In patients with coronavirus disease 2019 (COVID-19) and moderate or severe acute respiratory distress syndrome (ARDS), does intravenous dexamethasone plus standard care compared with standard care alone increase the number of days alive and free from mechanical ventilation?
In this randomized clinical trial that included 299 patients, the number of days alive and free from mechanical ventilation during the first 28 days was significantly higher among patients treated with dexamethasone plus standard care when compared with standard care alone (6.6 days vs 4.0 days).
Intravenous dexamethasone plus standard care, compared with standard of care alone, resulted in a statistically significant increase in the number of days alive and free of mechanical ventilation over 28 days.
Acute respiratory distress syndrome (ARDS) due to coronavirus disease 2019 (COVID-19) is associated with substantial mortality and use of health care resources. Dexamethasone use might attenuate lung injury in these patients.
To determine whether intravenous dexamethasone increases the number of ventilator-free days among patients with COVID-19–associated ARDS.
Design, Setting, and Participants
Multicenter, randomized, open-label, clinical trial conducted in 41 intensive care units (ICUs) in Brazil. Patients with COVID-19 and moderate to severe ARDS, according to the Berlin definition, were enrolled from April 17 to June 23, 2020. Final follow-up was completed on July 21, 2020. The trial was stopped early following publication of a related study before reaching the planned sample size of 350 patients.
Twenty mg of dexamethasone intravenously daily for 5 days, 10 mg of dexamethasone daily for 5 days or until ICU discharge, plus standard care (n =151) or standard care alone (n = 148).
Main Outcomes and Measures
The primary outcome was ventilator-free days during the first 28 days, defined as being alive and free from mechanical ventilation. Secondary outcomes were all-cause mortality at 28 days, clinical status of patients at day 15 using a 6-point ordinal scale (ranging from 1, not hospitalized to 6, death), ICU-free days during the first 28 days, mechanical ventilation duration at 28 days, and Sequential Organ Failure Assessment (SOFA) scores (range, 0-24, with higher scores indicating greater organ dysfunction) at 48 hours, 72 hours, and 7 days.
A total of 299 patients (mean [SD] age, 61  years; 37% women) were enrolled and all completed follow-up. Patients randomized to the dexamethasone group had a mean 6.6 ventilator-free days (95% CI, 5.0-8.2) during the first 28 days vs 4.0 ventilator-free days (95% CI, 2.9-5.4) in the standard care group (difference, 2.26; 95% CI, 0.2-4.38; P = .04). At 7 days, patients in the dexamethasone group had a mean SOFA score of 6.1 (95% CI, 5.5-6.7) vs 7.5 (95% CI, 6.9-8.1) in the standard care group (difference, −1.16; 95% CI, −1.94 to −0.38; P = .004). There was no significant difference in the prespecified secondary outcomes of all-cause mortality at 28 days, ICU-free days during the first 28 days, mechanical ventilation duration at 28 days, or the 6-point ordinal scale at 15 days. Thirty-three patients (21.9%) in the dexamethasone group vs 43 (29.1%) in the standard care group experienced secondary infections, 47 (31.1%) vs 42 (28.3%) needed insulin for glucose control, and 5 (3.3%) vs 9 (6.1%) experienced other serious adverse events.
Conclusions and Relevance
Among patients with COVID-19 and moderate or severe ARDS, use of intravenous dexamethasone plus standard care compared with standard care alone resulted in a statistically significant increase in the number of ventilator-free days (days alive and free of mechanical ventilation) over 28 days.
ClinicalTrials.gov Identifier: NCT04327401
Three months after the emergence of the coronavirus disease 2019 (COVID-19)1 caused by the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the World Health Organization declared it a pandemic.2 Estimates have suggested that up to 12% of patients hospitalized with COVID-19 have required invasive mechanical ventilation,3,4 with the majority developing acute respiratory distress syndrome (ARDS).5 Diffuse alveolar damage with hyaline membranes,6 hallmarks of ARDS, have been found on pulmonary histological examination of patients with COVID-19. Furthermore, an uncontrolled inflammatory state is frequent with COVID-197,8 and may contribute to multiorgan failure in these patients. Corticosteroids might exert an effect in controlling this exacerbated response.9
Several trials evaluated the role of corticosteroids for non–COVID-19 ARDS with conflicting results.10,11 Observational studies of other viral diseases suggested that corticosteroids might increase viral load in patients with SARS-CoV12 and Middle East respiratory syndrome (MERS).13 A meta-analysis identified an association between corticosteroids and higher mortality among patients with influenza.14 Findings from a randomized clinical trial involving patients with COVID-19 indicated that the use of dexamethasone decreased mortality in hospitalized patients requiring supplemental oxygen or mechanical ventilation.15
The COVID-19 Dexamethasone (CoDEX) randomized clinical trial was conducted to evaluate the efficacy of intravenous dexamethasone in patients with moderate to severe ARDS due to COVID-19. The hypothesis was that dexamethasone would increase the number of days alive and free from mechanical ventilation during the first 28 days.
Study Design and Oversight
We conducted an investigator-initiated, multicenter, randomized, open-label, clinical trial in 41 intensive care units (ICUs) in Brazil. The trial protocol (Supplement 1) and the statistical analysis plan were submitted for publication before the first interim analysis16 (Supplement 2). The study was approved at the Brazilian Health Regulatory Agency, the Brazilian National Commission for Research Ethics, and all ethics committees at the participating sites. Written or oral informed consent was obtained before randomization from each patient’s legal representative. The trial was overseen by an external and independent data and safety monitoring committee (DSMC).
Patients were enrolled who were at least 18 years old, had confirmed or suspected COVID-19 infection (eMethods in Supplement 3), and were receiving mechanical ventilation within 48 hours of meeting criteria for moderate to severe ARDS with partial pressure of arterial blood oxygen to fraction of inspired oxygen (Pao2:Fio2)ratio of 200 or less. An ARDS diagnosis was made according to the Berlin Definition criteria.17 Exclusion criteria were pregnancy or active lactation, known history of dexamethasone allergy, corticosteroid use in the past 15 days for nonhospitalized patients, use of corticosteroids during the present hospital stay for more than 1 day, indication for corticosteroid use for other clinical conditions (eg, refractory septic shock), use of immunosuppressive drugs, cytotoxic chemotherapy in the past 21 days, neutropenia due to hematological or solid malignancies with bone marrow invasion, consent refusal, or expected death in the next 24 hours (Figure 1). During the study period we refined some of the inclusion and exclusion criteria. Full details are provided in Supplement 3.
Randomization was performed through an online web-based system18 using computer-generated random numbers and blocks of 2 and 4, unknown to the investigators, and was stratified by center. The group treatment was disclosed to the investigator only after all information regarding patient enrollment was recorded in the online system (eMethods in Supplement 3).
Eligible patients were randomly assigned in a 1:1 ratio to receive dexamethasone 20 mg intravenously once daily for 5 days, followed by 10 mg intravenously once daily for additional 5 days or until ICU discharge, whichever occurred first, plus standard care. Patients in the control group received standard care only. Physicians, patients, and individuals who assessed the outcomes were not blinded for the assigned treatment. Each study center was encouraged to follow the best practice guidelines and their institutional protocol for the care of critically ill patients with COVID-19. All clinical interventions, such as use of antibiotics, ventilatory strategy, laboratory testing, and hemodynamic management were left at the discretion of the ICU team for both groups.
Protocol adherence was assessed daily until day 10. Unjustified corticosteroid use or use for treating ARDS or COVID-19 in the control group was not recommended and considered a protocol deviation. The use of nonstudy corticosteroids was permitted in the control group for usual ICU indications, such as bronchospasm and refractory septic shock.19 Additionally, any dexamethasone dosage change or early interruption in the intervention group was considered a protocol violation.
Clinical and Laboratory Data
Data on demographic characteristics, physiological variables, corticosteroid use before randomization, timing from ARDS diagnosis to randomization, insulin use for hyperglycemia, and other clinical and laboratory data were collected. Use of neuromuscular blocking agents, prone positioning, and extracorporeal membrane oxygenation (ECMO) were collected daily through day 14. Use of mechanical ventilation and other oxygen supportive therapies (high-flow nasal cannula, noninvasive positive pressure ventilation, and use of supplemental oxygen) were collected daily through 28 days. Diagnosis of new infections were reported daily through day 28. Individual patient data on infections were adjudicated by a blinded investigator (eMethods in Supplement 3). Patients were followed up for 28 days after randomization or until hospital discharge, whichever occurred first.
The primary outcome was ventilator-free days during the first 28 days, defined as the number of days alive and free from mechanical ventilation for at least 48 consecutive hours.20 Patients discharged from the hospital before 28 days were considered alive and free from mechanical ventilation at 28 days. Nonsurvivors at day 28 were considered to have no ventilator-free days. More details on the definitions are provided in the eMethods section of Supplement 3.
Prespecified secondary outcomes were all-cause mortality during 28 days, clinical status of patients at day 15 using a 6-point ordinal scale adapted from the World Health Organization R&D Blueprint expert group21—(1) not hospitalized, (2) hospitalized, not requiring supplemental oxygen, (3) hospitalized, requiring supplemental oxygen, (4) hospitalized, requiring noninvasive ventilation or nasal high-flow oxygen therapy, (5) hospitalized, requiring invasive mechanical ventilation or ECMO, and (6) death; ICU-free days during the first 28 days; mechanical ventilation duration at 28 days; and Sequential Organ Failure Assessment (SOFA) scores, which range from 0 to 24, with higher scores indicating greater dysfunction, at 48 hours, 72 hours, and 7 days. For post hoc analyses, we evaluated the components of ventilator-free days during the first 28 days, the cumulative proportions of the 6-point ordinal scale at 15 days, and the outcome of discharge from hospital alive within 28 days. For patients who died, the number of ventilator-free days was 0; for patients who were alive, the ventilator-free days were the days they did not require mechanical ventilation.
No reliable data were available at the trial design to allow for an accurate sample size calculation. Therefore, we used data from a multicenter randomized trial of non–COVID-19 ARDS in Brazil22 for our sample size calculation. We originally estimated a 2-sided α level of .05 and power of 80% to detect a difference of 3 ventilator-free days between groups; assuming a mean of 8 (SD, 9) ventilator-free days in the control group, 290 patients had to be enrolled. Before the first interim analysis, without any study data review and after discussing the protocol with the DSMC, the study steering committee decided to increase the sample size to 350 patients based on necessary adjustments regarding the uncertainty about the normality of the distribution of ventilator-free days. Thus, the original sample size was increased by 15% based on the Pitman asymptotic relative efficiency23 to preserve study power.
Two preplanned interim analyses for efficacy and safety evaluation after 96 and 234 patients with complete follow-up were programmed. The stopping rule for safety was P < .01 and for efficacy P < .001 (Haybittle–Peto boundary).24 There was no adjustment in the final threshold for statistical significance for sequential analysis.
To estimate treatment effects on the primary outcome, a generalized linear model was used with 0-1 inflated beta-binomial distribution, with center as random effect and adjusted for age and the Pao2:Fio2 ratio at randomization. The effect size was estimated as mean difference and its respective 95% confidence interval.
The all-cause mortality rate at 28 days was analyzed using a mixed Cox model, with centers as the random effects. The treatment effect on the SOFA score at 48 hours, 72 hours, and 7 days after randomization was analyzed by a linear mixed model with patients as random effects adjusted for the baseline SOFA score. For the clinical status of patients, if the proportional odds assumption was met, a mixed ordinal logistic regression was used. All secondary outcomes were adjusted for age and the Pao2:Fio2 ratio to increase statistical power and improve the efficiency of the analysis. Further details on model assumptions and model fit are provided in the eMethods section of Supplement 3. Adverse events are expressed as counts and percentages and compared between groups using the χ2 test.
All patients were included in the primary analysis. There was no loss to follow-up, and data on the primary outcome, mortality within 28 days, clinical status at day 15, ICU-free days at 28 days, and mechanical ventilation duration were available for all patients. Missing values on individual SOFA components were imputed as normal (eMethods in Supplement 3). We assessed the consistency of the primary analysis results through prespecified sensitivity analyses considering the per-protocol population, patients who received corticosteroids vs patients who did not (as-treated population), patients with confirmed COVID-19, and patients with confirmed or probable COVID-19 (eMethods in Supplement 3).
We performed prespecified subgroup analysis on the primary outcome testing interactions for age (<60 and ≥60 years), Pao2:Fio2 ratio (≤100 and >100), symptoms duration at randomization (≤7 and >7 days), Simplified Acute Physiology Score III (SAPS III) (<60 and ≥60), position at randomization (prone or supine), and use of vasopressor at randomization (eMethods in Supplement 3).
Patients were analyzed according to their randomization groups, and no adjustments for multiplicity were performed. Thus, the results of secondary outcomes and subgroup analyses should be interpreted as exploratory. A 2-sided P value of less than .05 was considered statistically significant. All analyses were performed using the R software version 4.0.2 (R Core Team).
On June 25, 2020, the DSMC discussed the implications of the results of the dexamethasone group in the RECOVERY (Randomized Evaluation of COVID-19 Therapy) trial,15 stating that given the study results,15 it was no longer ethical to continue the trial, which led to the recommendation to stop the trial. This recommendation was accepted by the CoDEX Steering Committee on June 25, 2020 (eMethods in Supplement 3).
From April 17 to June 23, 2020, 299 patients were randomized. Of the enrolled patients, 151 were randomly assigned to receive dexamethasone and 148 to the control group (Figure 1).
Baseline characteristics were well balanced between groups (Table 1; eTable 1 in Supplement 3), including severity of ARDS and the use of rescue therapies at randomization. Remdesivir was not available in Brazil during the trial period. Only 1 patient received lopinavir-ritonavir treatment. Other therapeutic strategies such as tocilizumab and convalescent plasma were limited and not widely available.
Only 1 patient in the intervention group did not receive any dexamethasone. The rate of dexamethasone use within 10 days was 94.8 per 100 patient-days (eTable 2 in Supplement 3). The median duration of dexamethasone treatment was 10 days (interquartile range [IQR], 6-10 days). In the standard care group, 52 patients (35.1%) received at least 1 dose of corticosteroids, of whom 38 (73.1%) had other established clinical indications for corticosteroid use. The use of corticosteroids in 14 patients (9.4%) was considered a protocol deviation, and the rate of corticosteroid use within 10 days was 16.5 per 100 patient-days (eTable 3 in Supplement 3).
The mean number of days alive and free from mechanical ventilation during the first 28 days was significantly higher in the dexamethasone group than in the standard care group (6.6; 95% CI, 5.0-8.2 days vs 4.0; 95% CI, 2.9-5.4 days; difference, 2.26; 95% CI, 0.2-4.38; P = .04) (Table 2; eFigure 1 in Supplement 3). The cumulative frequency of ventilator-free days according to study group is shown in Figure 2.
Secondary Outcomes and Adverse Events
There was no significant difference in all-cause mortality at 28 days (56.3% in the dexamethasone group vs 61.5% the standard care group; hazard ratio, 0.97; 95% CI, 0.72 to 1.31; P = .85), in the 6-point ordinal scale at day 15 (median, 5; IQR, 3-6 for the dexamethasone group vs median, 5; IQR, 5-6 for standard care group; odds ratio [OR], 0.66; 95% CI, 0.39 to 1.13; P = .07), ICU-free days at 28 days (mean, 2.1; 95% CI, 1.0 to 4.5 days for the dexamethasone group vs mean, 2.0; 95% CI, 0.8 to 4.2 days for the standard care group; difference, 0.28; 95% CI, −0.49 to 1.02; P = .50), and mechanical ventilation duration (12.5; 95% CI, 11.2 to 13.8 days for the dexamethasone group vs 13.9, 95% CI, 12.7 to 15.1 days for the standard care group; difference, −1.54; 95% CI, −3.24 to −0.12; P = .11). The mean SOFA score at 7 days was significantly lower in the treatment group (6.1; 95% CI, 5.5 to 6.7 for dexamethasone vs 7.5; 95% CI, 6.9 to 8.1 for standard care; difference, −1.16; 95% CI, −1.94 to −0.38; P = .004) (Table 2).
Both groups had a comparable need for insulin use for hyperglycemia: 47 patients (31.1%) in the dexamethasone group vs 42 (28.4%) in the standard care group. The number of new diagnoses of infection until day 28 was 33 (21.9%) vs 43 (29.1%). Twelve patients (7.9%) in the dexamethasone group had bacteremia vs 14 (9.5%) in the standard care group. Five patients (3.3%) had serious adverse events vs 9 (6.1%) (Table 3; eTable 4 in Supplement 3).
Subgroup and Exploratory Analyses
In subgroup analyses, tests for interaction were not statistically significant for subgroups defined by age (P = .21), Pao2:Fio2 ratio (P = .73), SAPS III (P = .75), time since symptom onset (P = .12), position at randomization (P = .89), and vasopressor use at randomization (P = .81) (eFigure 2 in Supplement 3).
The post hoc analyses showed no significant difference of the intervention in the components of the primary outcome or in the outcome of discharged alive within 28 days (eTable 6 in Supplement 3). Patients in the dexamethasone group had significantly lower cumulative probability of having died or being mechanically ventilated at day 15 (categories 5-6 on the 6-point scale) than the standard care group (67.5% vs 80.4%; OR, 0.46; 95% CI, 0.26 to 0.81; P = .01) (eTable 6 and eFigure 3 in Supplement 3). In the sensitivity analyses for the primary outcome of ventilator-free days, the treatment effect was not significantly different in the as-treated analysis. The mean number of ventilator-free days was 5.8 (95% CI, 4.6 to 7.3) among 203 patients in the dexamethasone group vs 4.1 (95% CI, 2.6 to 5.5) among 96 patients in the standard care group, for a mean difference of 2.38 (95% CI, −0.6 to 3.32; P = .16). In the per-protocol analysis, the mean number of ventilator-free days among dexamethasone group was 6.4 (95% CI, 5.1 to 8.1) among 125 patients vs 4.1 (95% CI, 2.6 to 5.5) among 96 patients in the standard care group for a difference of 2.36 (95% CI, −0.15 to 4.56; P = .06). The main results remained statistically significant among patients with confirmed COVID-19 in the dexamethasone group, which had a mean number of ventilator-free days of 6.8 (95% CI, 5.4 to 8.4) among 144 patients vs 3.9 (95% CI, 2.7 to 5.1) among 142 patients in the standard care group for a difference of 2.7 (95% CI, 0.8 to 4.74; P = .01). Among the patients with confirmed or probable COVID-19, the mean number of ventilator-free days was 6.6 (95% CI, 5.3 to 8.2) among 151 patients vs 4.1 (95% CI, 2.9 to 5.2) among 147 patients for a difference of 2.38 (95% CI, 0.48 to 4.33; P = .02) (eTable 7 in Supplement 3).
In this randomized clinical trial involving 299 adults with moderate or severe ARDS due to COVID-19, dexamethasone plus standard care compared with standard care alone significantly increased the number of days alive and free of mechanical ventilation during the first 28 days. Dexamethasone was not associated with increased risk of adverse events in this population of critically ill COVID-19 patients.15
This trial included only patients with COVID-19 and moderate or severe ARDS and provided laboratory, physiological, and adverse events data on the use of corticosteroids in this population. The ventilator-free days criterion was chosen as the primary outcome because it comprises both mortality and ventilation duration in surviving patients. The number of days alive and free from mechanical ventilation at 28 days was significantly lower than reported in other trials of non–COVID-19 ARDS,10,11,25 but consistent with COVID-19 ARDS studies, confirming the disease severity.26 The difference between groups of 2.26 days was lower than the effect size of 3 days used in the sample size calculation. This reduction is relevant in the context of a pandemic, in which an inexpensive, safe, and widely available intervention like dexamethasone increases even modestly the number of ventilator-free days and may reduce the risk of ventilatory complications, ICU length of stay, and burden to the health care system.
Mortality rates were high and not significantly different between groups, in contrast with the RECOVERY trial of dexamethasone in patients hospitalized for COVID-1915 and a trial of dexamethasone in patients with non-COVID-19 ARDS.11 The high mortality rate might be explained by several factors. The patients had a high risk of death as shown by the low mean Pao2:Fio2 ratio and mean SAPS III score of 70, which represents a mortality risk of 70.9% in South America.27,28 In a previous randomized clinical trial, moderate to severe ARDS not caused by COVID-19 had an elevated mortality rate in Brazil of 52%,22 and recent data collected by Brazilian Association of Critical Care demonstrated mortality rates of 66% to 70% for ventilated patients with COVID-19 in Brazilian ICUs.29 This may be explained by the pandemic and its burden to the health care system, especially in a country with limited resources like Brazil. However, even in high-income countries the mortality rate in ventilated patients with COVID-19 might range from 54% to 88%.30-32 This mortality rate may be similar to that of other low and middle-income countries and is important to consider when translating the scientific evidence to clinical practice. In this sense, the results of this trial expand those of the RECOVERY trial15 by showing that corticosteroids were effective even when the baseline mortality rate was high.
The dexamethasone dose was chosen based on a previous11 trial showing the benefit of dexamethasone to patients with non–COVID-19 ARDS. Previous data suggest that high doses of corticosteroids (the equivalent of 30 mg/d of dexamethasone) in viral pneumonia may be associated with unfavorable outcomes.33 However, there are no currently available data from patients with COVID-19 to determine if higher doses are harmful. In the present study, the number of adverse events, new infections, and the use of insulin were comparable in both groups, in line with previous studies that did not demonstrate an augmented risk of adverse events with corticosteroids in non-COVID-19 ARDS.10,11,19
This trial has several strengths. Bias was controlled by ensuring allocation concealment, all patients were analyzed according to their randomization group, and follow-up was complete. Also, adverse events data regarding corticosteroid use among patients with COVID-19 were provided, along with detailed data on ventilatory parameters, ARDS treatment, and laboratory and physiological variables.
This study has several limitations. First, it was an open-label trial due to time constraints of producing placebo in a pandemic scenario with an urgent need for reliable and randomized data. Second, 35% of the patients in the control group received corticosteroids during the study period, possibly related to the open-label design, the disease severity of the patients, and other diverse indications for corticosteroid use in critical care.19 However, the use of corticosteroids in the control group would have biased the results toward the null, and the study identified a benefit of the intervention on the primary outcome. Third, the open-label design and investigator-reported data on adverse events and infections may have led to bias in the description of these events. Fourth, the trial was underpowered for important secondary outcomes like mortality and the study was interrupted before the original sample size was obtained due to external evidence of benefit, and the obtained sample size was limited to demonstrate benefits in secondary outcomes.
In patients with COVID-19 and moderate or severe ARDS, use of intravenous dexamethasone plus standard care, compared with standard care alone, resulted in a statistically significant increase increase in the number of ventilator-free days (days alive and free of mechanical ventilation) over 28 days.
Corresponding Author: Luciano C. P. Azevedo, MD, PhD, Hospital Sirio-Libanes, Rua Prof Daher Cutait, 69, 01308-060, São Paulo, Brazil (firstname.lastname@example.org).
Accepted for Publication: August 20, 2020.
Published Online: September 2, 2020. doi:10.1001/jama.2020.17021
Author Contributions: Drs Tomazini and Azevedo had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Mr Damiani conducted and is responsible for the data analysis.
Concept and design: Tomazini, Maia, Cavalcanti, Berwanger, Veiga, Lopes, Bueno, Baldassare, Damiani, Lisboa, Zampieri, Fernandes, Morais, Zung, Machado, Azevedo.
Acquisition, analysis, or interpretation of data: Tomazini, Cavalcanti, Berwanger, Rosa, Veiga, Avezum, Lopes, Bueno, Silva, Baldassare, E. Costa, Moura, Honorato, A. Costa, Damiani, Lisboa, Kawano-Dourado, Olivato, Righy, Amendola, Roepke, D. Freitas, Forte, F. Freitas, Melro, Junior, Machado, Azevedo.
Drafting of the manuscript: Tomazini, Berwanger, Veiga, Bueno, Baldassare, Kawano-Dourado, Junior, Machado, Azevedo.
Critical revision of the manuscript for important intellectual content: Tomazini, Maia, Cavalcanti, Berwanger, Rosa, Veiga, Avezum, Lopes, Bueno, Silva, Baldassare, E. Costa, Moura, Honorato, A. Costa, Damiani, Lisboa, Kawano-Dourado, Zampieri, Olivato, Righy, Amendola, Roepke, D. Freitas, Forte, F. Freitas, Fernandes, Melro, Morais, Zung, Machado, Azevedo.
Statistical analysis: Tomazini, Berwanger, Bueno, Damiani, Lisboa.
Obtained funding: Berwanger, Bueno, Baldassare, Lisboa, Morais, Zung, Azevedo.
Administrative, technical, or material support: Tomazini, Rosa, Bueno, Silva, Baldassare, Moura, Honorato, A. Costa, Lisboa, Righy, Roepke, Fernandes, Junior, Morais, Zung, Azevedo.
Supervision: Tomazini, Maia, Veiga, Avezum, Bueno, Baldassare, E. Costa, Moura, Zampieri, Roepke, Fernandes, Junior, Machado, Azevedo.
Conflict of Interest Disclosures: Dr Tomazini reported receiving support from Aché pharmaceutical. Dr Maia reported receiving nonfinancial support from Aché Laboratórios Farmacêuticos. Dr Cavalcanti reported receiving grants from Bayer, Bactiguard, Johnson & Johnson, do Brasil, Hemaclear, Hillrom, and Pfizer. Dr Berwanger reported receiving grants from AstraZeneca, Novartis, Servier, Bayer, Amgen, and Boehringer-Ingelheim. Dr Lopes reported receiving personal fees from Bayer, Boehinger Ingleheim, Daiichi Sankyo, Merck, and Portola; grants and personal fees from Bristol-Myers Squibb, GlaxoSmithKline, Medtronic, Pfizer, Portola, and Sanofi. Ms Bueno reported receiving personal fees from Endpoint Health. Dr Silva reported receiving support from Aché Laboratórios Farmacêuticos. Mrs Baldassare reported receiving grants from Aché Laboratórios Farmacêuticos. Dr Moura reported receiving personal fees from Hospital Sírio-Libanês. Dr A. Costa reported receiving grants from Pfizer. Dr Fernandes reported receiving grants from Hospital Sírio Libanês and from Aché Laboratórios Farmacêuticos S.A. Mr Morais reported receiving personal fees and other support from Aché Laboratórios Farmacêuticos. Dr Zung reported receiving personal fees from Aché Laboratórios Farmacêuticos. Dr Machado reported receiving support from Laboratórios Farmacêuticos. Dr Azevedo reported receiving grants from Aché Laboratórios and personal fees from Pfizer and Halex-Istar. No other disclosures were reported.
Funding/Support: This trial was funded and supported by the Coalition COVID-19 Brazil. The Laboratórios Farmacêuticos provided the study drug, distribution logistics, and insurance for the study patients.
Role of the Funder/Sponsor: Laboratórios Farmacêuticos had no role in the design and conduct of the study. The Coalition COVID-19 Brazil was responsible for the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and the decision to submit the manuscript for publication.
Site Investigators: All in Brazil: Hospital Vila Santa Catarina, São Paulo: Adriano José Pereira, Guilherme Benfatti Olivato, Natalie Botelho Borges, and Ana Lucia Neves; Instituto Estadual do Cérebro, Rio de Janeiro: Cássia Righy, Pedro Kurtz, Ricardo Turon, and Marília Gomes e Silva; Hospital do Câncer de Barretos, Barretos: Cristina Prata Amendola, Luciana Coelho Sanches, Luis Henrique Simões Covello, and André Luiz Tosello Penteado; UTI Emergências Cirúrgicas e Trauma–HCFMUSP, São Paulo: Bruno M. Tomazini, Roberta Muriel Longo Roepke, and Estevão Bassi; UTI Respiratória–HCFMUSP, São Paulo: Eduardo Leite Vieira Costa, Marcelo Britto Passos Amato, Daniela Helena Machado de Freitas, and Carlos R. Carvalho; Hospital São Paulo, Universidade Federal de São Paulo–UNIFESP, São Paulo: Flavia Ribeiro Machado, Flávio Geraldo Rezende Freitas, Maria Aparecida de Souza, and Fernando José da Silva Ramos; UTI 09DN–HCFMUSP: Daniel Neves Forte, José Mauro Vieira Júnior, Sâmia Yasin Wayhs, Veridiana Schulz Casalechi, and Ricardo Antônio Bonifácio Moura; Hospital Estadual Mario Covas–FMABC, Santo André: Caio Cesar Ferreira Fernandes, Marcelo Rodrigues Bacci, Antônio Carlos Palandri Chagas, and Desirè Carlos Callegari; Hospital Samaritano, São Paulo: Livia Maria Garcia Melro, Yuri de Albuquerque Pessoa dos Santos, Anderson Roberto Dallazen, and Daniel Curitiba Marcellos; Hospital Evangélico de Vila Velha, Vila Velha: Gedealvares Francisco de Souza Júnior, Ana Carolina Simões Ramos, and Gláucia Gleine Souza Ferraz; Hospital Unimed Vitória, Vitória: Eliana Bernadete Caser and Danilo Hugo Brito Figueiredo; UTI da Disciplina de Emergências Clínicas—HCFMUSP: Bruno Adler Maccagnan Pinheiro Besen and Leandro Utino Taniguchi; Hospital Naval Marcílio Dias, Rio de Janeiro: Vicente Cés de Souza Dantas, Priscilla Alves Barreto, and Orlando Farias Jr; Hospital São José, Criciúma: Felipe Dal Pizzol and Cristiane Ritter; Hospital Israelita Albert Einstein, São Paulo: Otávio Berwanger, Remo H. M. Furtado, Thiago D. Correia, and Ary Serpa Neto; Hospital das Clínicas da Faculdade de Medicina de Botucatu–UNESP, Botucatu: Marina Politi Okoshi, Suzana Erico Tanni, and Aparecido Rios Queiroz; UTI Bloco Cirúrgico IV–HCFMUSP, São Paulo: Carlos Eduardo Pompilio and José Otto Reusing Jr; Hospital Sepaco, São Paulo: Flávio Geraldo Rezende de Freitas, Antônio Tonete Bafi, and Fernanda Regina de Campos Radziavicius; Hospital Municipal Dr. Moysés Deutsch (M’Boi Mirim), São Paulo: Felipe Maia de Toledo Piza, Airton L. O. Manoel, Niklas S. Campos; Hospital Regional Hans Dieter Schmidt, Joinville: Conrado Roberto Hoffmann Filho and Iara Caravajal Hoffmann; Unidade de Terapia Intensiva Cirúrgicas da Divisão de Anestesiologia–HCFMUSP, São Paulo: Luiz Marcelo Sá Malbouisson and Thiago Tavares dos Santos; Casa de Saúde Santa Marcelina, São Paulo: Luiz Relvas and Bruno Nunes Rodrigues; Beneficência Portuguesa, São Paulo: Viviane Cordeiro Veiga and Agnes Cohen Lisboa; Hospital Estadual Jayme dos Santos Neves, Serra: Priscila Aquino and Vinícius Santana Nunes; Hospital da Mulher do Recife, Recife: Mario Diego Teles Correia and Giselle Matias de Carvalho; Hospital Universitário de Maringá, Maringá: Sergio Yamada; Hospital do Coração, São Paulo: Alexandre Biasi Cavalcanti and Leticia Kawano-Dourado; UTI da Divisão de Anestesia–HCFMUSP, São Paulo: Pedro Vitale Mendes and João Manoel Silva Junior; Hospital Alemão Oswaldo Cruz, São Paulo: José Victor Gomes Costa and David J. B. Machado; Hospital Maternidade São Vicente de Paulo, Barbalha: Meton Soares De Alencar Filho and Jussara Alencar Arraes; Unimed Cariri, Juazeiro do Norte: Thales Anibal leite Barros Agostinho and Sérgio de Araújo; Santa Casa de Misericórdia de Passos, Passos: Priscila Freitas das Neves Gonçalves; Instituto do Coração (Incor)–FMUSP, São Paulo: Alexandre de Matos Soeiro; Hospital Baía Sul, Florianópolis: Israel Silva Maia and Ana Cristina Burigo; Hospital Sírio-Libanês, São Paulo: Bruno M. Tomazini and Luciano Cesar Pontes de Azevedo; Hospital Nereu Ramos, Florianópolis: Israel Silva Maia and Cassio Zandonai; Hospital Moinhos de Vento, Porto Alegre: Regis G. Rosa; Hospital de Brasília, Brasília: Rodrigo Santos Biondi; and UTI da Gastroenterologia–HCFMUSP, São Paulo: Rodolpho Augusto de Moura Pedro.
Trial Coordinating Center: Bruno Martins Tomazini, Flavia R. Bueno, Maria Vitoria A. O. Silva, Franca P. Baldassare, Eduardo Leite V. Costa, Ricardo A. B. Moura, Michele Honorato, Andre N. Costa, Camila S. J. C. Sampaio, Luciano CP Azevedo; Hospital Sirio-Libanes, São Paulo, Brazil.
Executive Committee: Luciano C. P. Azevedo, MD, PhD; Alexandre B. Cavalcanti, MD, PhD; Regis G. Rosa, MD, PhD; Alvaro Avezum, MD, PhD; Viviane C. Veiga, MD, PhD; Renato D. Lopes, MD, PhD; Flávia R. Machado, MD, PhD; and Otavio Berwanger, MD, PhD.
Steering Committee: Luciano C. P. Azevedo, MD, PhD; Alexandre B. Cavalcanti, MD, PhD; Regis G. Rosa, MD, PhD; Alvaro Avezum, MD, PhD; Viviane C. Veiga, MD, PhD; Renato D. Lopes, MD, PhD; Flávia R Machado, MD, PhD; Otavio Berwanger, MD, PhD; Fernando G. Zampieri, MD, PhD; Letícia Kawano-Dourado, MD, PhD; Thiago Lisboa, MD, PhD; Israel S. Maia, MD, MSc; Remo Furtado, MD, PhD; Henrique Fonseca, MD, PhD; Ary Serpa-Neto, MD, PhD; Thiago Correa, MD, PhD; Cláudio Galvão, MD, PhD; Leonardo R. Ferraz, MD, PhD; Guilherme Schettino, MD, PhD; Luiz V. Rizzo, MD, PhD; Maicon Falavigna, MD, PhD; Eduardo Leite Vieira Costa, MD, PhD; Bruno M. Tomazini, MD; Danielle Leão, MD, PhD; João Prats, MD, PhD; Philip Scheinberg MD, PhD; André Gobatto, MD, PhD; Cintia Grion, MD, PhD; Felipe Dal Pizzol, MD, PhD; Fernando A. Bozza, MD, PhD; Flavio G. R. Freitas, MD, PhD; Glauco Westphal, MD, PhD; Hugo Urbano, MD; Rodrigo Biondi, MD; and Rodrigo C. Figueiredo, MD.
Affiliations of the Executive Committee and Steering Committee: Hospital Sirio-Libanes, São Paulo, Brazil: Azevedo, Tomazini, and Eduardo Costa; Disciplina de Emergências Clínicas, Hospital das Clinicas HCFMUSP, Faculdade de Medicina, Universidade de São Paulo, São Paulo, Brazil: Azevedo; Brazilian Research in Intensive Care Network (BRICNet), Brazil: Azevedo, Cavalcanti, Rosa, Veiga, Machado, Zampieri, Lisboa, Maia, Gobatto, Grion, Dal Pizzol, Bozza, Freitas, Westphal, Urbano, Biondi, and Figueiredo; Hcor Research Institute, São Paulo, Brazil: Cavalcanti, Zampieri, Kawano-Dourado, Lisboa, and Maia; Hospital Moinhos de Vento, Porto Alegre, Brazil: Rosa and Falavigna; Hospital Alemão Oswaldo Cruz, São Paulo, Brazil: Avezum; BP–A Beneficência Portuguesa de São Paulo, São Paulo, Brazil: Veiga; Brazilian Clinical Research Institute, São Paulo, Brazil: Lopes; Duke University Medical Center-Duke Clinical Research Institute, Durham, North Carolina: Lopes; Anesthesiology, Pain, and Intensive Care Department, Federal University of São Paulo, São Paulo, Brazil: Machado and Freitas; and Academic Research Organization, Hospital Israelita Albert Einstein, São Paulo, Brazil: Berwanger, Furtado, Fonseca, Serpa-Neto, Correa, Galvão, Ferraz, Schettino, and Rizzo.
Data Monitoring and Safety Committee: Monash University, Melbourne, Australia; Carol Hodgson, PhD, FACP, BappSc(PT) Mphil PGDip(Cardio); Michael Bailey, BSc(Hons), MSc, PhD; University of Michigan, Ann Arbor: Theodore John Iwashyna, MD.
Disclaimer: This study was performed on behalf of the Coalition COVID-19 Brazil Group.
Data Sharing Statement: See Supplement 4.
Additional Contributions: We thank all the multidisciplinary teams in the participating centers for their support in following the study interventions in the challenging context of the COVID-19 pandemic.
et al; and the Northwell COVID-19 Research Consortium. Presenting characteristics, comorbidities, and outcomes among 5700 patients hospitalized with COVID-19 in the New York City area. JAMA
. 2020;323(20):2052-2059. doi:10.1001/jama.2020.6775PubMedGoogle ScholarCrossref
et al; ISARIC4C investigators. Features of 20 133 UK patients in hospital with COVID-19 using the ISARIC WHO Clinical Characterisation Protocol: prospective observational cohort study. BMJ
. Published online May 22, 2020. doi:10.1136/bmj.m1985PubMedGoogle Scholar
et al; COVID-19 Lombardy ICU Network. Baseline characteristics and outcomes of 1591 patients infected with SARS-CoV-2 admitted to ICUs of the Lombardy region, Italy. JAMA
. 2020;323(16):1574-1581. doi:10.1001/jama.2020.5394PubMedGoogle ScholarCrossref
et al; National Heart, Lung, and Blood Institute Acute Respiratory Distress Syndrome (ARDS) Clinical Trials Network. Efficacy and safety of corticosteroids for persistent acute respiratory distress syndrome. N Engl J Med
. 2006;354(16):1671-1684. doi:10.1056/NEJMoa051693PubMedGoogle ScholarCrossref
et al; RECOVERY Collaborative Group. Dexamethasone in hospitalized patients with Covid-19—preliminary report. N Engl J Med
. Published online July 7, 2020. doi:10.1056/NEJMoa2021436PubMedGoogle Scholar
et al. Guidelines for the diagnosis and management of critical illness-related corticosteroid insufficiency (CIRCI) in critically ill patients, I: Society of Critical Care Medicine (SCCM) and European Society of Intensive Care Medicine (ESICM) 2017. Crit Care Med
. 2017;45(12):2078-2088. doi:10.1097/CCM.0000000000002737PubMedGoogle ScholarCrossref
et al; WIND (Weaning according to a New Definition) Study Group and the REVA (Réseau Européen de Recherche en Ventilation Artificielle) Network ‡. Epidemiology of weaning outcome according to a new definition: the WIND Study. Am J Respir Crit Care Med
. 2017;195(6):772-783. doi:10.1164/rccm.201602-0320OCPubMedGoogle ScholarCrossref
et al; Writing Group for the Alveolar Recruitment for Acute Respiratory Distress Syndrome Trial (ART) Investigators. Effect of lung recruitment and titrated positive end-expiratory pressure (PEEP) vs low PEEP on mortality in patients with acute respiratory distress syndrome: a randomized clinical trial. JAMA
. 2017;318(14):1335-1345. doi:10.1001/jama.2017.14171PubMedGoogle ScholarCrossref
HJM. Nonparametrics: Statistical Methods Based on Ranks. Holden-Day; 1975.
et al; PRoVENT; PROVE Network investigators. Epidemiological characteristics, practice of ventilation, and clinical outcome in patients at risk of acute respiratory distress syndrome in intensive care units from 16 countries (PRoVENT): an international, multicentre, prospective study. Lancet Respir Med
. 2016;4(11):882-893. doi:10.1016/S2213-2600(16)30305-8PubMedGoogle ScholarCrossref
et al; COVID-19 Spanish ICU Network. Clinical features, ventilatory management, and outcome of ARDS caused by COVID-19 are similar to other causes of ARDS. Intensive Care Med
. Published online July 31, 2020. doi:10.1007/s00134-020-06192-2PubMedGoogle Scholar
et al; SAPS 3 Investigators. SAPS 3—from evaluation of the patient to evaluation of the intensive care unit, II: development of a prognostic model for hospital mortality at ICU admission. Intensive Care Med
. 2005;31(10):1345-1355. doi:10.1007/s00134-005-2763-5PubMedGoogle ScholarCrossref
et al; SAPS 3 Investigators. SAPS 3—from evaluation of the patient to evaluation of the intensive care unit, I: objectives, methods and cohort description. Intensive Care Med
. 2005;31(10):1336-1344. doi:10.1007/s00134-005-2762-6PubMedGoogle ScholarCrossref
et al; COVID-19 Lombardy ICU Network. Risk factors associated with mortality among patients with COVID-19 in intensive care units in Lombardy, Italy. JAMA Intern Med
. Published online July 2020. doi:10.1001/jamainternmed.2020.3539PubMedGoogle Scholar