key: cord-0796201-nkfx1gyo authors: Munch, Marie Warrer; Meyhoff, Tine Sylvest; Helleberg, Marie; Kjær, Maj‐Brit Nørregaard; Granholm, Anders; Hjortsø, Carl Johan Steensen; Jensen, Thomas Steen; Møller, Morten Hylander; Hjortrup, Peter Buhl; Wetterslev, Mik; Vesterlund, Gitte Kingo; Russell, Lene; Jørgensen, Vibeke Lind; Tjelle Kristiansen, Klaus; Benfield, Thomas; Ulrik, Charlotte Suppli; Andreasen, Anne Sofie; Bestle, Morten Heiberg; Poulsen, Lone Musaeus; Hildebrandt, Thomas; Knudsen, Lene Surland; Møller, Anders; Sølling, Christoffer Grant; Brøchner, Anne Craveiro; Rasmussen, Bodil Steen; Nielsen, Henrik; Christensen, Steffen; Strøm, Thomas; Cronhjort, Maria; Wahlin, Rebecka Rubenson; Jakob, Stephan M.; Cioccari, Luca; Venkatesh, Balasubramanian; Hammond, Naomi; Jha, Vivekanand; Myatra, Sheila Nainan; Jensen, Marie Qvist; Leistner, Jens Wolfgang; Mikkelsen, Vibe Sommer; Svenningsen, Jens S.; Laursen, Signe Bjørn; Hatley, Emma Victoria; Kristensen, Camilla Meno; Al‐Alak, Ali; Clapp, Esben; Jonassen, Trine Bak; Bjerregaard, Caroline Løkke; Østerby, Niels Christian Haubjerg; Jespersen, Mette Mindedahl; Abou‐Kassem, Dalia; Lassen, Mathilde Languille; Zaabalawi, Reem; Daoud, Mohammed Mahmoud; Abdi, Suhayb; Meier, Nick; la Cour, Kirstine; Derby, Cecilie Bauer; Damlund, Birka Ravnholt; Laigaard, Jens; Andersen, Lene Lund; Mikkelsen, Johan; Jensen, Jeppe Lundholm Stadarfeld; Rasmussen, Anders Hørby; Arnerlöv, Emil; Lykke, Mathilde; Holst‐Hansen, Mikkel Zacharias Bystrup; Tøstesen, Boris Wied; Schwab, Janne; Madsen, Emilie Kabel; Gluud, Christian; Lange, Theis; Perner, Anders title: Low‐dose hydrocortisone in patients with COVID‐19 and severe hypoxia: the COVID STEROID randomised, placebo‐controlled trial date: 2021-06-17 journal: Acta Anaesthesiol Scand DOI: 10.1111/aas.13941 sha: a85ef02e68d3f912bda3aeaa339657265f2546c3 doc_id: 796201 cord_uid: nkfx1gyo BACKGROUND: In the early phase of the pandemic, some guidelines recommended the use of corticosteroids for critically ill patients with COVID‐19, whereas others recommended against the use despite lack of firm evidence of either benefit or harm. In the COVID STEROID trial, we aimed to assess the effects of low‐dose hydrocortisone on patient‐centred outcomes in adults with COVID‐19 and severe hypoxia. METHODS: In this multicentre, parallel‐group, placebo‐controlled, blinded, centrally randomised, stratified clinical trial, we randomly assigned adults with confirmed COVID‐19 and severe hypoxia (use of mechanical ventilation or supplementary oxygen with a flow of at least 10 L/min) to either hydrocortisone (200 mg/day) versus a matching placebo for 7 days or until hospital discharge. The primary outcome was the number of days alive without life support at day 28 after randomisation. RESULTS: The trial was terminated early when 30 out of 1,000 participants had been enrolled because of external evidence indicating benefit from corticosteroids in severe COVID‐19. At day 28, the median number of days alive without life support in the hydrocortisone versus placebo group were 7 versus 10 (adjusted mean difference: ‐1.1 days, 95% CI ‐9.5 to 7.3, p = 0.79); mortality was 6/16 versus 2/14; and the number of serious adverse reactions 1/16 versus 0/14. CONCLUSIONS: In this trial of adults with COVID‐19 and severe hypoxia, we were unable to provide precise estimates of the benefits and harms of hydrocortisone as compared with placebo as only 3% of the planned sample size were enrolled. In the early phase of the pandemic, some guidelines recommended the use of corticosteroids for critically ill patients with COVID-19, whereas others recommended against the use despite lack of firm evidence of either benefit or harm. In the COVID STEROID trial, we aimed to assess the effects of low-dose hydrocortisone on patient-centred outcomes in adults with COVID-19 and severe hypoxia. In this multicentre, parallel-group, placebo-controlled, blinded, centrally randomised, stratified clinical trial, we randomly assigned adults with confirmed COVID-19 and severe hypoxia (use of mechanical ventilation or supplementary oxygen with a flow of at least 10 L/min) to either hydrocortisone (200 mg/day) versus a matching placebo for 7 days or until hospital discharge. The primary outcome was the number of days alive without life support at day 28 after randomisation. The trial was terminated early when 30 out of 1,000 participants had been enrolled because of external evidence indicating benefit from corticosteroids in severe COVID-19. At day 28, the median number of days alive without life support in the hydrocortisone versus placebo group were 7 versus 10 (adjusted mean difference: -1.1 days, 95% CI -9.5 to 7.3, p = 0.79); mortality was 6/16 versus 2/14; and the number of serious adverse reactions 1/16 versus 0/14. In this trial of adults with COVID-19 and severe hypoxia, we were unable to provide precise estimates of the benefits and harms of hydrocortisone as compared with placebo as only 3% of the planned sample size were enrolled. The novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) 1 may cause coronavirus disease 2019 (COVID-19); 2 the clinical presentation of which varies from mild upper respiratory infection to severe acute respiratory distress syndrome (ARDS). 3 Systemic corticosteroids have been reported to benefit some patients with ARDS and septic shock. [4] [5] [6] In the early phases of the COVID-19 pandemic, there were conflicting guidelines on the use of corticosteroids for COVID-19; 7-10 the Surviving Sepsis Campaign advocated its use in critically ill patients with COVID-19 9 whilst the World Health Organization (WHO) initially recommended against it. 7 To address this uncertainty, several clinical trials were initiated. [11] [12] [13] [14] [15] [16] [17] [18] [19] On April 15, 2020, we commenced the 'Low-dose hydrocortisone in patients with COVID-19 and severe hypoxia (COVID STEROID) trial' aiming to assess the effects of low-dose hydrocortisone versus placebo on patient-centred outcomes. We hypothesised that low-dose hydrocortisone would increase the number of days alive without life support as compared with placebo. On June 16, 2020, enrolment was paused when 30 participants had been randomised because of a press release from the Randomised Evaluation of COVid-19 thERapY (RECOVERY) trial reporting benefit from systemic corticosteroids on 28-day mortality in hospitalised patients with COVID-19. 20 The Management Committee of the COVID STEROID trial decided to terminate the trial on September 4, 2020, due to the results from a WHO-initiated prospective meta-analysis (PMA) of ongoing or recently completed trials demonstrating benefit from systemic corticosteroids on 28-day mortality in critically ill patients with COVID-19, 21 which in turn led to an update in the guideline from the WHO strongly recommending the use of systemic corticosteroids for patients with severe or critical COVID-19. 22 We provided data on 28-day mortality and serious adverse reactions (SARs) for 29 out of 30 trial participants to the PMA published in September 2020, 21 because of the importance of summarising the evidence from all trials assessing corticosteroids in patients with critical COVID-19. Here, we present the full trial report of the COVID STEROID trial, including 90-day follow-up. This article is protected by copyright. All rights reserved The COVID STEROID trial was an investigator-initiated, multicentre, parallel-group, placebo-controlled, blinded, centrally randomised, stratified clinical trial. We planned to randomise 1,000 adult patients with COVID-19 and severe hypoxia in Denmark, Sweden, Switzerland and India. The trial was commenced on April 15, 2020; paused on June 16, 2020; and terminated early on September 4, 2020, after 30 patients had been enrolled at 12 trial sites in Denmark. The COVID STEROID trial was conducted in accordance with the published trial protocol, 23 Patients fulfilling all inclusion criteria and no exclusion criteria were randomised in a 1:1 ratio to 200 mg of intravenous (IV) hydrocortisone versus placebo using a centralised and web-based randomisation system at the Copenhagen Trial Unit (CTU). The randomisation was performed according to a computer-generated allocation sequence, stratification variables (trial site, use of invasive mechanical ventilation (y/n), age below 70 years (y/n)), and varying block sizes. The allocation sequence was only known by the data manager at the CTU. This article is protected by copyright. All rights reserved The Management Committee, investigators, trial site staff registering outcome data, trial statistician, clinical staff, relatives, and participants were all blinded to the allocation. Trial medication was prepared daily using shelf-medication by a dedicated team of unblinded trial site staff (medical students and/or research nurses and doctors). The unblinded trial site staff were not involved in the care of trial participants, outcome data entry, or statistical analyses. For the analyses conducted in the present report, the trial statistician received a blinded dataset with the allocation coded as 0 or 1. Of note, due to the previous publication of data on 28-day mortality and SARs, it was possible to identify the groups 0 and 1 by referral to the PMA. 21 We screened adult patients (18 years Trial participants received either IV hydrocortisone (200 mg per day) or a matching placebo (isotonic saline) for 7 days or until hospital discharge (whichever came first). The injections had identical appearances. For patients allocated to hydrocortisone, the trial intervention was given either as a continuous infusion over 24 hours or as bolus injections every 6 hours (50 mg per bolus). For patients allocated to placebo, the trial intervention was given either as a continuous infusion over 24 hours or as bolus injections every 6 hours. This article is protected by copyright. All rights reserved All other interventions were given at the discretion of the treating clinicians. The procedure for trial medication preparation was described in detail in the published trial protocol. 23 Detailed definitions of primary and secondary outcomes are presented in the published trial protocol. 23 Primary outcome The primary outcome was days alive without the use of life support (i.e. invasive mechanical ventilation, circulatory support, or renal replacement therapy, including days in between intermittent renal replacement therapy) at day 28. The secondary outcomes were: • Number of participants with one or more serious adverse reactions at day 14 defined as new episodes of septic shock, invasive fungal infection, clinically important gastrointestinal bleeding, or anaphylactic reaction. • All-cause mortality at day 28. • Days alive without life support at day 90. • Days alive and out of hospital at day 90. • All-cause mortality at day 90. We estimated that 1,000 participants would be required to have 85% power to detect a 15% relative reduction in 28-day mortality combined with a 10% reduction in the days on life support among survivors. Baseline data were presented descriptively in each group and in total with medians with interquartile ranges (IQRs) for continuous variables and numbers with percentages for categorical variables. Because the final sample size was substantially smaller than planned, deviations from the pre-published statistical Accepted Article analysis plan were necessary (details are presented in Supplementary Information S3) . For continuous outcomes, results are presented as medians with IQRs in each group and adjusted mean differences with 95% confidence intervals (CIs). For binary outcomes, results are presented as numbers with percentages in each group and adjusted relative risks (RR) with 95% CIs as described previously. 23 All analyses were done in the intention-to-treat (ITT) population defined as all randomised participants for whom there were consent to use data. We had no missing outcome data, and only missing data for baseline lactate in one participant; accordingly, neither multiple imputation nor best-worst/worst-best case analyses were conducted. 23 Primary outcome The primary outcome was evaluated using a linear regression adjusted for invasive mechanical ventilation and age below 70 years. We were unable to adjust for site due to the reduced sample size. The primary analysis was supplemented by a Wilcoxon rank sum test. Descriptive data on the primary outcome are presented in 5 out of 6 pre-planned subgroups (Table 4) . For all secondary binary outcomes, we employed unadjusted generalised linear models with log links and binomial error distributions (except for SARs due to only 1 event). These were supplemented by Fisher's exact tests. The secondary continuous outcomes were evaluated as the primary outcome using linear regressions adjusted for invasive mechanical ventilation and age below 70 years and were supplemented by Wilcoxon rank sum tests. We conducted a post hoc fully anonymised survey to assess how clinical practice was influenced by the RECOVERY preprint. 20 We surveyed the clinical preferences for corticosteroid use for COVID-19 among doctors working at planned COVID STEROID trial sites in India, Sweden, Switzerland and Denmark. Moreover, we asked site investigators to submit aggregate data from participating hospitals on the use of corticosteroids in the last 10 COVID-19 patients admitted in July 2020 after the RECOVERY preprint had been published. The trial was initiated on April 15, 2020, and the first participant was enrolled on April 17, 2020. On June 16, 2020, enrolment was paused after a total of 30 participants had been randomised. Of these, 16 were allocated to hydrocortisone and 14 were allocated to placebo, and all completed follow-up of data at day 90 ( Figure 1 ). The baseline characteristics of the trial participants are presented in Table 1 and the trial medication administration and protocol violations in Table 2 . The trial participants received the trial intervention for a median of 7 days (IQR 6-7 days) in both groups. A total of 3 trial participants (2 allocated to hydrocortisone and 1 allocated to placebo) discontinued the trial intervention for the reasons presented in Figure 1 The outcome data for the 2 groups are presented in Table 3 . There was no statistically significant difference in the days alive without life support at day 28 after randomisation (the primary outcome) in patients allocated to hydrocortisone versus those allocated to placebo (adjusted mean difference: -1.1 days, 95% CI -9.5 to 7.3, p = 0.79). In total, there was 1 patient with one or more SARs in the trial (hydrocortisone: 1/16; placebo 0/14). There were no statistically significant differences in all-cause mortality at 28 and 90 days (28 days: RR 2.63, 95% CI 0.74 to 16.03, p = 0.19; 90 days: RR 2.04, 95% CI 0.71 to 8.16, p = 0.22), in the days alive without life support at day 90 (adjusted mean difference: -14.7 days, 95% CI -40.4 to 10.9, p = 0.25) or in the days alive and out of hospital at day 90 (adjusted mean difference: -6.5 days, 95% CI -29.6 to 16.7, p = 0.57). Some results of the survey have already been published elsewhere. 27 A total of 278 doctors from Denmark (53%), India (40%), Sweden (4%) and Switzerland (3%) responded to the survey on corticosteroid use in patients with COVID-19. The responders primarily worked in ICUs (56%) or departments of anaesthesiology (21%) or infectious diseases (15%). This article is protected by copyright. All rights reserved In this investigator-initiated, multicentre, parallel-group, blinded, centrally randomised, stratified clinical trial, we were unable to provide any precise estimates on the benefits and harms of hydrocortisone versus placebo for any outcomes as only 3% of the planned sample size was enrolled. In our post hoc survey, we found that the dose preferences for corticosteroid use in COVID-19 varied among doctors and that most would enrol their patients into a trial comparing higher versus lower doses of dexamethasone. The strengths include all those of a multicentre randomised, placebo-controlled trial. The trial design was pragmatic allowing all other interventions to follow standard care, and only patient-centred outcomes were reported. The trial was monitored externally by local GCP units in Denmark, and we had no missing values for any assessed outcome. The trial was reported in accordance with the CONSORT statement. 26 During the planning and commencement of the COVID STEROID trial, the number of COVID-19 patients in Denmark decreased leading to an unexpected low recruitment throughout the inclusion period. The recruitment was also lowered due to enrolment in other interventional trials prohibiting co-enrolment (16% of all excluded patients). The COVID STEROID trial was terminated early when only 3% of the planned sample size had been enrolled resulting in highly uncertain point estimates. 28 Yet, we consider it important to publish data on all outcomes out of respect for the trial participants and to allow for data to be included in future meta-analyses. The trial intervention was subject to protocol violations occurring in more than one-third of the participants, likely due to strain on the clinical staff. We planned to explore the effect of these protocol violations by performing a sensitivity analysis of the primary outcome in the per protocol population (i.e. participants without one or more protocol violations) but refrained from this due to the small sample size. Moreover, we were unable to adhere to the published statistical analysis plan due to the reduced sample size and few patients with events for the binary outcomes or the value zero for the continuous outcomes. Consequently, the results presented in this report are based on a post hoc decision on how to analyse the This article is protected by copyright. All rights reserved data in a sensible manner adhering to the protocol to the extent possible in accordance with the trial statistician. Several clinical questions remain unanswered regarding the use of corticosteroids in COVID-19, among others the timing of initiation, optimal dosing regime and duration of treatment. In our survey, we found that the dose preferences for corticosteroid use in COVID-19 varied among doctors and that most would enrol their patients into a trial comparing higher versus lower doses of dexamethasone. The doses used in clinical practice also varied between hospitals. Of note, the dose used was often higher than that used in the RECOVERY trial. Consequently, the higher vs. lower doses of dexamethasone in patients with COVID-19 and severe hypoxia (COVID STEROID 2) trial was commenced on August 27, 2020, to provide data on the effect of 12 mg vs. 6 mg of dexamethasone in a similar cohort of patients. 27, 29 In this trial, we also include patients using corticosteroids at the time of screening and patients who have limitations in the use of lifesupport to increase the generalisability of the results. In this early terminated randomised clinical trial of adult patients with COVID-19 and severe hypoxia, we were unable to provide any precise estimates on the benefits and harms of hydrocortisone versus placebo for any outcomes as only 3% of the planned sample size had been enrolled. We did learn important lessons that have informed the design and conduct of the COVID STEROID 2 trial. 27, 29 Coronaviridae Study Group of the International Committee on Taxonomy of V. The species Severe acute respiratory syndrome-related coronavirus: classifying 2019-nCoV and naming it SARS-CoV-2 Novel Coronavirus(2019-nCoV) COVID-19 diagnosis and management: a comprehensive review Low-dose corticosteroids for adult patients with septic shock: a systematic review with meta-analysis and trial sequential analysis Effect of different doses and time-courses of corticosteroid treatment in patients with acute respiratory distress syndrome: A meta-analysis Efficacy and safety of corticosteroids in COVID-19 based on evidence for COVID-19, other coronavirus infections, influenza, community-acquired pneumonia and acute respiratory distress syndrome: a systematic review and meta-analysis Clinical management of severe acute respiratory infection (SARI) when COVID-19 disease is suspected. Interim guidance COVID-19) Treatment Guidelines. National Institutes of Health Accessed 01 Surviving Sepsis Campaign: guidelines on the management of critically ill adults with Coronavirus Disease 2019 (COVID-19) Infectious Diseases Society of America Guidelines on the Treatment and Management of Patients with COVID-19 Multifactorial Adaptive Platform Trial for Community-Acquired Pneumonia (REMAP-CAP) (NCT02735707) Targeted Steroids for ARDS Due to COVID-19 Pneumonia: A Pilot Randomized Clinical Trial (NCT04360876) Steroid Dosing by bioMARker Guided Titration in Critically Ill Patients With Pneumonia (SMART) (NCT03852537) Glucocorticoid Therapy for COVID-19 Critically Ill Patients With Severe Acute Respiratory Failure (NCT04244591) Prophylactic Corticosteroid to Prevent COVID-19 Cytokine Storm (NCT04355247) Efficacy of Methylprednisolone for Patients With COVID-19 Severe Acute Respiratory Syndrome (MP-C19) (NCT04323592) Methylprednisolone in the Treatment of Patients With Signs of Severe Acute Respiratory Syndrome in Covid-19 (MetCOVID) (NCT04343729) Efficacy of Dexamethasone Treatment for Patients With ARDS Caused by COVID-19 (DEXA-COVID19) (NCT04325061) Evaluation of Corticosteroids (CAPE_COD) (NCT02517489). Available from Effect of Dexamethasone in Hospitalized Patients with COVID-19: Preliminary Report WHO Rapid Evidence Appraisal for COVID-19 Therapies Working Group Association Between Administration of Systemic Corticosteroids and Mortality Among Critically Ill Patients With COVID-19: A Meta-analysis World Health Organization. Corticosteroids for COVID-19. Living guidance Low-dose hydrocortisone in patients with COVID-19 and severe hypoxia (COVID STEROID) trial-Protocol and statistical analysis plan WMA Declaration of Helsinki -Ethical Principles for Medical Research involving Human Subjects statement: Updated guidelines for reporting parallel group randomised trials Higher vs. Lower Doses of Dexamethasone in Patients with Accepted Article This article is protected by copyright. All rights reserved COVID-19 and Severe Hypoxia (COVID STEROID 2) trial: protocol and statistical analysis plan Randomized trials stopped early for benefit: a systematic review Days alive without life support at day 28, median (IQR) IQR: interquartile range Computed using a linear regression adjusted for invasive mechanical ventilation (y/n) and age below 70 (y/n) ** Computed using Fisher's exact test ~Computed using an unadjusted generalised linear model with log link and binomial error distribution ≠ Computed using an unadjusted generalised linear model with log link and binomial error distribution Computed using a linear regression adjusted for invasive mechanical ventilation (y/n) and age below 70 (y/n) Computed using a linear regression adjusted for invasive mechanical ventilation (y/n) and age below 70 (y/n) This article is protected by copyright. All rights reserved Tables Oxygen supplementation This article is protected by copyright. All rights reserved This article is protected by copyright. All rights reserved Treatment duration in days, median (IQR) 7 (6-7) 7 (6-7)Mode of administration, no. (%)Continuous infusion 10 (63%) 10 (71%) Both used in the intervention period 2 (13%) 1 (7%) Participants who did not receive trial medication at all, no. This article is protected by copyright. All rights reserved This article is protected by copyright. All rights reserved We planned to conduct statistical analyses of subgroup differences on the primary outcome in the ITT population, but refrained from this due to the reduced sample size. Instead, descriptive data for each subgroup is reported. We were unable to report data for the subgroup analysis of geographical region as only Danish patients were included in the trial. Detailed definitions of each subgroup are provided in the published trial protocol.