&get_box_var;

ORIGINAL ARTICLE

Corticosteroid Therapy for Critically Ill Patients with Middle East

Respiratory Syndrome
Yaseen M. Arabi1,2, Yasser Mandourah3, Fahad Al-Hameed4, Anees A. Sindi5, Ghaleb A. Almekhlafi3,
Mohamed A. Hussein6, Jesna Jose6, Ruxandra Pinto7, Awad Al-Omari8,9, Ayman Kharaba10,11, Abdullah Almotairi12,
Kasim Al Khatib13, Basem Alraddadi8,14, Sarah Shalhoub15, Ahmed Abdulmomen16, Ismael Qushmaq14,
Ahmed Mady17,18, Othman Solaiman19, Abdulsalam M. Al-Aithan20, Rajaa Al-Raddadi21, Ahmed Ragab22,
Hanan H. Balkhy1,23, Abdulrahman Al Harthy17, Ahmad M. Deeb24, Hanan Al Mutairi24, Abdulaziz Al-Dawood1,2,
Laura Merson25, Frederick G. Hayden25,26, and Robert A. Fowler27,28,29; for the Saudi Critical Care Trial Group
1
College of Medicine, 6Department of Biostatistics and Bioinformatics, and 24Research Office, King Saud Bin Abdulaziz University for
Health Sciences, King Abdullah International Medical Research Center, Riyadh, Saudi Arabia; 2Intensive Care Department
and 23Department of Infection Prevention and Control, King Abdulaziz Medical City, National Guard Health Affairs, Riyadh, Saudi
Arabia; 3Department of Intensive Care Services, Prince Sultan Military Medical City, Riyadh, Saudi Arabia; 4Department of Intensive Care,
King Saud bin Abdulaziz University for Health Sciences, King Abdullah International Medical Research Center, King Abdulaziz Medical
City, Jeddah, Saudi Arabia; 5Department of Anesthesia and Critical Care, Faculty of Medicine, King Abdulaziz University, Jeddah, Saudi
Arabia; 7Sunnybrook Hospital, University of Toronto, Toronto, Ontario, Canada; 8College of Medicine, Alfaisal University, Riyadh, Saudi
Arabia; 9Department of Intensive Care, Dr. Sulaiman Al-Habib Group Hospitals, Riyadh, Saudi Arabia; 10Department of Critical Care, King
Fahad Hospital, Al-Madinah Al-Monawarah, Saudi Arabia; 11Department of Critical Care, Ohoud Hospital, Al-Madinah Al-Monawarah,
Saudi Arabia; 12Department of Critical Care Medicine, King Fahad Medical City, Riyadh, Saudi Arabia; 13Intensive Care Department,
Al-Noor Specialist Hospital, Makkah, Saudi Arabia; 14Department of Medicine, King Faisal Specialist Hospital and Research Center,
Jeddah, Saudi Arabia; 15Division of Infectious Diseases, Department of Medicine, King Fahad Armed Forces Hospital, Jeddah,
Saudi Arabia; 16Department of Critical Care Medicine, King Saud University, Riyadh, Saudi Arabia; 17Department of Anesthesiology
and Intensive Care, Tanta University Hospitals, Tanta, Egypt; 18Intensive Care Department, King Saud Medical City, Riyadh, Saudi
Arabia; 19King Faisal Specialist Hospital and Research Center, Riyadh, Saudi Arabia; 20Intensive Care Department, King Abdulaziz
Hospital, Al Ahsa, Saudi Arabia; 21Department of Research, Ministry of Health, Jeddah, Saudi Arabia; 22Intensive Care Department,
King Fahd Hospital, Jeddah, Saudi Arabia; 25International Severe Acute Respiratory and Emerging Infection Consortium, Infectious
Diseases Data Observatory, Oxford University, Oxford, United Kingdom; 26Division of Infectious Diseases and International Health,
Department of Medicine, University of Virginia School of Medicine, Charlottesville, Virginia; 27Institute of Health Policy Management
and Evaluation, University of Toronto, Toronto, Ontario, Canada; and 28Department of Critical Care Medicine and 29Department of
Medicine, Sunnybrook Hospital, Toronto, Ontario, Canada
ORCID ID: 0000-0001-5735-6241 (Y.M.A.).

Abstract received corticosteroids were more likely to receive invasive ventilation

(141 of 151 [93.4%] vs. 121 of 158 [76.6%]; P , 0.0001) and had higher
Rationale: Corticosteroid therapy is commonly used among 90-day crude mortality (112 of 151 [74.2%] vs. 91 of 158 [57.6%]; P =
critically ill patients with Middle East Respiratory Syndrome 0.002). Using marginal structural modeling, corticosteroid therapy was
(MERS), but its impact on outcomes is uncertain. Analyses of not significantly associated with 90-day mortality (adjusted odds ratio,
observational studies often do not account for patients’ clinical 0.75; 95% confidence interval, 0.52–1.07; P = 0.12) but was associated
condition at the time of corticosteroid therapy initiation. with delay in MERS coronavirus RNA clearance (adjusted hazard ratio,
0.35; 95% CI, 0.17–0.72; P = 0.005).
Objectives: To investigate the association of corticosteroid therapy
on mortality and on MERS coronavirus RNA clearance in critically ill Conclusions: Corticosteroid therapy in patients with MERS was not
patients with MERS. associated with a difference in mortality after adjustment for time-
varying confounders but was associated with delayed MERS
Methods: ICU patients with MERs were included from 14 Saudi coronavirus RNA clearance. These findings highlight the challenges
Arabian centers between September 2012 and October 2015. We and importance of adjusting for baseline and time-varying
performed marginal structural modeling to account for baseline and confounders when estimating clinical effects of treatments using
time-varying confounders. observational studies.
Measurements and Main Results: Of 309 patients, 151 received
corticosteroids. Corticosteroids were initiated at a median of 3.0 days Keywords: respiratory distress syndrome; coronavirus;
(quartile 1 [Q1]–Q3, 1.0–7.0) from ICU admission. Patients who pneumonia; Saudi Arabia; corticosteroid

( Received in original form June 15, 2017; accepted in final form November 20, 2017 )
Am J Respir Crit Care Med Vol 197, Iss 6, pp 757–767, Mar 15, 2018
Copyright © 2018 by the American Thoracic Society
Originally Published in Press as DOI: 10.1164/rccm.201706-1172OC on November 21, 2017
Internet address: www.atsjournals.org

Arabi, Mandourah, Al-Hameed, et al.: Corticosteroids for Critically Ill Patients with MERS 757
 ORIGINAL ARTICLE

another related coronavirus infection, the initiation and thus do not account for
At Glance Commentary severe acute respiratory syndrome (SARS) potential emerging between-patient time-
have been mainly observational and yielded varying differences that can confound the
Scientific Knowledge on the inconsistent results (4, 5). In a systematic relationship being investigated. Data from
Subject: Corticosteroid therapy is review of studies on SARS, of 29 studies RCTs on corticosteroid therapy in MERS of
commonly used among critically ill documenting corticosteroid use, 25 were sufficient size (to have a better chance of
patients with Middle East respiratory inconclusive and 4 were classified as balancing such known and unknown
syndrome (MERS), but its impact causing possible harm (6). In one confounders between groups receiving or not
on outcome is uncertain. Analyses randomized controlled trial (RCT) that receiving corticosteroid therapy) are lacking.
of observational studies often do included 16 non-ICU patients, “early” The objective of this study is to
not account for time-varying (,7 d of illness) hydrocortisone therapy was investigate the association of corticosteroid
confounders, such as worsening associated with a higher subsequent plasma therapy on mortality and on MERS-CoV
clinical status and the decision to viral load (7). Similar controversy exists with RNA clearance accounting for potential
prescribe corticosteroids, leading to severe influenza. A meta-analysis of multiple immortal time bias and indication bias.
potentially biased estimates of observational studies has shown that Some of this work has been previously
treatment effect. corticosteroid therapy was associated with presented in abstract form (14).
increased mortality in patients with severe
What This Study Adds to the influenza (8). Other serious adverse events,
Field: This multicenter study including opportunistic infections, prolonged Methods
investigated the association of virus replication, and antiviral resistance
corticosteroid therapy on mortality emergence, have also been observed (9). Setting
and MERS coronavirus RNA clearance Challenging analytic issues with these We analyzed data from a multicenter,
accounting for time-varying studies include the risk of immortal time retrospective cohort study from 14 participating
confounders during critical illness, up bias and indication bias from time-varying Saudi Arabian tertiary care hospitals (3). The
to the time of corticosteroid therapy confounding (10–13). Immortal time bias institutional review boards of all participating
initiation, using marginal structural refers to the requirement for patients to centers approved the study. Patient-level
models. Corticosteroid therapy in survive long enough to receive the intervention informed consent was not required.
patients with MERS was not associated of interest, leading to a potential incorrect
with differences in mortality but was estimation of a positive treatment effect (11, Patients
associated with delay in MERS 13). Indication bias from time-varying We included in this analysis all patients with
coronavirus RNA clearance. confounding refers to having an association MERS admitted to the participating ICUs
related to the indication of the intervention between September 2012 and October 2015.
that is evolving during the course of the illness. We excluded patients known to be receiving
Middle East respiratory syndrome coronavirus For example, a patient who becomes sicker has chronic corticosteroid therapy before the onset
(MERS-CoV) is a pathogenic respiratory virus an inherent higher risk of death but may also of critical illness. Details of management and
that often causes severe acute respiratory be more likely to receive second- or third-line laboratory testing of patients with MERS in
illness with substantial mortality. To date, therapies because first-line therapies have not the cohort have been reported previously (3).
there is no specific treatment for MERS, and led to improvement, possibly leading to an For MERS-CoV–positive patients, follow-up
management is largely supportive (1, 2). incorrect estimation of a negative or harmful respiratory samples were collected at
Systemic corticosteroid therapy is treatment effect. Studies that adjust only for the discretion of the treating teams
commonly used among critically ill patients characteristics present at time of hospital or approximately one to two times per
with MERS, but its impact on the clinical ICU admission do not account for the clinical week to assess clearance of viral RNA for
outcomes is uncertain (3). Data from condition at the time of corticosteroid therapy infection control purposes.

Author Contributions: Y.M.A. and A.M.D.: Conception and design, data acquisition, analytical plan, interpretation of data for the work, drafting of the
manuscript, critical revision of the manuscript for important intellectual content, approval of the final version to be published, and agreement to be accountable
for all aspects of the work. Y.M., F.A.-H., A.A.S., G.A.A., A.A.-O., A.K., A. Almotairi, K.A.K., B.A., S.S., A. Abdulmomen, I.Q., A.M., O.S., A.M.A.-A., R.A.-R.,
A.R., H.H.B., A.A.H., H.A.M., and A.A.-D.: Data acquisition, critical revision of the manuscript for important intellectual content, approval of the final version to
be published, and agreement to be accountable for all aspects of the work. M.A.H. and J.J.: Conception and design, analytical plan, data analysis, critical
revision of the manuscript for important intellectual content, approval of the final version to be published, and agreement to be accountable for all aspects of
the work. R.P.: Conception and design, analytical plan, critical revision of the manuscript for important intellectual content, approval of the final version to be
published, and agreement to be accountable for all aspects of the work. L.M.: Conception and design, critical revision of the manuscript for important
intellectual content, approval of the final version to be published, and agreement to be accountable for all aspects of the work. F.G.H. and R.A.F.: Conception
and design, analytical plan, interpretation of data for the work, critical revision of the manuscript for important intellectual content, approval of the final version to
be published, and agreement to be accountable for all aspects of the work
Correspondence and requests for reprints should be addressed to Yaseen M. Arabi, M.D., Intensive Care Department, MC 1425, College of Medicine, King Saud
Bin Abdulaziz University for Health Sciences, King Abdullah International Medical Research Center, P.O. Box 22490, Riyadh 11426, Kingdom of Saudi Arabia.
E-mail: [email protected].
This article has an online supplement, which is accessible from this issue’s table of contents at www.atsjournals.org.

758 American Journal of Respiratory and Critical Care Medicine Volume 197 Number 6 | March 15 2018
ORIGINAL ARTICLE

Corticosteroid Therapy Association of Corticosteroid Therapy probability of treatment weighting to

The main exposure was corticosteroid and 90-Day Mortality account for time-varying confounders that
therapy, defined as the use of systemic We tested the associations between are likely to influence the corticosteroid
corticosteroids. We converted all corticosteroid therapy and mortality using therapy initiation and at the same time are
preparations to hydrocortisone-equivalent three approaches: one approach that adjusts likely to be correlated with the risk of
doses (methylprednisolone 1:5, for baseline differences only, but does not mortality (10, 17–19). We included the
dexamethasone 1:25, prednisolone 1:4). account for immortal time bias or indication above-mentioned baseline confounding
bias from time-varying confounding variables as well. In this model, we
(logistic regression); a second approach— considered ventilation status and SOFA
Cox proportional hazards regression scores on the day of corticosteroid therapy
Data Collection accounting for time-varying exposure—that initiation and the day before as the time-
Data were collected using standardized adjusts for baseline differences and varying variables. We reasoned that these
International Severe Acute Respiratory and accounts for immortal time bias (with variables capture the patient condition that
Emerging Infection Consortium case report corticosteroids as a time-varying exposure); physicians would most often consider when
forms (15). We extracted data on patient and a third approach, using marginal initiating corticosteroid therapy.
demographic features, underlying structural models, that adjusts for baseline This process involves calculation of two
comorbidities, radiographic findings, and differences and accounts for indication bias weights for each observation: treatment
the durations from symptom onset to (by examining the impact of time-varying selection weight and censoring weight. The
presentation to the emergency department, confounders on the daily “risk” of treatment selection weight at time k is a ratio
ICU admission, and intubation. We prescription of corticosteroids) and of two weights. The numerator is the
assessed severity of illness using the immortal time bias (using corticosteroids as product of probabilities that a patient
sequential organ failure assessment (SOFA) a time-varying exposure). receives his observed treatment at time k,
score, as well as laboratory and ventilator given the baseline covariates. The
parameters on Days 1, 3, 7, 14, and 28 Logistic Regression denominator is calculated similarly by
of ICU admission (15, 16). We collected To examine the independent association of incorporating also the time-varying
data on the type, maximum daily corticosteroid therapy on 90-day mortality, covariates (SOFA on the day, SOFA on the
dose, and duration of corticosteroids, we performed multivariable logistic regression previous day, mode of ventilation on the day
and we documented the use of antiviral analysis, with corticosteroid therapy being the and mode of ventilation on the previous
therapy. independent variable. We included in the day) as well (10, 17). The weights are
The primary outcome was 90-day all- multivariable model a priori–decided baseline updated until the first day of corticosteroid
cause mortality. In patients who had at least variables of clinical interest and all significant therapy and kept constant afterward.
one follow-up real-time RT-PCR (rRT- variables at the univariable level (P < 0.2), Because SOFA scores were recorded on
PCR) performed after the diagnostic test, we which included: age, sex, Day 1 SOFA, Days 1, 3, 7, 14, and 28, we imputed
examined the time to MERS-CoV RNA asthma/chronic pulmonary disease, chronic missing values for the remaining days (see
clearance in respiratory secretions by cardiac disease, chronic neurological online supplement). Ventilation status was
rRT-PCR, defined as the time from ICU disease, diabetes with chronic complications, recorded also on Days 1, 3, 7, 14, and 28,
admission until the test was negative on two obesity, days from onset of symptom to ICU and we coded the mode of ventilation as
occasions, without a positive test afterward. admission, and healthcare worker status, by follows: 0 = no ventilation, 1 = noninvasive
Secondary outcomes were ICU and hospital applying the PROC GENMOD procedure ventilation, 2 = invasive ventilation, 3 =
mortality and length of stay in the ICU and (SAS version 9.4; SAS Institute). advanced ventilation support (e.g.,
the hospital. extracorporeal membrane oxygenation,
oscillatory, prone, nitric oxide). We imputed
Cox Proportional Hazards Regression the mode of ventilation between these days by
Accounting for Time-Varying the last-observation-carried-forward method.
Statistical Analysis Exposure The same approach is used to calculate
We compared baseline characteristics, To examine outcomes as a time to event the censoring weight for early patient
cointerventions, and outcomes of patients (i.e., death), we performed a Cox dropout. We have censored patients at
who received corticosteroid therapy during proportional hazards model adjusting for hospital discharge or at Day 90, and the
ICU admission and those who did not the same above-mentioned baseline weights for censoring are calculated as the
receive any corticosteroid therapy using chi- covariates with corticosteroid therapy as a ratio of a subject’s probability of remaining
square test or Fisher exact test for categorical time-varying covariate. uncensored up to day k. The final weight
variables and Student t test or Mann- for each observation is obtained by
Whitney U test for continuous variables Marginal Structural Model multiplying the treatment selection weights
as appropriate. For serial measurements, Because corticosteroid therapy often was not and the censoring weights. We used a
we tested differences between the two started at the time of ICU admission, but weight-trimming approach to deal with
groups over time using repeated measures rather during the course of the disease extreme weights; weights larger than the
analysis of variance with no imputation on the basis of a change in the patient 95th percentile value were fixed at the 95th
for missing values or correction for condition, we performed marginal percentile value, and weights smaller than
multiple comparisons. structural model analysis with inverse the fifth percentile value were fixed at the

Arabi, Mandourah, Al-Hameed, et al.: Corticosteroids for Critically Ill Patients with MERS 759
 ORIGINAL ARTICLE

fifth percentile value. This process marginal structural model used for 90-day RNA clearance might have been influenced
continued until the average weight reached mortality above. by the practice of repeating rRT-PCR
approximately 1. The time-varying among different sites, we conducted a
intercept was modeled by a smoothing Subgroup and Sensitivity Analyses sensitivity analysis restricting to centers
function of time, using restricted cubic Because a corticosteroid therapy that had repeated rRT-PCR on more than
splines with five knots for days since effect may be dose dependent and may 50% of their patients.
beginning of follow-up day (17, 19). vary according to the time of initiation Tests were two-sided, with significance
In step 2, weighted pooled logistic (20, 21), we performed all previous set at a , 0.05. Results from all
regression with robust SE was used to models on the following stratified groups: multivariable analyses are reported as
estimate the effect of corticosteroids on patients who received high-dose odds ratios (OR) or hazard ratios (HR) with
mortality after adjusting for baseline corticosteroid therapy (highest daily dose of 95% confidence intervals (CIs) as
characteristics. We modeled the probability .300 mg of hydrocortisone equivalent), appropriate. Analyses were conducted
of receiving corticosteroid therapy with the patients who received low-dose using SAS version 9.4.
assumption that once the patient was started corticosteroid therapy (highest daily dose of
on corticosteroid therapy the patient will <300 mg of hydrocortisone equivalent),
remain on that treatment. and patients who had corticosteroid
therapy initiated in the first 7 days of
Results
Association of Corticosteroid Therapy ICU admission, each compared with
Patient Characteristics
and MERS-CoV RNA Clearance patients who did not receive any
Within the study period, 309 patients with
We tested the associations between corticosteroid therapy. We also compared
MERS met the eligibility criteria for this
corticosteroid therapy and MERS-CoV corticosteroid therapy started after Day 7
study. Almost, half of these patients, 151
RNA clearance in two approaches: one compared with no corticosteroid therapy;
of 309 (48.9%), received corticosteroid
approach—Cox proportional hazards in this analysis, we only included patients
therapy. Patients receiving corticosteroid
regression accounting for time-varying who were alive after Day 7, to ensure that
therapy and those not receiving
exposure—that adjusts for baseline patients in both groups were comparable in
corticosteroid therapy were similar in most
differences and accounts for immortal time having “the opportunity” to receive the
baseline characteristics (Table 1). However,
bias (with corticosteroids as a time-varying treatment, thus minimizing the risk of
patients given corticosteroid therapy were
exposure); and a second approach, using confounding due to immortal time bias.
more likely to have one or more
marginal structural Cox proportional To further assess the potential effect
comorbidity than those without
hazards modeling, that adjusts for baseline modification of the time of initiation of
corticosteroid therapy (132 of 151 [87.4%]
differences and accounts for indication bias corticosteroid therapy on the association
compared with 115 of 158 [72.8%]; P =
(by examining the impact of time-varying between corticosteroid therapy and 90-day
0.001), including diabetes with chronic
confounders on the daily “risk” of mortality, we performed landmark
complications, chronic pulmonary disease,
prescription of corticosteroids) and analyses at different time cutoff points
and chronic cardiac disease (Table 1). The
immortal time bias (using corticosteroids as by logistic regression analysis and Cox
use of corticosteroid therapy varied by
a time-varying exposure). proportional hazards regression model
site (median percentage of patients on
accounting for time-varying exposure.
corticosteroid therapy, 50%; quartile 1
Cox Proportional Hazards Regression These analyses compared patients who
[Q1]–Q3, 34–70%; see Figure E1 in the
Accounting for Time-Varying were alive and with/without the exposure
online supplement).
Exposure (corticosteroid/no corticosteroid) by each
We used Cox proportional hazards time cutoff point. To account for the
regression to examine the time to MERS- possible variation by site, we performed a Corticosteroid Use
CoV RNA clearance rRT-PCR. For this sensitivity analysis using a logistic Hydrocortisone was the most frequently
analysis, we censored patients if they never regression model adjusting for clustering by administered corticosteroid (103 of
cleared MERS-CoV RNA or at hospital centers in addition to the previously 151 [68.2%] patients), followed by
discharge if they were discharged alive mentioned baseline variables. To examine methylprednisolone (61 of 151 [40.4%])
before they had cleared MERS-CoV RNA. whether imputation of missing data had an (Table 2). Corticosteroid therapy was
We adjusted for the same baseline impact on the association observed in the started at a median (Q1–Q3) of 3.0 days
covariables used in the logistic regression marginal structural model, and considering (1.0–7.0 d) from ICU admission. The
model, with corticosteroid therapy as a that corticosteroid therapy was initiated in median (Q1–Q3) of the maximum daily
time-varying covariate. 80% of patients in the first 7 days when hydrocortisone-equivalent dose was
imputation was relatively limited, we 300.0 mg (200.0–400.0 mg), with a median
Marginal Structural Cox Proportional conducted a sensitivity analysis comparing (Q1–Q3) duration of 7.0 days (4.0–14.0 d).
Hazards Model patients who were started on corticosteroid Figure 1 shows the distribution of the
The marginal structural Cox proportional therapy in the first 7 days with patients in time to initiate corticosteroid therapy from
hazards model was performed incorporating the no corticosteroid therapy group with ICU admission. Before corticosteroid
the stabilized weights to estimate the effect imputation up to Day 7 only. To examine therapy administration, the median
of corticosteroid therapy on MERS-CoV whether the results of the association of (Q1–Q3) PaO2/FIO2 was 108.8 mm Hg
RNA clearance in a similar approach to the corticosteroid therapy and MERS-CoV (68.8–166.2 mm Hg), and the median

760 American Journal of Respiratory and Critical Care Medicine Volume 197 Number 6 | March 15 2018
ORIGINAL ARTICLE

Table 1. Baseline Characteristics and Physiological Parameters on Day 1 of Admission to the ICU among Patients with Middle East
Respiratory Syndrome in the Corticosteroid Therapy and No Corticosteroid Therapy Groups

Variable Corticosteroid Group (N = 151) No Corticosteroid Group (N = 158) P Value

Age, yr, mean 6 SD 57.8 6 17.2 55.3 6 17.3 0.20*

BMI, kg/m2 28.9 (24.3–33.8) 28.4 (24.2–33.0) 0.64
Male sex 107 (70.9) 106 (67.1) 0.47
Source of infection 0.25
Community acquired 71 (47.0) 60 (38.0)
Healthcare worker, hospital acquired 15 (9.9) 17 (10.8)
Non–healthcare worker, hospital acquired 55 (36.4) 62 (39.2)
Days from onset of symptoms to the emergency 5.0 (3.0–8.0) 4.0 (3.0–7.0) 0.03
room
Days from onset of symptoms to ICU admission 8.0 (5.0–12.0) 6.5 (4.0–11.0) 0.07
Days from onset of symptoms to intubation 8.0 (6.0–12.0) 8.0 (5.0–12.0) 0.40
Comorbidities
Any comorbidity 132 (87.4) 115 (72.8) 0.001
Diabetes with chronic complications 87 (57.6) 69 (43.7) 0.01
Asthma/chronic pulmonary disease 30 (19.9) 15 (9.5) 0.01
Liver disease 12 (7.9) 9 (5.7) 0.43
Renal disease 43 (28.5) 47 (29.7) 0.81
Chronic cardiac disease 75 (49.7) 53 (33.5) 0.004
Chronic neurological disease/hemiplegia or 20 (13.2) 13 (8.2) 0.15
paraplegia or dementia
Obesity 21 (13.9) 15 (9.5) 0.23
Rheumatological disease 2 (1.3) 1 (0.6) 0.62†
HIV/AIDS 1 (0.7) 1 (0.6) .0.99†
Any malignancy, including leukemia or 14 (9.3) 13 (8.2) 0.75
lymphoma/solid tumors
Physiologic parameters
SOFA score 9.0 (6.0–12.0) 8.0 (5.0–11.0) 0.09
Tidal volume, ml 400 (350–434) 400 (348–454) .0.99
PEEP, cm H2O 10.0 (8.0–14.0) 12.0 (10.0–15.0) 0.46
Plateau pressure, cm H2O 28.0 (22.0–30.0) 28.0 (20.0–31.0) 0.57
PaO2/FIO2 ratio, mm Hg 99.0 (64.0–151) 115.5 (73.8–176) 0.12
Mean arterial pressure, mm Hg 68.5 (59.0–80.0) 70.0 (61.0–84.7) 0.05
Lactate, mmol/L 1.8 (1.1–2.7) 1.6 (1.1–2.7) 0.50
INR 1.1 (1.0–1.3) 1.1 (1.0–1.3) 0.88
Creatinine, mmol/L 127.0 (74.0–251.0) 118.0 (72.0–255.0) 0.84
Bilirubin level, mmol/L 12.5 (7.0–24.0) 11.3 (7.8–20.9) 0.70
Platelets, 3109/L 180 (111.5–254) 160 (113–241) 0.42
No. of quadrants with infiltrates on chest 3.0 (2.0–4.0) 3.0 (2.0–4.0) 0.19
radiograph

Definition of abbreviations: BMI = body mass index; INR = international normalized ratio; PEEP = positive end-expiratory pressure; Q = quartile; SOFA =
Sequential Organ Failure Assessment.
Data presented as n (%) or median (Q1–Q3) unless otherwise noted. For continuous variables, Mann-Whitney U test was used to calculate the P value
unless otherwise noted. For categorical variables, chi-square test was used to calculate the P value unless otherwise noted.
*t test was used to calculate P value.
†
Fisher exact test was used to calculate P value.

(Q1–Q3) positive end-expiratory pressure (Table 3). Changes of physiological compared with 15.0 d [8.0–30.0 d];
was 12.0 cm H2O (10.0–14.0 cm H2O) parameters over time and between the P = 0.0006) (Table 3).
(Table 2). corticosteroid therapy group and no
corticosteroid therapy group are described Mortality
Clinical Course and Outcomes in Figure E2. After adjustment for baseline variables
Throughout their ICU stay, patients Patients who received corticosteroid using multivariable logistic regression,
in the corticosteroid therapy group therapy compared with those who did not corticosteroid therapy was associated with
received more invasive ventilation, high- had higher crude 90-day mortality (112 of higher 90-day mortality (adjusted OR
frequency oscillation ventilation, nitric 151 [74.2%] compared with 91 of 158 [aOR], 1.87; 95% CI, 1.02–3.44; P = 0.04).
oxide, neuromuscular blockers, [57.6%]; P = 0.002), longer ICU length of Adjustment for clustering by site did not
vasopressors, blood transfusion, and stay (median [Q1–Q3], 12.5 d [8.0–23.0 d] significantly alter this association (Table
renal replacement therapy and received compared with 7.0 d [5.0–13.0 d]; E1). Using Cox proportional hazards
more ribavirin and IFN than patients who P , 0.0001), and longer hospital length of regression model accounting for time-
did not receive corticosteroid therapy stay (median [Q1–Q3], 21.0 d [13.0–38.0 d] varying exposures, with adjustment for the

Arabi, Mandourah, Al-Hameed, et al.: Corticosteroids for Critically Ill Patients with MERS 761
 ORIGINAL ARTICLE

Table 2. Corticosteroid Therapy among Critically Ill Patients with Middle East and HRs when using increasing time cutoff
Respiratory Syndrome (n = 151) points of initiation of corticosteroid therapy
(Figure 2). Using the marginal structural
Medication Variable Result
model, there was no significant association
between corticosteroid therapy and
90-day mortality (aOR, 0.75; 95% CI,
Dexamethasone 9 (6.0)*
Hydrocortisone 103 (68.2)* 0.52–1.07; P = 0.12). Sensitivity analysis
Methylprednisolone 61 (40.4)* comparing patients who were started on
Prednisolone 20 (13.2)* corticosteroid therapy in the first 7 days to
Duration of corticosteroids, d patients in the no corticosteroid therapy
All patients 7.0 (4.0–14.0)
Survivors 10.0 (4.0–19.0)
group and imputing missing data only up
Nonsurvivors 7.0 (4.0–12.0) to Day 7 did not alter the association
Dose, hydrocortisone equivalent/d, mg 300.0 (200.0–400.0) (n = 260; aOR, 0.76; 95% CI, 0.50–1.15;
Duration between onset of illness and corticosteroid 10.0 (7.0–17.0) P = 0.20). The dose and time of initiation of
initiation, d corticosteroid therapy were not associated
Duration between hospital admission and 7.0 (3.0–15.0)
corticosteroid initiation, d with differences in 90-day mortality
Duration between ICU admission and corticosteroid 3.0 (1.0–7.0) (Table 4).
initiation, d
Duration between onset of ventilation and 3.0 (1.0–7.0)
corticosteroid initiation, d
MERS-CoV RNA Clearance
PaO2/FIO2 before corticosteroid initiation 108.8 (68.8–166.2) Crude analyses showed no statistically
Positive end expiratory pressure before corticosteroid 12.0 (10.0–14.0) significant differences in the proportion
initiation, cm H2O of patients who had MERS-CoV RNA
SOFA cardiovascular score, before corticosteroid 1.0 (0.0–3.0) clearance between the corticosteroid therapy
initiation
group and the no corticosteroid therapy
Definition of abbreviations: Q = quartile; SOFA = Sequential Organ Failure Assessment. group, among all patients and among
Data presented as n (%) or median (Q1–Q3). survivors (Table 3). Cox proportional
*Percentages add to more than 100% because some patients received more than one formulation of hazards regression accounting for time-
corticosteroids during ICU stay.
varying exposure showed no significant
association of corticosteroid therapy with
baseline variables, corticosteroid therapy demonstrated that the associations the time to MERS-CoV RNA clearance
was not associated with mortality difference observed with logistic regression and Cox (aHR, 1.06; 95% CI, 0.61–1.84; P = 0.84)
(adjusted HR [aHR], 1.20; 95% CI, proportional hazards regression analyses (Table 4). However, using marginal
0.88–1.63; P = 0.24). Landmark analyses were time dependent, with increasing ORs structural Cox proportional hazards
modeling, corticosteroid therapy was
35 associated with a significant delay in
MERS-CoV RNA clearance (aHR, 0.35;
95% CI, 0.17–0.72; P = 0.005) (Table 4).
30 29
Sensitivity analysis restricting to centers
27 who had repeated rRT-PCR on more than
25 50% of their patients (11 of 14 centers, 168
patients) demonstrated a similar
Number of patients

association (aHR, 0.33; 95% CI, 0.15–0.72;

20
20 P = 0.005). A similar association with
delayed MERS-CoV RNA clearance was
15 observed when restricting analysis to low-
15 14 dose, high-dose, and early initiation of
corticosteroid therapy. The association of
late initiation of corticosteroid therapy
10
(after 7 d) and MERS-CoV RNA clearance
6 6 was not statistically significant (Table 4).
5
5 4
3 3 3
2 2 2
1 1 1 1 1 1 1 Discussion
0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 26 32 Our study investigates the association of
Day from ICU addmission corticosteroid therapy on mortality and
Figure 1. Time to corticosteroid therapy initiation from ICU admission. Day 0 includes patients who MERS-CoV RNA clearance accounting for
were already on corticosteroid therapy when admitted to the ICU. Date of initiation of corticosteroid time-varying confounders during critical
therapy was missing in three patients. illness. Our study shows that corticosteroid

762 American Journal of Respiratory and Critical Care Medicine Volume 197 Number 6 | March 15 2018
ORIGINAL ARTICLE

Table 3. ICU Course and Outcomes among Patients with Middle East Respiratory Syndrome in the Corticosteroid Therapy and No
Corticosteroid Therapy Groups

Variable Corticosteroids (n = 151) No Corticosteroids (n = 158) P Value

Noninvasive positive pressure ventilation 54 (35.8) 41 (25.9) 0.06

Invasive ventilation 141 (93.4) 121 (76.6) ,0.0001
Neuromuscular blockade 74 (49.0) 46 (29.1) 0.0003
High-frequency oscillation ventilation 17 (11.3) 7 (4.4) 0.03
ECMO 10 (6.6) 8 (5.1) 0.56
Nitric oxide 29 (19.2) 10 (6.3) 0.0007
Prone positioning 20 (13.2) 10 (6.3) 0.04
Vasopressors 134 (88.7) 111 (70.3) ,0.0001
Blood transfusion 74 (49.0) 34 (21.5) ,0.0001
Ribavirin and/or IFN 76 (50.3) 59 (37.3) 0.02
Ribavirin and IFN 64 (42.4) 47 (29.7)
IFN only 2 (1.3) 6 (3.8) 0.03*
Ribavirin only 10 (6.6) 6 (3.8)
Oseltamivir 89 (58.9) 80 (50.6) 0.14
Renal replacement therapy 83 (55.0) 69 (43.7) 0.05
ICU mortality 114 (75.5) 89 (56.3) 0.0004
Hospital mortality 117 (77.5) 92 (58.2) 0.0003
90-d mortality 112 (74.2) 91 (57.6) 0.002
ICU length of stay, d 12.5 (8.0–23.0) 7.0 (5.0–13.0) ,0.0001
Hospital length of stay, d 21.0 (13.0–38.0) 15.0 (8.0–30.0) 0.0006
MERS-CoV RNA clearance 34 of 99 (34.3) 31 of 104 (29.8) 0.49
Time to MERS-CoV RNA clearance,† d 21.0 (10.0–31.0) 17.0 (7.0–34.0) 0.31
MERS-CoV RNA clearance among 90-d survivors 20 of 28 (71.4) 24 of 49 (49.0) 0.06
Time to MERS-CoV RNA clearance among 90-d 23.0 (11.0–32.5) 17.5 (9.5–36.5) 0.62
survivors,‡ d

Definition of abbreviations: ECMO = extracorporeal membrane oxygenation; MERS-CoV = Middle East respiratory syndrome coronavirus; Q = quartile.
Data presented as n (%) or median (Q1–Q3). For continuous variables, Mann-Whitney U test was used to calculate the P value. For categorical variables,
chi-square test was used to calculate the P value unless otherwise noted.
*Fisher exact test was used to calculate P value.
†
Time to MERS-CoV RNA clearance was calculated for 34 patients in the corticosteroid group and 31 patients in the no corticosteroid group.
‡
Time to MERS-CoV RNA clearance among 90-d survivors was calculated for 20 patients in the corticosteroid group and 24 patients in the no
corticosteroid group.

therapy in patients with MERS was not therapy was longer in our study than what with adverse effects and increased
associated with significant change in 90-day has been reported in other respiratory mortality (5, 7).
mortality after adjustment for time-varying infection case series (10), which may reflect RCTs are the best design to account for
confounders but was associated with current recommendations to avoid confounding factors. Multivariable analysis
delayed MERS-CoV RNA clearance. corticosteroid therapy for patients with often uses baseline and not time-varying
Our study shows that the corticosteroid MERS unless indicated for other reasons confounder adjustment, leading to residual
therapy was commonly used for critically ill (25), and suggests that physicians used confounding from immortal time bias and
patients with MERS. Corticosteroid therapy corticosteroids when the clinical condition evolving indication bias from time-varying
was initiated at variable times during the was not improving. confounders. The landmark analyses in our
course of the disease. Patients were generally Studies examining corticosteroid study demonstrate the time-dependent
hypoxemic and on moderate levels of therapy and vital outcomes for patients with nature of associations. Time-varying
positive end-expiratory pressure at the community-acquired pneumonia and acute confounding adjustment, including
time of starting therapy, indicating that respiratory distress syndrome have yielded marginal structural models, has been used
corticosteroids were used for patients conflicting results (26–29). The results of fairly extensively in other fields that have
who were quite ill and/or were not showing studies on unselected patients with longer periods of follow-up, such as studies
signs of improvement. Patients with community-acquired pneumonia may not on treatment of HIV infection (17). In
comorbidities were more likely to receive be generalizable to viral pneumonia; a viral critical care, observational studies have
corticosteroid therapy, as shown in other etiology was documented in a small typically not considered time-varying
studies (10, 22–24). The median daily dose number of the included patients in these confounding adjustment in shorter follow-
in our study was in the range reported by studies. Observational studies in patients up periods. However, such adjustment is
others (10, 20), taking into consideration with SARS coronavirus found that high- particularly relevant when the likelihood of
that we used the maximum daily dose in dose systemic corticosteroid therapy was receiving a treatment and the likelihood
our analyses. The median time between the associated not only with increased of the outcome are both influenced by
onset of critical illness and corticosteroid subsequent blood viral loads but also dynamic factors that change between

Arabi, Mandourah, Al-Hameed, et al.: Corticosteroids for Critically Ill Patients with MERS 763
 ORIGINAL ARTICLE

A
Alive and Corticosteroids No Corticosteroids
exposure by Events/N (%) Events/N (%) OR (95% CI) P Value

Day 3 60/82 (73.2) 75/139 (54.0) 1.28 ( 0.60, 2.72) 0.52

Day 4 68/94 (72.3) 73/137 (53.3) 1.61 ( 0.79, 3.28) 0.19

Day 6 66/93 (71.0) 51/115 (44.3) 1.83 ( 0.89, 3.77) >0.99

Day 8 64/91 (70.3) 29/93 (31.2) 3.67 ( 1.63, 8.23) 0.002

Day 10 54/82 (65.9) 23/87 (26.4) 4.48 ( 1.90, 10.56) 0.001

Day 12 46/77 (59.7) 17/81 (21.0) 7.29 ( 2.62, 20.35) <0.0001

Day 14 42/72 (58.3) 12/76 (15.8) 10.1 ( 3.35, 30.27) <0.0001

1 5 10 20 30
OR (95% CI)

B
Alive and Corticosteroids No Corticosteroids
exposure by Events/N (%) Events/N (%) HR (95% CI) P Value

Day 3 60/82 (73.2) 75/139 (54.0) 0.88 (0.61, 1.28) 0.52

Day 4 68/94 (72.3) 73/137 (53.3) 0.99 (0.69, 1.41) 0.94

Day 6 66/93 (71.0) 51/115 (44.3) 1.22 (0.81, 1.85) 0.34

Day 8 64/91 (70.3) 29/93 (31.2) 2.17 (1.31, 3.60) 0.003

Day 10 54/82 (65.9) 23/87 (26.4) 2.50 (1.42, 4.40) 0.002

Day 12 46/77 (59.7) 17/81 (21.0) 3.15 (1.67, 5.93) <0.0001

Day 14 42/72 (58.3) 12/76 (15.8) 4.71 (2.26, 9.81) <0.0001

1 2 3 4 5 6 8 10
HR (95% CI)
Figure 2. Landmark analyses for 90-day mortality at different time cutoff points by logistic regression analysis (A) and Cox proportional hazards regression
model accounting for time-varying exposure (B) comparing patients who were alive and with/without the exposure (corticosteroid/no corticosteroid) by
each time cutoff point. CI = confidence interval; HR = hazard ratio; OR = odds ratio.

baseline and the time at which treatment is (H1N1) in critically ill patients that showed therapy and mortality are critically
received (in those who receive) and every no association between corticosteroid dependent on the analytic model chosen
time point at which the treatment might therapy and mortality when accounting and highlights the need for RCTs as
have been administered and was not (in for time-varying confounders using a the most accurate way to obtain an
those who do not receive). Our results are marginal structural model (10). Our estimate of the treatment effect. However,
in accordance with the results derived from study demonstrates that measures of as cases of MERS are relatively infrequent
an observational study on influenza A the association of corticosteroid and sporadic, analyses of observational

764 American Journal of Respiratory and Critical Care Medicine Volume 197 Number 6 | March 15 2018
 ORIGINAL ARTICLE

Table 4. Ninety-Day Mortality and Middle East Respiratory Syndrome Coronavirus RNA Clearance of Critically Ill Patients with Middle East Respiratory Syndrome
Using Various Adjustment Methodologies

Day-90 Mortality MERS-CoV RNA Clearance

Cox Proportional Hazards Cox Proportional Hazards Marginal Structural Cox
Logistic Regression Regression Model MSM Regression Model Proportional Hazards Model
P P P P P
Variables n aOR (95% CI) Value n aHR (95% CI) Value n aOR (95% CI) Value n aHR (95% CI) Value n aHR (95% CI) Value

All patients treated with corticosteroids vs. 291 1.87 (1.02–3.44) 0.04 291 1.20 (0.88–1.63) 0.24 290 0.75 (0.52–1.07) 0.12 194 1.06 (0.61–1.84) 0.84 189 0.35 (0.17–0.72) 0.005
patients not treated with corticosteroids
(reference)
Patients treated with .300 mg vs. patients 185 1.43 (0.53–3.82) 0.48 185 1.05 (0.65–1.69) 0.84 184 0.99 (0.55–1.80) 0.98 121 0.87 (0.34–2.22) 0.78 118 0.26 (0.09–0.77) 0.02
not treated with corticosteroids
(reference)
Patients treated with <300 mg vs. patients 241 1.89 (0.94–3.82) 0.08 241 1.21 (0.86–1.71) 0.28 240 0.75 (0.51–1.11) 0.15 164 1.01 (0.54–1.89) 0.98 159 0.41 (0.19–0.88) 0.02
not treated with corticosteroids
(reference)
Patients treated for <7 d vs. not treated with 261 1.81 (0.91–3.59) 0.09 261 1.05 (0.76–1.46) 0.75 260 0.88 (0.61–1.26) 0.48 174 0.59 (0.31–1.13) 0.11 169 0.23 (0.09–0.63) 0.004
corticosteroids (reference)
Patients treated for .7 d vs. patients not 141 2.49 (0.85–7.25) 0.10 142 1.60 (0.88–2.89) 0.12 142 0.51 (0.26–1.00) 0.05 87 3.12 (1.31–7.40) 0.01 85 0.94 (0.36–2.47) 0.90
treated with corticosteroids who survived
.7 d (reference)

Definition of abbreviations: aHR = adjusted hazard ratio; aOR = adjusted odds ratio; CI = confidence interval; MERS-CoV = Middle East respiratory syndrome coronavirus; MSM = marginal
structural model.

Arabi, Mandourah, Al-Hameed, et al.: Corticosteroids for Critically Ill Patients with MERS
For logistic regression, Hosmer-Lemeshow goodness-of-fit test was used to assess the model fitness, and P values were not significant for all analyses except for “patients treated for .7 d versus
not treated with corticosteroids and who survived .7 d” data. After removing one observation, which has the highest influence on the chi-square goodness of fit (on the basis of residuals and
deviance influence statistics), a nonsignificant P value was obtained; we report the results after removing that observation. HR , 1 signifies delay in Middle East respiratory syndrome coronavirus
RNA clearance.

765
 ORIGINAL ARTICLE

data may ultimately provide the best possible that patients with persistent therapy in the first 7 days, when missing
available evidence. viral shedding trajectories were more data were relatively limited, thus
Our results reveal that corticosteroid likely to get repeat RNA testing. This minimizing the impact of imputation.
therapy delayed the MERS-CoV RNA could be a source of bias, because Furthermore, sensitivity analysis
clearance. Similarly, corticosteroid therapy only patients with repeat testing were restricting to the first week showed no
delayed the viral clearance in avian influenza included in the analysis. However, we do change in the estimate of treatment effect.
A(H7N9) (20) and SARS (7). This may be not believe the decision to repeat testing Because of the retrospective observational
related to immune-suppressing effects of was related to the decision to treat nature of our study, repeat rRT-PCR
corticosteroid therapy, which are mediated with corticosteroids. In addition, we testing was not protocolized and varied
mainly by T-cell responses (7). However, performed a sensitivity analysis restricting among centers. However, sensitivity
it is important to note that persistent to centers that had repeated testing on analysis restricting to centers with more
positivity of MERS-CoV RNA does not more than 50% of patients and found that frequent testing showed that associations
necessarily indicate persistent shedding of the association of corticosteroid therapy of corticosteroid therapy and viral
live virus. with MERS-CoV RNA remained the clearance were robust. Importantly,
Our study is the first to address same. We did not assess other associated our findings cannot necessarily be
corticosteroid therapy in MERS and is outcomes with corticosteroid therapy, generalized to patients with other types
derived from the largest collaborative such as opportunistic infections, of viral pneumonia or, more broadly,
multicenter observational database on hyperglycemia, and neuromyopathy. patients with acute respiratory distress
critically ill patients with MERS, using Marginal structural models require the syndrome.
standardized and detailed data collection. availability of time-varying data on
The retrospective observational design is a each time point around when the Conclusions
main limitation of our study. We used intervention may be initiated (each day, Corticosteroid therapy was commonly used
statistical methods that consider measured in our case). Although we used multiple in critically ill patients with MERS. After
baseline and time-varying confounders. a priori selected data time points to adjustment for baseline and time-varying
However, these adjustments may not be adjust for potential changes in severity of confounders, the use of corticosteroid
fully account for measured confounders and illness, we do not have data for each day; therapy was not associated 90-day mortality
would not account for indication bias that we needed to use imputation for days but was associated with delayed MERS-CoV
could occur in the presence of unmeasured with missing data, and it is possible that RNA clearance. n
confounding. Because follow-up additional data points would lead to
MERS-CoV RNA testing was at the even more valid estimates. However, 80% Author disclosures are available with the text
discretion of the treating team, it is of patients were started on corticosteroid of this article at www.atsjournals.org.

References 9. Rodrigo C, Leonardi-Bee J, Nguyen-Van-Tam JS, Lim WS. Effect of

corticosteroid therapy on influenza-related mortality: a systematic
1. World Health Organization. Middle East respiratory syndrome review and meta-analysis. J Infect Dis 2015;212:183–194.
coronavirus (MERS-CoV). 2017 [accessed 2017 Apr 17]. Available 10. Delaney JW, Pinto R, Long J, Lamontagne F, Adhikari NK,
from: http://www.who.int/emergencies/mers-cov/en/. Kumar A, et al.; Canadian Critical Care Trials Group H1N1
2. Arabi YM, Balkhy HH, Hayden FG, Bouchama A, Luke T, Baillie JK, Collaborative. The influence of corticosteroid treatment on the
et al. Middle East respiratory syndrome. N Engl J Med 2017;376: outcome of influenza A(H1N1pdm09)-related critical illness. Crit
584–594. Care 2016;20:75.
3. Arabi YM, Al-Omari A, Mandourah Y, Al-Hameed F, Sindi AA, Alraddadi B, 11. Shintani AK, Girard TD, Eden SK, Arbogast PG, Moons KG, Ely EW.
et al.; Saudi Critical Care Trial Group. Critically ill patients with the Immortal time bias in critical care research: application of time-
Middle East respiratory syndrome: a multicenter retrospective cohort varying Cox regression for observational cohort studies. Crit Care
study. Crit Care Med 2017;45:1683–1695. Med 2009;37:2939–2945.
4. Yam LY, Lau AC, Lai FY, Shung E, Chan J, Wong V; Hong Kong Hospital 12. Sjoding MW, Luo K, Miller MA, Iwashyna TJ. When do confounding by
Authority SARS Collaborative Group (HASCOG). Corticosteroid indication and inadequate risk adjustment bias critical care studies?
treatment of severe acute respiratory syndrome in Hong Kong. J Infect A simulation study. Crit Care 2015;19:195.
2007;54:28–39. 13. Suissa S. Effectiveness of inhaled corticosteroids in chronic obstructive
5. Auyeung TW, Lee JS, Lai WK, Choi CH, Lee HK, Lee JS, et al. The use pulmonary disease: immortal time bias in observational studies.
of corticosteroid as treatment in SARS was associated with Am J Respir Crit Care Med 2003;168:49–53.
adverse outcomes: a retrospective cohort study. J Infect 2005;51: 14. Arabi YM, Mandourah Y, Al-Hameed F, Al Omari A, Sindi A,
98–102. Alraddadi B, et al. The association of corticosteroid therapy
6. Stockman LJ, Bellamy R, Garner P. SARS: systematic review of and the outcome of critically ill patients with the Middle East
treatment effects. PLoS Med 2006;3:e343. respiratory syndrome [abstract]. Am J Respir Crit Care Med 2017;
7. Lee N, Allen Chan KC, Hui DS, Ng EK, Wu A, Chiu RW, et al. Effects of 195:A6868.
early corticosteroid treatment on plasma SARS-associated 15. The International Severe Acute Respiratory and Emerging Infection
coronavirus RNA concentrations in adult patients. J Clin Virol 2004;31: Consortium (ISARIC) [accessed 2017 Nov 15]. Available from:
304–309. https://isaric.tghn.org/.
8. Lee N, Leo YS, Cao B, Chan PK, Kyaw WM, Uyeki TM, et al. 16. Vincent JL, de Mendonça A, Cantraine F, Moreno R, Takala J,
Neuraminidase inhibitors, superinfection and corticosteroids Suter PM, et al. Use of the SOFA score to assess the
affect survival of influenza patients. Eur Respir J 2015; incidence of organ dysfunction/failure in intensive care
45:1642–1652. units: results of a multicenter, prospective study. Working

766 American Journal of Respiratory and Critical Care Medicine Volume 197 Number 6 | March 15 2018
ORIGINAL ARTICLE

group on “sepsis-related problems” of the European 24. Martin-Loeches I, Lisboa T, Rhodes A, Moreno RP, Silva E,
Society of Intensive Care Medicine. Crit Care Med 1998;26: Sprung C, et al.; ESICM H1N1 Registry Contributors. Use of early
1793–1800. corticosteroid therapy on ICU admission in patients affected by
17. Hernán MA, Brumback B, Robins JM. Marginal structural models severe pandemic (H1N1)v influenza A infection. Intensive Care Med
to estimate the causal effect of zidovudine on the survival of 2011;37:272–283.
HIV-positive men. Epidemiology 2000;11:561–570. 25. The World Health Orgnization. Clinical management of severe acute
18. Faries DE, Kadziola ZA. Analysis of longitudinal observational data respiratory infection when Middle East respiratory syndrome
using marginal structural models. In: Faries DE, Leon AC, Haro JM, coronavirus (MERS-CoV) infection is suspected. Interim guidance
Obenchain RL, editors. Analysis of observational health care data [updated 2015, July 2; accessed 2017 Sept 6]. Available from:
using SAS. Cary, NC: SAS Institute; 2010. pp. 211–230. http://www.who.int/csr/disease/coronavirus_infections/case-
19. Robins JM, Hernán MA, Brumback B. Marginal structural models management-ipc/en/.
and causal inference in epidemiology. Epidemiology 2000;11: 26. Torres A, Sibila O, Ferrer M, Polverino E, Menendez R, Mensa J, et al.
550–560. Effect of corticosteroids on treatment failure among hospitalized
20. Cao B, Gao H, Zhou B, Deng X, Hu C, Deng C, et al. Adjuvant patients with severe community-acquired pneumonia and high
corticosteroid treatment in adults with influenza A (H7N9) viral inflammatory response: a randomized clinical trial. JAMA 2015;313:
pneumonia. Crit Care Med 2016;44:e318–e328. 677–686.
21. Rochwerg B, Oczkowski S, Siemieniuk RA, Menon K, Szczeklik W, 27. Blum CA, Nigro N, Briel M, Schuetz P, Ullmer E, Suter-Widmer I,
English S, et al. Corticosteroids in sepsis: an updated systematic et al. Adjunct prednisone therapy for patients with community-
review and meta-analysis (protocol). BMJ Open 2017;7:e016847. acquired pneumonia: a multicentre, double-blind, randomised,
22. Kim SH, Hong SB, Yun SC, Choi WI, Ahn JJ, Lee YJ, et al.; Korean placebo-controlled trial. Lancet 2015;385:1511–1518.
Society of Critical Care Medicine H1N1 Collaborative. Corticosteroid 28. Snijders D, Daniels JM, de Graaff CS, van der Werf TS, Boersma WG.
treatment in critically ill patients with pandemic influenza A/H1N1 Efficacy of corticosteroids in community-acquired pneumonia: a
2009 infection: analytic strategy using propensity scores. Am J randomized double-blinded clinical trial. Am J Respir Crit Care Med
Respir Crit Care Med 2011;183:1207–1214. 2010;181:975–982.
23. Brun-Buisson C, Richard JC, Mercat A, Thiébaut AC, Brochard L; 29. Ruan SY, Lin HH, Huang CT, Kuo PH, Wu HD, Yu CJ. Exploring the
REVA-SRLF A/H1N1v 2009 Registry Group. Early corticosteroids in heterogeneity of effects of corticosteroids on acute respiratory
severe influenza A/H1N1 pneumonia and acute respiratory distress distress syndrome: a systematic review and meta-analysis. Crit Care
syndrome. Am J Respir Crit Care Med 2011;183:1200–1206. 2014;18:R63.

Arabi, Mandourah, Al-Hameed, et al.: Corticosteroids for Critically Ill Patients with MERS 767