6):473-6.
292. Hernández Madrid A, Potpara TS, Dagres N, Chen J, Larsen TB, Estner H, et al.
Differences in attitude, education, and knowledge about oral anticoagulation therapy among
patients with atrial fibrillation in Europe: result of a self-assessment patient survey conducted
by the European Heart Rhythm Association. EP Europace. 2016;18(3):463-7.
293. Clarkesmith DE, Pattison HM, Borg Xuereb C, Lane DA. Developing a Complex
Educational-Behavioural Intervention: The TREAT Intervention for Patients with Atrial
Fibrillation. Healthcare (Basel). 2016;4(1).
294. Clarkesmith DE, Pattison HM, Lane DA. Educational and behavioural interventions for
anticoagulant therapy in patients with atrial fibrillation. Cochrane Database Syst Rev.
2013(6):CD008600.
295. Beyth RJ, Quinn L, Landefeld CS. A multicomponent intervention to prevent major
bleeding complications in older patients receiving warfarin. A randomized, controlled trial. Ann
Intern Med. 2000;133(9):687-95.
296. Khan TI, Kamali F, Kesteven P, Avery P, Wynne H. The value of education and self-
monitoring in the management of warfarin therapy in older patients with unstable control of
anticoagulation. Br J Haematol. 2004;126(4):557-64.
352
297. Hori M, Matsumoto M, Tanahashi N, Momomura S, Uchiyama S, Goto S, et al.
Rivaroxaban vs. warfarin in Japanese patients with atrial fibrillation - the J-ROCKET AF study.
Circ J. 2012;76(9):2104-11.
298. Schulman S, Shortt B, Robinson M, Eikelboom JW. Adherence to anticoagulant
treatment with dabigatran in a real-world setting. J Thromb Haemost. 2013;11(7):1295-9.
299. Gorst-Rasmussen A, Skjøth F, Larsen TB, Rasmussen LH, Lip GYH, Lane DA.
Dabigatran adherence in atrial fibrillation patients during the first year after diagnosis: a
nationwide cohort study. J Thromb Haemost. 2015;13(4):495-504.
300. Shore S, Carey EP, Turakhia MP, Jackevicius CA, Cunningham F, Pilote L, et al.
Adherence to dabigatran therapy and longitudinal patient outcomes: insights from the veterans
health administration. Am Heart J. 2014;167(6):810-7.
301. Beyer-Westendorf J, Ehlken B, Evers T. Real-world persistence and adherence to oral
anticoagulation for stroke risk reduction in patients with atrial fibrillation. EP Europace.
2016;18(8):1150-7.
302. Zhou M, Chang H-Y, Segal JB, Alexander GC, Singh S. Adherence to a Novel Oral
Anticoagulant Among Patients with Atrial Fibrillation. J Manag Care Spec Pharm.
2015;21(11):1054-62.
303. McHorney CA, Crivera C, Laliberté F, Nelson WW, Germain G, Bookhart B, et al.
Adherence to non-vitamin-K-antagonist oral anticoagulant medications based on the
Pharmacy Quality Alliance measure. Curr Med Res Opin. 2015;31(12):2167-73.
304. Crivera C, Nelson WW, Bookhart B, Martin S, Germain G, Laliberté F, et al. Pharmacy
quality alliance measure: adherence to non-warfarin oral anticoagulant medications. Curr Med
Res Opin. 2015;31(10):1889-95.
305. Cutler TW, Chuang A, Huynh TD, Witt RG, Branch J, Pon T, et al. A Retrospective
Descriptive Analysis of Patient Adherence to Dabigatran at a Large Academic Medical Center.
J Manag Care Pharm. 2014;20(10):1028-34.
306. Tsai K, Erickson SC, Yang J, Harada AS, Solow BK, Lew HC. Adherence, persistence,
and switching patterns of dabigatran etexilate. Am J Manag Care. 2013;19(9):e325-32.
307. Davis NJ, Billett HH, Cohen HW, Arnsten JH. Impact of Adherence, Knowledge, and
Quality of Life on Anticoagulation Control. Ann Pharmacother. 2005;39(4):632-6.
308. Castellucci LA, Shaw J, van der Salm K, Erkens P, Le Gal G, Petrcich W, et al. Self-
reported adherence to anticoagulation and its determinants using the Morisky medication
adherence scale. Thromb Res. 2015;136(4):727-31.
309. Kimmel SE, Chen Z, Price M, et al. The influence of patient adherence on
anticoagulation control with warfarin: Results from the international normalized ratio
adherence and genetics (in-range) study. Arch Intern Med. 2007;167(3):229-35.
353
310. Gohil KJ, Patel JA. Herb-drug interactions: A review and study based on assessment
of clinical case reports in literature. Indian J Pharmacol. 2007;39(3):129.
311. Rindone JP, Murphy TW. Warfarin-cranberry juice interaction resulting in profound
hypoprothrombinemia and bleeding. Am J Ther. 2006;13(3):283-4.
312. Jaspers Focks J, van Vugt SPG, Albers-Akkers MTH, Lamfers EJP, Bloem-de Vries
LM, Verheugt FWA, et al. Low performance of bleeding risk models in the very elderly with
atrial fibrillation using vitamin K antagonists. J Thromb Haemost. 2016;14(9):1715-24.
313. Friberg L, Rosenqvist M, Lip GYH. Net Clinical Benefit of Warfarin in Patients with
Atrial Fibrillation. Circulation. 2012;125(19):2298.
314. Lip GY, Lane DA. Bleeding risk assessment in atrial fibrillation: observations on the
use and misuse of bleeding risk scores. J Thromb Haemost. 2016;14(9):1711-4.
315. Devereaux PJ, Fahey T, Anderson DR, Gardner MJ, Putnam W, Flowerdew GJ, et al.
Differences between perspective of physicians and patients on anticoagulation in patients with
atrial fibrillation: observational study. BMJ. 2001;323(7323):1218.
316. Freedman B, Potpara TS, Lip GYH. Stroke prevention in atrial fibrillation. Lancet.
2016;388(10046):806-17.
317. Pivatto Júnior F, RS S, l R, RR W, R M, SS B, et al. SAMe-TT2R2 Score in the
Outpatient Anticoagulation Clinic to Predict Time in
Therapeutic Range and Adverse Events. Arq Bras Cardiol. 2017;108(4):290-6.
318. Roldan V, Cancio S, Galvez J, Valdes M, Vicente V, Marin F, et al. The SAMe-TTR
Score Predicts Poor Anticoagulation Control in AF Patients: A Prospective 'Real-world'
Inception Cohort Study. Am J Med. 2015;128(11):1237-43.
319. Abumuaileq RR, Abu-Assi E, Raposeiras-Roubin S, Lopez-Lopez A, Redondo-
Dieguez A, Alvarez-Iglesias D, et al. Evaluation of SAMe-TT2R2 risk score for predicting the
quality of anticoagulation control in a real-world cohort of patients with non-valvular atrial
fibrillation on vitamin-K antagonists. Europace. 2015;17(5):711-7.
320. Skov J, Bladbjerg E-M, Bor MV, Gram J. SAmett2r2 does not predict time in
therapeutic range of the international normalized ratio in patients attending a high-quality
anticoagulation clinic. Chest. 2014;145(1):187-8.
321. Poli D, Antonucci E, Testa S, Lip GY. A prospective validation of the SAME-TT2R 2
score: how to identify atrial fibrillation patients who will have good anticoagulation control on
warfarin. Intern Emerg Med. 2014;9(4):443-7.
322. Gallego P, Roldan V, Marin F, Galvez J, Valdes M, Vicente V, et al. SAMe-TT2R2
score, time in therapeutic range, and outcomes in anticoagulated patients with atrial fibrillation.
Am J Med. 2014;127(11):1083-8.
354
323. Szymanski FM, Lip GYH, Filipiak KJ, Platek AE, Karpinski G. Usefulness of the SAME-
TT2R2 score to predict anticoagulation control on VKA in patients with atrial fibrillation and
obstructive sleep apnea. Int J Cardiol. 2016;204:200-5.
324. Gorzelak-Pabis P, Zyzak S, Krewko L, Broncel M. Assessment of the mean time in the
therapeutic INR range and the SAME-TT2R2 score in patients with atrial fibrillation and
cognitive impairment. Pol Arch Med Wewn. 2016;126(7-8):494-501.
325. Chan PH, Hai JJ, Chan EW, Li WH, Tse HF, Wong ICK, et al. Use of the SAMe-
TT(2)R(2) Score to Predict Good Anticoagulation Control with Warfarin in Chinese Patients
with Atrial Fibrillation: Relationship to Ischemic Stroke Incidence. PLoS One.
2016;11(3):e0150674.
326. Bernaitis N, Ching CK, Chen L, Hon JS, Teo SC, Davey AK, et al. The Sex, Age,
Medical History, Treatment, Tobacco Use, Race Risk (SAMe TT2R2) Score Predicts Warfarin
Control in a Singaporean Population. J Stroke Cerebrovasc Dis. 2016;26(1):64-9.
327. Proietti M, Lane DA, Lip GYH. Relation of the SAMe-TT2R2 score to quality of
anticoagulation control and thromboembolic events in atrial fibrillation patients: Observations
from the SPORTIF trials. Int J Cardiol. 2016;216:168-72.
328. Lobos-Bejarano JM, Barrios V, Polo-García J, Escobar C, Vargas-Ortega D, Marín-
Montañés N, et al. Evaluation of SAMe-TT2R2 score and other clinical factors influencing the
quality of anticoagulation therapy in non-valvular atrial fibrillation: a nationwide study in Spain.
Curr Med Res Opin. 2016;32(7):1-7.
329. Ruiz-Ortiz M, Bertomeu V, Cequier A, Marin F, Anguita M. Validation of the SAMe-
TT2R2 score in a nationwide population of nonvalvular atrial fibrillation patients on vitamin K
antagonists. Thromb Haemost. 2015;114(4):695-701.
330. Lip GYH, Haguenoer K, Saint-Etienne C, Fauchier L. RElationship of the same-tt2r2
score to poor-quality anticoagulation, stroke, clinically relevant bleeding, and mortality in
patients with atrial fibrillation. Chest. 2014;146(3):719-26.
331. DeSantis G, Hogan-Schlientz J, Liska G, Kipp S, Sallee R, Wurster M, et al. STABLE
results: warfarin home monitoring achieves excellent INR control. Am J Manag Care.
2014;20(3):202-9.
332. Samsa GP, Matchar DB. Relationship between test frequency and outcomes of
anticoagulation: a literature review and commentary with implications for the design of
randomized trials of patient self-management. J Thromb Thrombolysis. 2000;9(3):283-92.
333. Singer DE, Albers GW, Dalen JE, Go AS, Halperin JL, Manning WJ. Antithrombotic
therapy in atrial fibrillation: the Seventh ACCP Conference on Antithrombotic and Thrombolytic
Therapy. Chest. 2004;126(3).
355
334. Heneghan CJ, Garcia-Alamino JM, Spencer EA, Ward AM, Perera R, Bankhead C, et
al. Self-monitoring and self-management of oral anticoagulation. Cochrane Database Syst
Rev. 2016(7).
335. Gadisseur APA, Kaptein AA, Breukink-Engbers WGM, Van Der Meer FJM, R.
Rosendaal F. Patient self-management of oral anticoagulant care vs. management by
specialized anticoagulation clinics: positive effects on quality of life. J Thromb Haemost.
2004;2(4):584-91.
336. Zulkifly H, Lip GYH, Lane DA. Use of the SAMe-TT2R2 score to predict anticoagulation
control in atrial fibrillation and venous thromboembolism patients receiving vitamin K
antagonists: A review. Heart Rhythm. 2018;15(4):615-23.
337. Heidbuchel H, Verhamme P, Alings M, Antz M, Diener H-C, Hacke W, et al. Updated
European Heart Rhythm Association practical guide on the use of non-vitamin-K antagonist
anticoagulants in patients with non-valvular atrial fibrillation: Executive summary. Eur Heart J.
2017;38(27):2137-49.
338. Mekaj YH, Mekaj AY, Duci SB, Miftari EI. New oral anticoagulants: their advantages
and disadvantages compared with vitamin K antagonists in the prevention and treatment of
patients with thromboembolic events. Ther Clin Risk Manag. 2015;11:967-77.
339. Mahaffey KW, Hellkamp AS, Patel MR, Hannan KL, Schwabe K, Nessel CC, et al. End
of Study Transition From Study Drug to Open-Label Vitamin K Antagonist Therapy. Circ
Cardiovasc Qual Outcomes. 2013;6(4):470.
340. Ruff CT, Giugliano RP, Braunwald E, Mercuri M, Curt V, Betcher J, et al. Transition of
Patients From Blinded Study Drug to Open-Label Anticoagulation: The ENGAGE AF–TIMI 48
Trial. J Am Coll Cardiol. 2014;64(6):576-84.
341. Bista D, Chalmers L, Bereznicki L, Peterson G. Potential use of NOACs in developing
countries: pros and cons. Eur J Clin Pharmacol. 2014;70(7):817-28.
342. Ling C, Siaw Han M, Len T, Siang Ning S, Ying K, Sing Yee C, et al. Validation of the
SAMe-TT2R2 score in atrial fibrillation patients on warfarin therapy: a multicenter, multiethnic
Asian cohort. Conference Abstract NHAM Annual Scientific Meeting 8-10 April. 2016.
343. Kataruka A, Kong X, Haymart B, Kline-Rogers E, Almany S, Kozlowski J, et al. SAMe-
TT2R2 predicts quality of anticoagulation in patients with acute venous thromboembolism:
The MAQI2 experience. Vasc Med. 2017;22(3):197-203.
344. National Institute for Clinical Excellence (2014) Atrial Fibrillation: Management. NICE
Guideline (CG180).
345. Gallego P, Roldan V, Marín F, Romera M, Valdés M, Vicente V, et al. Cessation of oral
anticoagulation in relation to mortality and the risk of thrombotic events in patients with atrial
fibrillation. Thromb Haemost. 2013;110(6):1189-98.
356
346. Fauchier L, Poli D, Olshansky B. The SAMe-TT2R2 score and quality of anticoagulation
in AF: Can we predict which patient benefits from anticoagulation? Thromb Haemost.
2015;114(4):657-9.
347. Schmitt L, Speckman J, Ansell J. Quality Assessment of Anticoagulation Dose
Management: Comparative Evaluation of Measures of Time-in-Therapeutic Range. J Thromb
Thrombolysis. 2003;15(3):213-6.
348. Saksena D, Muralidharan S, Mishra YK, Kanhere V, Mohanty BB, Srivastava CP, et
al. Anticoagulation Management in Patients with Valve Replacement. J Assoc Physicians
India. 2018;66:59.
349. Lip GYH, Collet JP, de Caterina R, Fauchier L, Lane DA, Larsen TB, et al.
Antithrombotic Therapy in Atrial Fibrillation Associated with Valvular Heart Disease: Executive
Summary of a Joint Consensus Document from the European Heart Rhythm Association
(EHRA) and European Society of Cardiology Working Group on Thrombosis, Endorsed by the
ESC Working Group on Valvular Heart Disease, Cardiac Arrhythmia Society of Southern
Africa (CASSA), Heart Rhythm Society (HRS), Asia Pacific Heart Rhythm Society (APHRS),
South African Heart (SA Heart) Association and Sociedad Latinoamericana de Estimulacion
Cardiaca y Electrofisiologia (SOLEACE). Thromb Haemost. 2017;117(12):2215-36.
350. Iung B, Vahanian A. Epidemiology of acquired valvular heart disease. Can J Cardiol.
2014;30(9):962-70.
351. Lip GYH, Collet JP, Caterina Rd, Fauchier L, Lane DA, Larsen TB, et al. Antithrombotic
therapy in atrial fibrillation associated with valvular heart disease: a joint consensus document
from the European Heart Rhythm Association (EHRA) and European Society of Cardiology
Working Group on Thrombosis, endorsed by the ESC Working Group on Valvular Heart
Disease, Cardiac Arrhythmia Society of Southern Africa (CASSA), Heart Rhythm Society
(HRS), Asia Pacific Heart Rhythm Society (APHRS), South African Heart (SA Heart)
Association and Sociedad Latinoamericana de Estimulación Cardíaca y Electrofisiología
(SOLEACE). Europace. 2017;0:1-21.
352. Coffey S, Cairns BJ, Iung B. The modern epidemiology of heart valve disease. Heart.
2016;102(1):75.
353. de Dassel JL, Ralph AP, Carapetis JR. Controlling acute rheumatic fever and
rheumatic heart disease in developing countries: are we getting closer? Curr Opin Pediatr.
2015;27(1):116-23.
354. Coffey S, Cox B, Williams MJA. Lack of progress in valvular heart disease in the pre–
transcatheter aortic valve replacement era: Increasing deaths and minimal change in mortality
rate over the past three decades. Am Heart J. 2014;167(4):562-7.
355. Wolf PA, Dawber TR, Thomas HE, Kannel WB. Epidemiologic assessment of chronic
atrial fibrillation and risk of stroke: The Framingham Study. Neurology. 1978;28(10):973-.
357
356. Oldgren J, Healey JS, Ezekowitz M, Commerford P, Avezum A, Pais P, et al. Variations
in etiology and management of atrial fibrillation in a prospective registry of 15,400 emergency
department patients in 46 countries: the RE-LY AF registry. Circulation. 2014;129(15):1568-
76.
357. Thomas KL, Jackson LR, Shrader P, Ansell J, Fonarow GC, Gersh B, et al.
Prevalence, Characteristics, and Outcomes of Valvular Heart Disease in Patients With Atrial
Fibrillation: Insights From the ORBIT-AF (Outcomes Registry for Better Informed Treatment
for Atrial Fibrillation). J Am Heart Assoc. 2017;6(12).
358. Baumgartner H, Falk V, Bax JJ, De Bonis M, Hamm C, Holm PJ, et al. 2017
ESC/EACTS Guidelines for the management of valvular heart disease. Eur Heart J.
2017;38(36):2739-91.
359. Nishimura RA, Otto CM, Bonow RO, Carabello BA, Erwin JP, Fleisher LA, et al. 2017
AHA/ACC focused update of the 2014 AHA/ACC guideline for the management of patients
with valvular heart disease: a report of the American College of Cardiology/American Heart
Association Task Force on Clinical Practice Guidelines. Circulation. 2017;135(25):e1159-e95.
360. Whitlock RP, Sun JC, Fremes SE, Rubens FD, Teoh KH. Antithrombotic and
thrombolytic therapy for valvular disease: antithrombotic therapy and prevention of
thrombosis: American College of Chest Physicians Evidence-Based Clinical Practice
Guidelines. Chest. 2012;141(2):e576S-e600S.
361. Keeling D, Baglin T, Tait C, Watson H, Perry D, Baglin C, et al. Guidelines on oral
anticoagulation with warfarin - fourth edition. Br J Haematol. 2011;154(3):311-24.
362. Cannegieter SC, Rosendaal FR, Briët E. Thromboembolic and bleeding complications
in patients with mechanical heart valve prostheses. Circulation. 1994;89(2):635.
363. Massel DR, Little SH. Antiplatelet and anticoagulation for patients with prosthetic heart
valves. The Cochrane database of Sys Rev. 2013(7):CD003464.
364. Eikelboom JW, Connolly SJ, Brueckmann M, Granger CB, Kappetein AP, Mack MJ, et
al. Dabigatran versus warfarin in patients with mechanical heart valves. N Engl J Med.
2013;369(13):1206-14.
365. Heras M, Chesebro JH, Fuster V, Penny WJ, Grill DE, Bailey KR, et al. High risk of
thromboemboli early after bioprosthetic cardiac valve replacement. J Am Coll Cardiol.
1995;25(5):1111-9.
366. Cohn LH, Mudge GH, Pratter F, Collins Jr JJ. Five to eight-year follow-up of patients
undergoing porcine heart-valve replacement. N Engl J Med. 1981;304(5):258-62.
367. Durães AR, Durães MAO, Correia LCL, Aras R. Antithrombotic strategy in the three
first months following bioprosthetic heart valve implantation. Arq Bras Cardiol.
2013;101(5):466-72.
358
368. El-Husseiny M, Salhiyyah K, Raja SG, Dunning J. Should warfarin be routinely
prescribed for the first three months after a bioprosthetic valve replacement? Interact
Cardiovasc Thorac Surg. 2006;5(5):616-23.
369. Iung B, Rodes-Cabau J. The optimal management of anti-thrombotic therapy after
valve replacement: certainties and uncertainties. Eur Heart J. 2014;35(42):2942-9.
370. Mérie C, Køber L, Olsen PS, Andersson C, Gislason G, Jensen JS, et al. Association
of warfarin therapy duration after bioprosthetic aortic valve replacement with risk of mortality,
thromboembolic complications, and bleeding. JAMA. 2012;308(20):2118-25.
371. Brennan JM, Edwards FH, Zhao Y, O'Brien S, Booth ME, Dokholyan RS, et al. Early
anticoagulation of bioprosthetic aortic valves in older patients: results from the Society of
Thoracic Surgeons Adult Cardiac Surgery National Database. J Am Coll Cardiol.
2012;60(11):971-7.
372. De Caterina R, Renda G, Carnicelli AP, Nordio F, Trevisan M, Mercuri MF, et al.
Valvular heart disease patients on edoxaban or warfarin in the ENGAGE AF-TIMI 48 trial. J
Am Coll Cardiol. 2017;69(11):1372-82.
373. Carnicelli AP, De Caterina R, Halperin JL, Renda G, Ruff CT, Trevisan M, et al.
Edoxaban for the Prevention of Thromboembolism in Patients With Atrial Fibrillation and
Bioprosthetic Valves. Circulation. 2017;135(13):1273-5.
374. Renda G, De Caterina R, Carnicelli A, Nordio F, Mercuri M, Ruff C, et al. Outcomes in
2824 patients with valvular heart disease treated with edoxaban or warfarin in the ENGAGE
AF-TIMI 48 trial. J Am Coll Cardiol. 2016;67(13 Supplement):2194.
375. Duraes AR, de Souza Roriz P, de Almeida Nunes B, Albuquerque FP, de Bulhoes FV,
de Souza Fernandes AM, et al. Dabigatran Versus Warfarin After Bioprosthesis Valve
Replacement for the Management of Atrial Fibrillation Postoperatively: DAWA Pilot Study.
Drugs R D. 2016;16(2):149-54.
376. Pokorney SD, Rao MP, Wojdyla DM, Gersh BJ, Lopes RD, Lewis BS, et al. Apixaban
use in patients with atrial fibrillation with bioprosthetic valves: insights from ARISTOTLE.
Circulation. 2015;132:A172277.
377. Poli D, Antonucci E, Pengo V, Migliaccio L, Testa S, Lodigiani C, et al. Mechanical
prosthetic heart valves: Quality of anticoagulation and thromboembolic risk. The observational
multicenter PLECTRUM study. Int J Cardiol. 2018;267:68-73.
378. Havers-Borgersen E, Haider Butt J, Vinding NE, Torp-Pedersen C, Gislason G, Koeber
L, et al. P4512Time in therapeutic range and risk of thromboembolism and bleeding in patients
with mechanical heart valve prosthesis. Eur Heart J. 2018;39(suppl_1):ehy563.P4512-
ehy563.P.
359
379. Grzymala-Lubanski B, Labaf A, Englund E, Svensson PJ, Själander A. Mechanical
heart valve prosthesis and warfarin–Treatment quality and prognosis. Thromb Res.
2014;133(5):795-8.
380. Grzymala-Lubanski B, Svensson PJ, Renlund H, Jeppsson A, Sjalander A. Warfarin
treatment quality and prognosis in patients with mechanical heart valve prosthesis. Heart.
2017;103(3):198-203.
381. Butchart EG, Payne N, Li HH, Buchan K, Mandana K, Grunkemeier GL. Better
anticoagulation control improves survival after valve replacement. J Thorac Cardiovasc Surg.
2002;123(4):715-23.
382. Smith DE, Xuereb CB, Pattison HM, Lip GYH, Lane DA. TRial of an Educational
intervention on patients' knowledge of Atrial fibrillation and anticoagulant therapy, INR control,
and outcome of Treatment with warfarin (TREAT). BMC Cardiovasc Disord. 2010;10(1):21.
383. Lip GYH, Kamath S, Jafri M, Mohammed A, Bareford D. Ethnic differences in patient
perceptions of atrial fibrillation and anticoagulation therapy the West Birmingham atrial
fibrillation project. Stroke. 2002;33(1):238-42.
384. Nadar S, Begum N, Kaur B, Sandhu S, Lip GYH. Patients' understanding of
anticoagulant therapy in a multiethnic population. J R Soc Med. 2003;96(4):175-9.
385. Amara W, Larsen TB, Sciaraffia E, Hernández Madrid A, Chen J, Estner H, et al.
Patients’ attitude and knowledge about oral anticoagulation therapy: results of a self-
assessment survey in patients with atrial fibrillation conducted by the European Heart Rhythm
Association. Ep Europace. 2015;18(1):151-5.
386. Hendriks JML, Vrijhoef HJM, Crijns HJGM, Brunner-La Rocca HP. The effect of a
nurse-led integrated chronic care approach on quality of life in patients with atrial fibrillation.
Europace. 2014;16(4):491-9.
387. Clarkesmith DE, Pattison HM, Khaing PH, Lane DA. Educational and behavioural
interventions for anticoagulant therapy in patients with atrial fibrillation. Cochrane Database
Syst Rev. 2017;4:Cd008600.
388. Morisky DE, Green LW, Levine DM. Concurrent and predictive validity of a self-
reported measure of medication adherence. Med Care. 1986;24(1):67-74.
389. Lam WY, Fresco P. Medication Adherence Measures: An Overview. Biomed Res Int
2015;2015:217047-.
390. Warren SR, Raisch DW, Campbell HM, Guarino PD, Kaufman JS, Petrokaitis E, et al.
Medication adherence assessment in a clinical trial with centralized follow-up and direct-to-
patient drug shipments. Clinical Trials. 2011;10(3):441-8.
391. Jayaraman S, Rieder MJ, Matsui DM. Compliance assessment in drug trials: has there
been improvement in two decades? Can J Clin Pharmacol. 2005;12(3):e251-3.
360
392. Schulman S, Kearon C, the SOCOAOTS, Standardization Committee Of The
International Society On T, Haemostasis. Definition of major bleeding in clinical investigations
of antihemostatic medicinal products in non-surgical patients. J Thromb Haemost.
2005;3(4):692-4.
393. Hohnloser SH, Shestakovska O, Eikelboom J, Franzosi MG, San Tan R, Zhu J, et al.
The effects of apixaban on hospitalizations in patients with different types of atrial fibrillation:
insights from the AVERROES trial. Eur Heart J. 2013;34(35):2752-9.
394. Ivanovs R, Kivite A, Ziedonis D, Mintale I, Vrublevska J, Rancans E. Association of
depression and anxiety with cardiovascular co-morbidity in a primary care population in Latvia:
a cross-sectional study. BMC Public Health. 2018;18(1):328-.
395. Haddad M, Walters P, Phillips R, Tsakok J, Williams P, Mann A, et al. Detecting
depression in patients with coronary heart disease: a diagnostic evaluation of the PHQ-9 and
HADS-D in primary care, findings from the UPBEAT-UK study. PLoS One.
2013;8(10):e78493.
396. Meader N, Mitchell AJ, Chew-Graham C, Goldberg D, Rizzo M, Bird V, et al. Case
identification of depression in patients with chronic physical health problems: a diagnostic
accuracy meta-analysis of 113 studies. Br J Gen Pract. 2011;61(593):e808-20.
397. Hendriks JML, Crijns HJGM, Tieleman RG, Vrijhoef HJM. The atrial fibrillation
knowledge scale: Development, validation and results. Int J Cardiol. 2013;168(2):1422-8.
398. Horne R, Weinman J, Hankins M. The beliefs about medicines questionnaire: The
development and evaluation of a new method for assessing the cognitive representation of
medication. Psychol & Health. 1999;14(1):1-24.
399. Kotecha D, Ahmed A, Calvert M, Lencioni M, Terwee CB, Lane DA. Patient-reported
outcomes for quality of life assessment in atrial fibrillation: a systematic review of
measurement properties. PLoS One. 2016;11(11):e0165790.
400. Kroenke K, Spitzer RL, Williams JBW. The Phq-9: Validity of a Brief Depression
Severity Measure. J Gen Intern Med. 2001;16(9):606-13.
401. Spitzer RL, Kroenke K, Williams JW, Löwe B. A brief measure for assessing
generalized anxiety disorder: The gad-7. Arch Intern Med. 2006;166(10):1092-7.
402. National Institute for Health and Care Excellence (2018) Depression in adults:
treatment and management. NICE guideline.
403. National Institute for Health and Care Excellence (2018) Identifying and addressing
common mental health disorders. NICE Pathways
404. Conway A, Sheridan J, Maddicks-Law J, Fulbrook P, Ski CF, Thompson DR, et al.
Accuracy of anxiety and depression screening tools in heart transplant recipients. Appl Nurs
Res. 2016;32:177-81.
361
405. Spertus J, Dorian P, Bubien R, Lewis S, Godejohn D, Reynolds MR, et al.
Development and validation of the Atrial Fibrillation Effect on QualiTy-of-Life (AFEQT)
Questionnaire in patients with atrial fibrillation. Circ Arrhythm Electrophysiol. 2011;4(1):15-25.
406. IBM SPSS Statistics for Windows, Version 23. Armonk, NY:IBM Corp.
407. Thrall G, Lane D, Carroll D, Lip GYH. Quality of Life in Patients with Atrial Fibrillation:
A Systematic Review. Am J Med. 2006;119(5):448.e1-.e19.
408. Galli F, Borghi L, Carugo S, Cavicchioli M, Faioni EM, Negroni MS, et al. Atrial
fibrillation and psychological factors: a systematic review. PeerJ. 2017;5:e3537-e.
409. Perret-Guillaume C, Briancon S, Wahl D, Guillemin F, Empereur F. Quality of life in
elderly inpatients with atrial fibrillation as compared with controlled subjects. J Nutr Health
Aging. 2010;14(2):161-6.
410. Lioni L, Vlachos K, Letsas KP, Efremidis M, Karlis D, Asvestas D, et al. Differences in
Quality of Life, Anxiety and Depression in Patients with Paroxysmal Atrial Fibrillation and
Common Forms of Atrioventricular Reentry Supraventricular Tachycardias. Indian Pacing
Electrophysiol J. 2014;14(5):250-7.
411. Schnabel RB, Michal M, Wilde S, Wiltink J, Wild PS, Sinning CR, et al. Depression in
atrial fibrillation in the general population. PLoS One. 2013;8(12):e79109.
412. Jamous RM, Sweileh WM, El-Deen Abu Taha AS, Zyoud SeH. Beliefs About
Medicines and Self-reported Adherence Among Patients with Chronic Illness: A Study in
Palestine. J Family Med Prim Care. 2014;3(3):224-9.
413. AlHewiti A. Adherence to Long-Term Therapies and Beliefs about Medications. Int J
Family Med. 2014;2014(479596).
414. Piccini JP, Simon DN, Steinberg BA, et al. Differences in clinical and functional
outcomes of atrial fibrillation in women and men: Two-year results from the orbit-af registry.
JAMA Cardiology. 2016;1(3):282-91.
415. Ferguson C, Inglis SC, Newton PJ, Middleton S, Macdonald PS, Davidson PM.
Education and practice gaps on atrial fibrillation and anticoagulation: a survey of
cardiovascular nurses. BMC Med Educ. 2016;16(1):9.
416. Mayet AY. Association between oral anticoagulation knowledge, anticoagulation
control, and demographic characteristics of patients attending an anticoagulation clinic in
Saudi Arabia: a cross-sectional prospective evaluation. Trop J Pharm Res. 2015;14(7):1285-
91.
417. Alrasheedy AA, Hassali MA, Wong ZY, Saleem F. Pharmacist-managed medication
therapy adherence clinics: The Malaysian experience. Res Social Adm Pharm.
2017;13(4):885-6.
362
418. Thanimalai S, Shafie AA, Hassali MA, Sinnadurai J. Comparing effectiveness of two
anticoagulation management models in a Malaysian tertiary hospital. Int J Clin Pharm.
2013;35(5):736-43.
419. Aidit S, Soh YC, Yap CS, Khan TM, Neoh CF, Shaharuddin S, et al. Effect of
Standardized Warfarin Treatment Protocol on Anticoagulant Effect: Comparison of a Warfarin
Medication Therapy Adherence Clinic with Usual Medical Care. Front Pharmacol. 2017;8:637-
.
420. Kaatz S. Determinants and measures of quality in oral anticoagulation therapy. J
Thromb Thrombolysis. 2007;25(1):61-6.
421. Rosendaal FR, Cannegieter SC, Van der Meer FJ, Briet E. A method to determine the
optimal intensity of oral anticoagulant therapy. Thromb Haemost. 1993;69(3):236-9.
422. Poller L, Keown M, Ibrahim S, Lowe G, Moia M, Turpie AG, et al. An international
multicenter randomized study of computer-assisted oral anticoagulant dosage vs. medical
staff dosage. J Thromb Haemost. 2008;6(6):935-43.
423. Fauchier L, Angoulvant D, Lip GY. The SAMe-TT2R2 score and quality of
anticoagulation in atrial fibrillation: a simple aid to decision-making on who is suitable (or not)
for vitamin K antagonists. Europace. 2015;17(5):671-3.
424. National Institute for Clinical Excellence (2014) Chronic kidney disease in adults:
assessment and management. NICE guideline (CG182).
425. Morisky DE, Green LW, Levine DM. Concurrent and predictive validity of a self-
reported measure of medication adherence. Med Care. 1986;24(1):67-74.
426. Optimal Oral Anticoagulant Therapy in Patients with Nonrheumatic Atrial Fibrillation
and Recent Cerebral Ischemia. N Engl J Med. 1995;333(1):5-10.
427. Cannegieter SC, Rosendaal FR, Wintzen AR, van der Meer FJM, Vandenbroucke JP,
Briet E. Optimal Oral Anticoagulant Therapy in Patients with Mechanical Heart Valves. N Engl
J Med. 1995;333(1):11-7.
428. Perera MA, Gamazon E, Cavallari LH, Patel SR, Poindexter S, Kittles RA, et al. The
missing association: sequencing-based discovery of novel SNPs in VKORC1 and CYP2C9
that affect warfarin dose in African Americans. Clin Pharmacol Ther. 2011;89(3):408-15.
429. Lindh JD, Holm L, Andersson ML, Rane A. Influence of CYP2C9 genotype on warfarin
dose requirements—a systematic review and meta-analysis. Eur J Clin Pharmacol.
2009;65(4):365-75.
430. Limdi NA, McGwin G, Goldstein JA, Beasley TM, Arnett DK, Adler BK, et al. Influence
of CYP2C9 and VKORC1 1173C/T Genotype on the Risk of Hemorrhagic Complications in
African-American and European-American Patients on Warfarin. Clin Pharmacol Ther.
2008;83(2):312-21.
363
431. Higashi MK, Veenstra DL, Kondo L, et al. Association between cyp2c9 genetic variants
and anticoagulation-related outcomes during warfarin therapy. JAMA. 2002;287(13):1690-8.
432. Aithal GP, Day CP, Kesteven PJL, Daly AK. Association of polymorphisms in the
cytochrome P450 CYP2C9 with warfarin dose requirement and risk of bleeding complications.
The Lancet. 1999;353(9154):717-9.
433. Sistonen J, Fuselli S, Palo JU, Chauhan N, Padh H, Sajantila A. Pharmacogenetic
variation at CYP2C9, CYP2C19, and CYP2D6 at global and microgeographic scales.
Pharmacogenet Genomics. 2009;19(2):170-9.
434. Solus JF, Arietta BJ, Harris JR, Sexton DP, Steward JQ, McMunn C, et al. Genetic
variation in eleven phase I drug metabolism genes in an ethnically diverse population.
Pharmacogenomics. 2004;5(7):895-931.
435. Lee CR, Goldstein JA, Pieper JA. Cytochrome P450 2C9 polymorphisms: a
comprehensive review of the in-vitro and human data. Pharmacogenet Genomics. 2002;12(3).
436. Johnson J, Caudle K, Gong L, Whirl-Carrillo M, Stein C, Scott S, et al. Clinical
Pharmacogenetics Implementation Consortium (CPIC) Guideline for Pharmacogenetics-
Guided Warfarin Dosing: 2017 Update.
437. Pirmohamed M, Burnside G, Eriksson N, Jorgensen AL, Toh CH, Nicholson T, et al. A
Randomized Trial of Genotype-Guided Dosing of Warfarin. N Engl J Med. 2013;369(24):2294-
303.
438. Kimmel SE, French B, Kasner SE, Johnson JA, Anderson JL, Gage BF, et al. A
Pharmacogenetic versus a Clinical Algorithm for Warfarin Dosing. N Engl J Med.
2013;369(24):2283-93.
439. Limdi NA, Brown TM, Yan Q, Thigpen JL, Shendre A, Liu N, et al. Race influences
warfarin dose changes associated with genetic factors. Blood. 2015;126(4):539-45.
440. Hart RG, Pearce LA, Asinger RW, Herzog CA. Warfarin in Atrial Fibrillation Patients
with Moderate Chronic Kidney Disease. Clin J Am Soc Nephrol. 2011.
441. Jones MR, Horner RD, Edwards LJ, Hoff J, Armstrong SB, Smith-Hammond CA, et al.
Racial Variation in Initial Stroke Severity. Stroke. 2000;31(3):563.
442. Sacco RL, Boden-Albala B, Gan R, Chen X, Kargman DE, Shea S, et al. Stroke
Incidence among White, Black, and Hispanic Residents of an Urban CommunityThe Northern
Manhattan Stroke Study. Am J Epidemiol. 1998;147(3):259-68.
443. Bush D, Martin LW, Leman R, Chandler M, Haywood LJ, Investigators NA. Atrial
fibrillation among African Americans, Hispanics and Caucasians: clinical features and
outcomes from the AFFIRM trial. J Natl Med Assoc. 2006;98(3):330.
444. Pokorney SD, Simon DN, Thomas L, Fonarow GC, Kowey PR, Chang P, et al. Patients'
time in therapeutic range on warfarin among US patients with atrial fibrillation: Results from
ORBIT-AF registry. Am Heart J.170(1):141-8.e1.
364
445. Apostolakis S, Guo Y, Lane DA, Buller H, Lip GYH. Renal function and outcomes in
anticoagulated patients with non-valvular atrial fibrillation: the AMADEUS trial. Eur Heart J.
2013;34(46):3572-9.
446. Soliman EZ, Prineas RJ, Go AS, Xie D, Lash JP, Rahman M, et al. Chronic kidney
disease and prevalent atrial fibrillation: the Chronic Renal Insufficiency Cohort (CRIC). Am
Heart J. 2010;159(6):1102-7.
447. Go AS, Fang MC, Udaltsova N, Chang Y, Pomernacki NK, Borowsky L, et al. Impact
of proteinuria and glomerular filtration rate on risk of thromboembolism in atrial fibrillation: the
anticoagulation and risk factors in atrial fibrillation (ATRIA) study. Circulation.
2009;119(10):1363-9.
448. Zimmerman D, Sood MM, Rigatto C, Holden RM, Hiremath S, Clase CM. Systematic
review and meta-analysis of incidence, prevalence and outcomes of atrial fibrillation in patients
on dialysis. Nephrol Dial Transplant. 2012;27(10):3816-22.
449. Ansell J, Hollowell J, Pengo V, Martinez-Brotons F, Caro J, Drouet L. Descriptive
analysis of the process and quality of oral anticoagulation management in real-life practice in
patients with chronic non-valvular atrial fibrillation: the international study of anticoagulation
management (ISAM). J Thromb Thrombolysis. 2007;23(2):83-91.
450. Fitzmaurice DA, Hobbs F, Murray ET, Holder RL, Allan TF, Rose PE. Oral
anticoagulation management in primary care with the use of computerized decision support
and near-patient testing: A randomized, controlled trial. Arch Intern Med. 2000;160(15):2343-
8.
451. Ming LC. Use of herbal products in Southeast Asian countries. Arch Pharm Prac.
2016;7(5):1.
452. Ekor M. The growing use of herbal medicines: issues relating to adverse reactions and
challenges in monitoring safety. Front Pharmacol. 2014;4:177.
453. Biswal BM, Sulaiman SA, Ismail HC, Zakaria H, Musa KI. Effect of Withania somnifera
(Ashwagandha) on the Development of Chemotherapy-Induced Fatigue and Quality of Life in
Breast Cancer Patients. Integr Cancer Ther. 2012;12(4):312-22.
454. Butchart EG, Gohlke-Bärwolf C, Antunes MJ, Tornos P, De Caterina R, Cormier B, et
al. Recommendations for the management of patients after heart valve surgery. Eur Heart J.
2005;26(22):2463-71.
455. Grunkemeier GL, Wu Y. “Our complication rates are lower than theirs”: Statistical
critique of heart valve comparisons. J Thorac Cardiovasc Surg. 2003;125(2):290-300.
456. Vestergaard AS, Skjøth F, Larsen TB, Ehlers LH. The importance of mean time in
therapeutic range for complication rates in warfarin therapy of patients with atrial fibrillation: A
systematic review and meta-regression analysis. PLoS One. 2017;12(11):e0188482.
365
457. Blustin JM, McBane RD, Ketha SS, Wysokinski WE. Distribution of thromboembolism
in valvular versus non-valvular atrial fibrillation. Expert Rev Cardiovasc Ther.
2014;12(10):1129-32.
458. Manteuffel M, Williams S, Chen W, Verbrugge RR, Pittman DG, Steinkellner A.
Influence of patient sex and gender on medication use, adherence, and prescribing alignment
with guidelines. J Womens Health (Larchmt). 2014;23(2):112-9.
459. Heneghan C, Ward A, Perera R, Bankhead C, Fuller A, Stevens R, et al. Self-
monitoring of oral anticoagulation: systematic review and meta-analysis of individual patient
data. Lancet. 2012;379(9813):322-34.
460. Razouki Z, Ozonoff A, Zhao S, Jasuja Guneet K, Rose Adam J. Improving Quality
Measurement for Anticoagulation. Circ Cardiovasc Qual Outcomes. 2014;7(5):664-9.
461. Population Distribution and Basic Demographics Characteristics Report 2010
(Updated: 05/08/2011). Department of Statistics Malaysia.
462. Pastori D, Farcomeni A, Saliola M, Del Sole F, Pignatelli P, Violi F, et al. Temporal
trends of time in therapeutic range and incidence of cardiovascular events in patients with
non-valvular atrial fibrillation. Eur J Intern Med. 2018;54:34-9.
366

Zulkifly 2019 PH D

Uploaded by

Copyright:

Available Formats

Zulkifly 2019 PH D

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Zulkifly 2019 PH D

Uploaded by

Copyright:

Available Formats

ANTICOAGULATION CONTROL IN PATIENTS WITH ATRIAL FIBRILLATION

HANIS HANUM BINTI ZULKIFLY

A thesis submitted to the University of Birmingham for the degree of

Institute of Cardiovascular Sciences

Study 1 (TREAT-2) examined self-report assessment of depression, anxiety, beliefs about

Publications arising from thesis ____________________________________________________ 3

List of tables ___________________________________________________________________ 6

List of figures __________________________________________________________________ 10

List of abbreviations ____________________________________________________________ 12

Table of Contents ______________________________________________________________ 14

Chapter 1. Literature review _______________________________________________ 18

1.1 Epidemiology of atrial fibrillation ___________________________________________ 18

1.2 Overview of atrial fibrillation _______________________________________________ 30

1.4 Stroke prevention in atrial fibrillation ________________________________________ 45

1.5 Anticoagulation control in stroke prevention __________________________________ 78

1.8 Aims and objectives _____________________________________________________ 121

Chapter 2. A prospective study examining non-vitamin K antagonist oral anticoagulants

2.1 Abstract _______________________________________________________________ 122

2.2 Background and rationale ________________________________________________ 125

2.3 Methods ______________________________________________________________ 128

2.4 Results ________________________________________________________________ 141

2.5 Discussion _____________________________________________________________ 164

2.6 Conclusion _____________________________________________________________ 170

Chapter 3. Anticoagulation control in different ethnic groups receiving vitamin K

3.1 Abstract _______________________________________________________________ 171

3.2 Introduction ___________________________________________________________ 173

3.3 Methods ______________________________________________________________ 174

3.4 Results ________________________________________________________________ 184

3.5 Discussion _____________________________________________________________ 223

3.6 Conclusions ____________________________________________________________ 236

4.2 Introduction ___________________________________________________________ 240

4.3 Methods ______________________________________________________________ 241

4.4 Results ________________________________________________________________ 247

4.5 Discussion _____________________________________________________________ 265

4.6 Conclusion _____________________________________________________________ 272

Chapter 5. General discussion and conclusions ________________________________ 273

Appendix 1 __________________________________________________________________ 280

Appendix 2 __________________________________________________________________ 282

Appendix 3 __________________________________________________________________ 295

Appendix 4 __________________________________________________________________ 299

Appendix 5 __________________________________________________________________ 311

1.1 Epidemiology of atrial fibrillation

greater risk of hospital admission, with 10-40% of AF patients hospitalised annually.

Additionally, irrespective of other cardiovascular-related conditions, AF patients have poorer

develop vascular dementia and a decline in cognitive function (3).

1.1.1 Incidence and prevalence of atrial fibrillation worldwide

1990 2010 1990 2010

-Years data Study population

Singapore Yap et al † 1,839 Community-based ≥55 1.4 (†)

quality of data obtained (5).

1.1.3 Ethnicity, age and prevalence of atrial fibrillation

this thesis, ethnicity is classified as Whites, (Europeans, Americans) Afro-Caribbeans,

and others, as reported in individual studies (31, 32).

of 10 studies examining the prevalence of AF among African-Americans compared to Whites

Country/State Study a) Study design Mean (SD) age Prevalence of AF (%)

Michigan, Lahiri et al a) Retrospective African American: 33.0% 29 19 N/A N/A

USA Winkelmayer et al a) Cross sectional 70.9 (11.8) 14 6.5 9.0* N/A

California, Shen et al a) Cross-sectional White: 70 (64-77) 8 3.8 3.9 3.6

15 U.S states, Marcus et al a) Combination of CHS CHS 23 15 N/A N/A