Improving The Outcomes of Anticoagulation in Rural Australia: An Evaluation of Pharmacist-Assisted Monitoring of Warfarin Therapy
Improving The Outcomes of Anticoagulation in Rural Australia: An Evaluation of Pharmacist-Assisted Monitoring of Warfarin Therapy
Improving The Outcomes of Anticoagulation in Rural Australia: An Evaluation of Pharmacist-Assisted Monitoring of Warfarin Therapy
ORIGINAL ARTICLE
anticoagulated patients (3). This is particularly (Australia), Castle Hill, NSW, Australia) INR
relevant in rural or remote settings because of the monitor and given educational material relating to
relative lack of access to pathology services. It is not warfarin (16). The training typically involved
uncommon for doctors from rural or remote approximately 2–3 h with the pharmacists dis-
regions in Australia to have to wait 2 days for cussing anticoagulation and the use of the INR
pathology results. monitor. Pharmacists were shown how to conduct
The key objective of point-of-care (POC) testing INR tests and were also observed conducting tests
is to generate a result quickly so that appropriate on consenting subjects or pharmacy staff. Problems
treatment can be implemented, leading to an or difficulties encountered by the researchers
improved clinical or economic outcome (4). POC [through previous research activities (17, 18) and
testing in pharmacies has two main functions. The personal experience] were raised with the phar-
first is in disease screening or risk assessment (e.g. macists and potential solutions to these difficulties
blood glucose or cholesterol determination). The were discussed. The pharmacists were provided
second is in the monitoring of chronic diseases and with a laminated colour brochure on the INR
the effects of medicines used to manage these dis- monitor, and ongoing assistance if needed. Phar-
eases, as with anticoagulant therapy. macists were provided with INR monitors, tests
The increasing number of patients on warfarin strips and other consumables free of charge for the
and concerns over the ability of conventional health duration of the trial.
services to cope is one of the reasons for the Local GPs were visited, informed of the availab-
expansion of POC testing and moves to find alter- ility of the CoaguChek S monitor in their region, and
native models of service provision within primary were invited to refer their patients to the pharmacy
care for anticoagulated patients. Many models of for POC testing. During the visits to the GPs, the
anticoagulant management have been proposed in accuracy of the CoaguChek S monitor (17, 18) and its
the literature: innovative models, such as commu- use in several overseas countries was discussed.
nity and general practice-based anticoagulant clin- Patients referred to the pharmacy or who were
ics run by nurses (5) and pharmacists (6–13), have identified as taking warfarin were given an infor-
been proposed as solutions to reducing the risks of mation sheet and gave written informed consent to
anticoagulant misadventure (14). Very few studies undergo fingerprick testing at the pharmacy.
have studied the effect of community pharmacists’ Patients could have two types of testing performed
involvement in anticoagulant management (15). in the pharmacy: comparison testing was defined
The aim of the project was to assess whether as a pharmacy-based test taken within 4 h of con-
rural pharmacist involvement in the management ventional laboratory testing, and additional testing
of targeted ‘high-risk’ patients (i.e. those receiving was a pharmacy-based test with no direct com-
warfarin therapy) has the potential to lead to safer parison laboratory test taken. All results were sent
and more effective anticoagulation, and is valued to the patient’s GP via a specially designed fax
and welcomed by patients and their general prac- form. The results of the testing, such as INR, time
titioners (GPs). taken, outcome of test (dosage changes) were
recorded. It was recommended to pharmacists that
all results were recorded for patients in the stand-
METHODS
ard warfarin educational booklet.
A convenience sample of rural pharmacists was Pharmacists and GPs were instructed that this
identified by the researchers through a composite type of testing was not to replace conventional
of previous research activities with the research pathology testing. The service was offered free of
team and via contact through an electronic mailing charge to patients for the duration of the trial.
list (AusPharmList). Pharmacies needed to have an Pharmacies were remunerated at a rate of $4 per
area where POC testing within the pharmacy could test for the duration of the trial.
be reasonably completed privately during normal The CoaguChek S INR monitor was left with
workflow. each participating pharmacy for approximately
The pharmacists were trained in the use 3 months. GPs and pharmacists were later sent
of the CoaguChek S (Roche Diagnostics Pty Ltd an anonymous questionnaire (using a visual
2005 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 30, 345–353
Rural pharmacy monitoring of warfarin 347
CoaguChek S INR
4
centiles were plotted to represent the responses
to the evaluation questionnaires. Participating 3
patients were given an anonymous satisfaction
survey from community pharmacists after the 2
A total of 22 pharmacies were identified and invi- done’ and ‘lack of time (at present) to implement
ted to participate in the project. Sixteen pharmacies the service.’
agreed to participate and were visited by one of the Characteristics of the patients and the test results
authors (SLJ) and trained to use the CoaguChek S are shown in Table 1. A total of 518 tests were
monitor. Three pharmacies did not conduct any conducted in the pharmacies and 137 different
testing during the trial period. Reasons cited for patients were tested. Over three-quarters of the
testing not being performed included a ‘lack of pharmacy-based tests were taken in addition to
interest by patients in having additional testing conventional laboratory testing.
The majority of tests (67Æ0%) were in the expan-
Table 1. Characteristics of pharmacy-based Interna- ded therapeutic range of 2Æ0–3Æ5, with over one-
tional Normalized Ratio (INR) testing and patients quarter (27Æ8%) being £1Æ9 and the remainder
involved (5Æ2%) were ‡3Æ6. A total of 120 pharmacy-based
INR tests had comparison laboratory tests taken for
Testing characteristics Result which the results were made available to the
research team. The mean (SD) INR values for the
No. of tests conducted 518
Tests per pharmacy 24 (0–171)
No. of patients tested 137 1.5
Tests per patient 2 (1–30)
Age of patient (years) 72 (23–100)
CoaguChek S INR - Laboratory INR
1.0
Test duration (minutes) 5 (2–15)
Reason for use of warfarin (N = 122*) 0.5
Atrial fibrillation 63 (52)
Valve replacement 20 (16) 0.0
Type of testing
–1.5
Comparison with conventional 120 (23) 0 1 2 3 4 5 6
laboratory testing Mean of CoaguChek S INR and Laboratory INR
Additional testing in the pharmacy 398 (77)
Fig. 2. Bland–Altman style bias plot for CoaguChek S
Values are presented as median (range) or n (%). and laboratory International Normalized Ratio (INR)
*As reason for use not recorded in some cases. values.
2005 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 30, 345–353
348 S. L. Jackson et al.
Table 2. Comparison of International Normalized Ratio I found this to be a valuable service provided to my patient(s)
(INR) categories for CoaguChek S and laboratory results
(values given as percentage of laboratory readings)
Strongly Strongly
agree disagree
CoaguChek S INR
I would feel comfortable operating this service if it was ongoing
I believe that more patients would benefit from this type of service
Table 3. Outcomes of additional testing conducted in
the pharmacy
Strongly Strongly
INR Result (%) Dosage Result (%) agree disagree
range (n = 398) changes (n = 390*)
I believe that this service would increase the compliance of patients on warfarin
Strongly Strongly
I found this to be a valuable service provided to my patient(s) agree disagree
2005 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 30, 345–353
Rural pharmacy monitoring of warfarin 349
2005 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 30, 345–353
350 S. L. Jackson et al.
The responses from patients to the satisfaction many potential users. The performance of the
questionnaire are displayed in Table 4. Responses CoaguChek S in the community pharmacy setting
were received from 62 patients, giving a response was excellent, especially when considering that data
rate from patients tested of 45%. Most of the were collected from many sites and was compared
comments from patients reflected an appreciation with different laboratories. Previous studies have
of the convenience with using pharmacy-based found that the variation between portable INR
testing. They seemed to be very happy with POC monitors and laboratory was not larger than the
monitoring but perhaps found it a barrier to util- variation encountered between different laborator-
ization of the service that pharmacists were unable ies measuring a single sample (24). To our know-
to adjust warfarin dosages on the basis of these ledge, this is the first published study comparing the
results. Some patients expressed views that gov- accuracy of the CoaguChek S INR monitor in com-
ernments or other organizations should pay for this parison with laboratory testing in a sample of com-
type of monitoring. munity pharmacies.
In nearly 10% of cases the additional testing
conducted in the pharmacy resulted in a change in
DISCUSSION
warfarin dosage. Clinical decisions were made
This project was designed to test the feasibility of (generally followed by pathology testing in the next
community pharmacy-based INR monitoring in day or so) on the basis of INR tests in the phar-
rural Australia. We also wanted to gauge the macy. It is difficult to quantify the impact that the
acceptance of this type of service by community changes in therapy had on clinical outcomes. In
pharmacists, GPs and patients. We recruited a rel- most cases, pharmacists recorded only dosage
atively large number of pharmacies (16), with a changes and not clinical outcomes, although the
broad spread or rurality, for a study of this type study was not designed to assess the latter. How-
and the pharmacists involved conducted a large ever, it is likely that the availability of POC testing
number of tests (518) on 137 patients. in community pharmacies may have a large impact
Use of the CoaguChek S monitor in community on clinical outcomes if pharmacy-based testing is
pharmacies resulted in accurate performance com- utilized in rural areas. Further research needs to be
pared with conventional laboratory testing. Previ- conducted on the impact of community pharmacy-
ous studies by the research team have obtained conducted INR monitoring on patient care and
correlation coefficients (r) of 0Æ90 in an anticoagula- outcomes.
tion clinic (17) and r = 0Æ89 in general practices (18), The quality of the INR monitoring service was
compared with r = 0Æ88 in this study. Other studies rated as good or excellent by nearly all patients.
have shown that the CoaguChek (previous version Importantly, the majority of patients indicated that
of the INR monitor) and CoaguChek S devices pro- the pharmacy-based monitoring helped them deal
duce INR values that are highly correlated with more effectively with warfarin. This emphasizes
laboratory INR values (r = 0Æ91–0Æ97) (20–25). the ability of pharmacists to educate patients
Eighty-five per cent of all dual measurements regarding their anticoagulant therapy, and in fact,
were within 0Æ5 INR units in this study, which the effects of education delivered through phar-
compares well to the figure of 79% reported by macy-based INR monitoring may have larger
Douketis et al. (25), and 83% by the authors in an long-term impacts on compliance and anticoagu-
outpatient anticoagulant clinic (17). This study also lant-related misadventure (7). It is a limitation of
found that 76% of all comparison tests were within the study and room for further research that nearly
10% of the laboratory value. This compares one-third of patients indicated that there was more
exceptionally well with other studies conducted by information that they would like about warfarin.
the research team, which have shown results of Structured education programmes could be devel-
55% of results within 10% of the laboratory in oped through community pharmacies and could
general practices (18) and 44% in an anticoagula- complement pharmacy-based INR monitoring.
tion clinic (17). Nearly 50% of respondents indicated that they
This study was unique in evaluating the Coagu- found the testing convenient and beneficial.
Chek S in a number of diverse settings and with Patients receiving anticoagulation are often elderly,
2005 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 30, 345–353
Rural pharmacy monitoring of warfarin 351
cannot endure long waiting times, and find trav- identified previously as a concern for implementa-
elling to hospitals or pathology laboratories diffi- tion of professional services (29). It is recognized that
cult (26). Pharmacies are usually located in a number of community pharmacies may have to
convenient and accessible locations, improving undergo structural modifications to cope with the
access for patients. A convenient approach could array of professional services that may be provided
be to provide long-term monitoring through in the future.
accredited community pharmacies. As noted by Studies have found that elderly rural patients
Moffat (27), the pharmacy profession has a great with chronic atrial fibrillation (AF) receive warfarin
opportunity to use diagnostic testing to monitor less frequently than urban patients, despite pos-
patients’ responses in assisting them to manage sessing a similar high-risk profile for stroke and
their chronic conditions and their prescribed fewer relative contraindications (30, 31). POC test-
medicines. Clearly, however, ongoing quality ing of rural and remote Australians taking warfarin
assurance schemes will need to be developed to has the potential to improve the prescribing rates of
ensure pharmacy-based INR monitoring systems warfarin in AF, thereby lessening the risk of debi-
are of a high standard. litating strokes. It also clearly has the potential to
The testing process was well received by the improve the safety of use of warfarin, and be par-
community pharmacists; they felt confident ticularly beneficial to the care of rural and remote
delivering this type of service and found it be a patients.
valuable service for patients. Pharmacists, how- The National Pharmaceutical Association in the
ever, will need to address issues such as the need UK has claimed that ‘community-based pharmacy
for separate counselling and consultation areas anticoagulation clinics could help the National
for testing, and occupational health and safety Health Service save 5000 lives and £24 million a
issues associated with handling bodily fluids. year’ (32). This figure was based on the under-
Responding pharmacists indicated that they utilization of anticoagulants for stroke prevention
thought that pharmacy-based INR monitoring in AF, suggesting a rise to 50% of patients being
would improve the compliance of patients on appropriately treated if community pharmacies
warfarin. In fact, pharmacist-managed anticoagu- were involved in anticoagulant monitoring. If
lation clinics have reported increased compliance similar data from the Australian community were
after instituting education programmes (7). Con- used, documenting the underuse of antithrombot-
sistent with other reports of POC testing by ics for AF (3, 33), and if pharmacy-based antico-
pharmacists (28), the pharmacists in this project agulation clinics were able to assist in the
indicated that testing should be part of a larger management of anticoagulated patients and there-
programme (28), such as comprehensive antico- fore increase the proportion of anticoagulated
agulant education. patients with AF by 10% (from 40 to 50%), this
The evaluation from the GPs was generally pos- would save approximately Aust$12 million dollars
itive, but variations in a number of the responses per annum due to reduced stroke incidence and
were evident. A negative response was observed to would likely reduce the incidence of bleeding
the question ‘I found the comments made by the complications.
pharmacist to be useful.’ This may reflect barriers
between general practice and pharmacists regarding
ACKNOWLEDGEMENTS
perceived encroachment of pharmacy services on
traditional general practice territory (29). More for- We thank the participating community pharma-
mal integration of pharmacists into health care and cists. The project was funded by the Common-
the development of partnerships with GPs are cru- wealth of Australia through the Rural and Remote
cial to the establishment of pharmacy-based INR Pharmacy Infrastructure Grants Program as a
monitoring. Professional barriers to the implemen- component of the Rural and Remote Pharmacy
tation of expanded services in pharmacy have been Workforce Development Program (administered
identified in the literature. These include the ‘shop- by the Pharmacy Guild of Australia). Shane
keeper image’ of community pharmacy (29). The Jackson is an Australian Commonwealth Depart-
design of many community pharmacies has been ment of Health and Aged Care, Quality Use of
2005 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 30, 345–353
352 S. L. Jackson et al.
Medicines Evaluation Program (QUMEP) PhD hospital. American Journal of Health-System Pharmacy,
scholar. Roche Diagnostics Pty Ltd (Australia) 53, 1151–1157.
contributed INR monitors and test strips for the 13. Wilt VM, Gums JG, Ahmed OI, Moore LM (1995)
duration of the project. Outcome analysis of a pharmacist-managed anti-
coagulation service. Pharmacotherapy, 15, 732–739.
14. Ansell J, Hirsh J, Dalen J et al. (2001) Managing oral
REFERENCES anticoagulant therapy (Review). Chest, 119, 22S–38S.
15. Knowlton CH, Thomas OV, Williamson A et al. (1999)
1. National Rural Health Policy Forum (1999) Healthy
Establishing community pharmacy-based anticoagu-
horizons: a framework for improving the health of
lation education and monitoring programs. Journal of
rural, regional and remote Australians. Canberra:
the American Pharmaceutical Association, 39, 368–374.
National Rural Health Policy Forum.
16. Hirsh J, Fuster V, Ansell J, Halperin JL (2003)
2. Australian Institute of Health and Welfare (1998)
American Heart Association/American College of
Health in rural and remote Australia. In: AIHW Cat
Cardiology Foundation guide to warfarin therapy.
no. PHE6. Canberra: Australian Institute of Health
Circulation, 107, 1692–1711.
and Welfare.
17. Jackson SL, Bereznicki LR, Peterson GM et al. (2004)
3. Peterson GM, Boom K, Jackson SL, Vial JH (2002)
Accuracy, reproducibility and clinical utility of the
Doctors’ beliefs on the use of antithrombotic ther-
CoaguChek S portable international normalized ratio
apy in atrial fibrillation: identifying barriers to
monitor in an outpatient anticoagulation clinic.
stroke prevention. Internal Medicine Journal, 32,
Clinical and Laboratory Haematology, 26, 49–55.
15–23.
18. Jackson SL, Bereznicki LR, Peterson GM et al. (2004)
4. Price CP (2001) Regular review – point of care testing
Accuracy and clinical usefulness of the near-patient
(Review). British Medical Journal, 322, 1285–1288.
testing CoaguChek S INR monitor in rural medical
5. Fitzmaurice DA, Hobbs FDR, Murray ET (1998)
practice. Australian Journal of Rural Health, 12, 137–142.
Primary care anticoagulant clinic management using
19. Bland JM, Altman DG (!995) Comparing methods of
computerized decision support and near patient
measurement – why plotting difference against
international normalized ratio (INR) testing – routine
standard method is misleading. Lancet, 346, 1085–
data from a practice nurse-led clinic. Family Practice,
1087.
15, 144–146.
20. Kapiotis S, Quehenberger P, Speiser W (1995) Eval-
6. Garabedian-Ruffalo SM, Gray DR, Sax MJ, Ruffalo
uation of the new method Coaguchek (R) for the
RL (1985) Retrospective evaluation of a pharmacist-
determination of prothrombin time from capillary
managed warfarin anticoagulation clinic. American
blood – comparison with Thrombotest(R) on Kc-1.
Journal of Hospital Pharmacy, 42, 304–308.
Thrombosis Research, 77, 563–567.
7. Ellis RF, Stephens MA, Sharp GB (1992) Evaluation
21. Chapman DC, Stephens MA, Hamann GL, Bailey LE,
of a pharmacy-managed warfarin-monitoring ser-
Dorko CS (1999) Accuracy, clinical correlation, and
vice to coordinate inpatient and outpatient therapy.
patient acceptance of two handheld prothrombin
American Journal of Hospital Pharmacy, 49, 387–394.
time monitoring devices in the ambulatory setting.
8. Foss MT, Schoch PH, Sintek CD (1999) Efficient
Annals of Pharmacotherapy, 33, 775–780.
operation of a high-volume anticoagulation clinic.
22. Gosselin R, Owings JT, White RH et al. (2000) A
American Journal of Health-System Pharmacy, 56, 443–
comparison of point-of-care instruments designed
449.
for monitoring oral anticoagulation with standard
9. Willey ML, Chagan L, Sisca TS et al. (2003) A phar-
laboratory methods. Thrombosis and Haemostasis, 83,
macist-managed anticoagulation clinic: six-year
698–703.
assessment of patient outcomes. American Journal of
23. Hobbs FD, Fitzmaurice DA, Murray ET, Holder R,
Health-System Pharmacy, 60, 1033–1037.
Rose PE, Roper JL (1999) Is the international nor-
10. Lewis SM, Kroner BA (1997) Patient survey of a
malised ratio (INR) reliable? A trial of comparative
pharmacist-managed anticoagulation clinic. Managed
measurements in hospital laboratory and primary
Care Interface, 10, 66–70.
care settings. Journal of Clinical Pathology, 52, 494–497.
11. Lee YP, Schommer JC (1996) Effect of a pharmacist-
24. Kaatz SS, White RH, Hill J, Mascha E, Humphries JE,
managed anticoagulation clinic on warfarin-related
Becker DM (1995) Accuracy of laboratory and port-
hospital readmissions. American Journal of Health-
able monitor international normalized ratio deter-
System Pharmacy, 53, 1580–1583.
minations – comparison with a criterion standard.
12. Norton JL, Gibson DL (1996) Establishing an out-
Archives of Internal Medicine, 155, 1861–1867.
patient anticoagulation clinic in a community
2005 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 30, 345–353
Rural pharmacy monitoring of warfarin 353
25. Douketis JD, Lane A, Milne J, Ginsberg JS (1998) cists and general practitioners: a qualitative
Accuracy of a portable international normalization assessment. British Journal of General Practice, 53,
ratio monitor in outpatients receiving long-term oral 600–606.
anticoagulant therapy: comparison with a laboratory 30. Gage BF, Boechler M, Doggette AL et al. (2000)
reference standard using clinically relevant criteria Adverse outcomes and predictors of underuse of
for agreement. Thrombosis Research, 92, 11–17. antithrombotic therapy in medicare beneficiaries
26. Macgregor SH, Hamley JG, Dunbar JA, Dodd TR, with chronic atrial fibrillation. Stroke, 31, 822–827.
Cromarty JA (1996) Evaluation of a primary care 31. Flaker GC, McGowan DJ, Boechler M, Fortune G,
anticoagulant clinic managed by a pharmacist. Gage B (1999) Underutilization of antithrombotic
British Medical Journal, 312, 560. therapy in elderly rural patients with atrial fibrilla-
27. Moffat T (2001) Point-of-care testing in the commu- tion. American Heart Journal, 137, 307–312.
nity pharmacy. The Pharmaceutical Journal, 267, 267– 32. National Pharmaceutical Association (1999) Phar-
268. macy anticoagulation clinics could help save 5,000
28. Gutierres SL, Welty TE (2004) Point-of-care testing: lives. The Pharmaceutical Journal, 263, 408.
an introduction. Annals of Pharmacotherapy, 38, 119– 33. Jackson SL, Peterson GM, Vial JH, Daud R, Ang SY
125. (2001) Outcomes in the management of atrial fibril-
29. Hughes CM, McCann S (2003) Perceived inter- lation: clinical trial results can apply in practice.
professional barriers between community pharma- Internal Medicine Journal, 31, 329–336.
2005 Blackwell Publishing Ltd, Journal of Clinical Pharmacy and Therapeutics, 30, 345–353