Corticosteroids on the Management of Coronavirus Disease 2019 (COVID-19): A Systemic Review and Meta-Analysis

Background: We aimed to examine the available evidence regarding the efficacy and safety of corticosteroids on the management of coronavirus disease 2019 (COVID-19), severe acute respiratory syndrome (SARS-CoV) and Middle East respiratory syndrome (MERS-CoV). Method: An extensive search was conducted in Medline, Embase, and Central databases until the end of March 2020, using keywords related to corticosteroids, COVID-19, SARS-CoV and MERS-CoV. The main outcome was considered to be the mortality rate, length of stay, virus clearance time, symptom improvement, and lung function improvement. The findings are presented as odds ratio (OR) with 95% confidence interval (95% CI). Results: Fifteen paper compromising 5 studies on COVID-19, 8 studies on SARS-CoV and 2 studies on MERS-CoV were included. One study was clinical trial and the rest were cohort. The analyses showed that corticosteroids were not reduce the mortality rate of COVID-19 (OR=1.08; 95% CI: 0.34 to 3.50) and SARS-CoV (OR=0.77; 95% CI: 0.34 to 1.3) patients, while they were associated with higher mortality rate of patients with MERS-CoV (OR = 2.52; 95% CI: 1.41 to 4.50). Moreover, it appears that corticosteroids administration would not be effective in shortening viral clearance time, length of hospitalization, and duration of relief symptoms following viral severe acute respiratory infections. Conclusion: There is no evidences that corticosteroids are safe and effective on the treatment of severe acute respiratory infection when COVID-19 disease is suspected. Therefore, corticosteroids prescription in COVID-19 patients should be avoided.


Introduction
Coronavirus disease 2019 (COVID-19) is a global pandemic, starting since the December of 2019 and spreading to all parts of the world, except the Antarctica. The number of affected patients is significantly increasing ever since, and its mortality rate varies in different regions of the world. The mortality rate was 6% compared to active cases, and 21% compared to closed cases (1). Current treatments for COVID-19 are supportive and symptomatic including the use of antivirals agents, antibiotics, intravenous interferons and gamma globulins, invasive and non-invasive oxygen therapy, and corticosteroids (2,3). Corticosteroids are widely used as a therapeutic option in COVID-19 and two previous epidemic of coronavirus related severe acute respiratory infection, severe acute respiratory syndrome (SARS-CoV) and Middle East respiratory syndrome (MERS-CoV) (4,5). Recently, WHO has prohibited corticosteroids administration as a routine treatment for COVID-19 patients (6). However, corticosteroids may decrease mortality rate in COVID-19 patients, casting doubt over WHO recommendation (7). As a result, a thorough consensus is yet to be achieved in order to provide adequate evidence to determine whether corticosteroids administration is beneficial in the management of COVID-19 patients or not. Hence, we aimed to answer the important question: Do corticosteroids have any beneficial effects in the treatment of severe acute respiratory infection when COVID-19 disease is suspected?

Study design
PICO was determined as follows: problem (P): COVID-19, SARS-CoV, and MERS-CoV patients; intervention (I): corticosteroids therapy; comparison (C): compared with noncorticosteroids treated patients; and outcome (O): mortality, length of stay, virus clearance time, symptom improvement and lung function improvement.

Eligibility criteria
All clinical trial and observational studies on assessment of corticosteroids therapy in COVID-19, SARS-CoV, and MERS-CoV patients were included. Exclusion criteria were lack of placebo or control group (non-corticosteroids treated patients), non-coronaviruses related disease, nonviral infection, duplicate reports, and review articles.

Search strategy
An extensive search was conducted on Medline, Embase, and Central databases, from the inception of databases until the end of March 2020. The search was performed using keywords related to corticosteroids, COVID-19, SARS-CoV, and MERS-CoV. A manual search was performed on Google Scholar, Google, and preprinted articles databases.

Data extraction
Four independent researchers screened the titles and abstracts. Next, potentially relevant studies were identified, and by assessing the full texts, related articles were included. Then, each researcher reviewed and summarized the articles, independently. The data extracted from the articles included first author name, year of publication, country, study type, sample size, name of administered corticosteroid, dosage and route of administration, duration of treatment and outcomes. Any disagreement was resolved by discussing with a third researcher.

Outcome
The main outcome of the present study was be the mortality rate. Secondary outcomes included length of stay, virus clearance time, symptom improvement, and lung function improvement.

Risk of bias assessment
The risk of bias assessment in the present study was performed based on two guidelines; the clin-ical trials were assessed using Cochrane's risk of bias tools, (8) and for the observational studies, the National Heart Lung and Blood Institute Quality Assessment Tool for Observational Studies was adopted (9). Two researchers reviewed the articles independently, and assigned one of the low risk, high risk, and unclear risk scores for each item of the instructions. Any disagreement was again resolved using a third researcher's opinion.

Quality of evidence
The GRADE approach was used to evaluate the quality of evidence and strength of recommendations (10). In this section, two researchers independently assessed the papers, and at the end, a third researcher resolved any disagreements.

Statistical analysis
All analyses were performed using STATA 14.0 statistical software. The analysis could only be performed on the mortality data. In this section, the analyses were stratified according to the type of the severe acute respiratory infection, meaning the efficacies of corticosteroids treatment on mortalities following COVID-19, SARS-CoV, and MERS-CoV were evaluated separately. Odds ratio (OR) with 95% confidence interval (95% CI) was included to perform meta-analysis. Some of the included studies reported a hazard ratio, but since the mortality rate was less than 10% in the included studies, the odds ratio and the hazard ratio were very close to each other and were able to be pooled. Heterogeneity between the studies was evaluated using I 2 test (I 2 greater than 50% indicated heterogeneity), and since there existed a heterogeneity, random effect model was used for the analyses. Furthermore, Egger's test was used to evaluate the publication bias. P < 0.05 was considered as the significance level in all of the analyses.

Risk of bias
Risk of bias assessment of the only included clinical trial reveals that allocation concealment, blinding of participants and personnel and selective reporting in the trial have unclear risk of bias. Furthermore, considering the virus clearance time as the main outcome and not assessment of mortality, length of hospital stay or lung function, resulted in the other bias of this study to be considered high risk. Risk of bias assessments of the observational studies reveals that all of the studies have high risk of bias in sample size justification and blinding of outcome assessor. Besides, only two studies assessed the effects of different doses of corticosteroids in the treatment of SARS-CoV infection (low risk). Other studies are consequently considered as high risk (Table 2).    Were inclusion and exclusion criteria for being in the study prespecified and applied uniformly to all participants?; 5. Was a sample size justification, power description, or variance and effect estimates provided?; 6. For the analyses in this paper, were the exposure(s) of interest measured prior to the outcome(s) being measured?; 7. Was the timeframe sufficient so that one could reasonably expect to see an association between exposure and outcome if it existed?; 8. For exposures that can vary in amount or level, did the study examine different levels of the exposure as related to the outcome (e.g., categories of exposure, or exposure measured as continuous variable)?; 9. Were the exposure measures (independent variables) clearly defined, valid, reliable, and implemented consistently across all study participants?; 10. Was the exposure(s) assessed more than once over time?; 11. Were the outcome measures (dependent variables) clearly defined, valid, reliable, and implemented consistently across all study participants?; 12. Were the outcome assessors blinded to the exposure status of participants?; 13. Was loss to follow-up after baseline 20% or less?; 14. Were key potential confounding variables measured and adjusted statistically for their impact on the relationship between exposure(s) and outcome(s)? Items for Cochrane risk of bias tools: 1. Random sequence generation. 2. Allocation concealment; 3. Blinding of participant and personnel; 4. Blinding of outcome assessor; 5. Incomplete outcome data; 6. Selective reporting; 7. Other bias

Quality of Evidence
According the GRADE guideline, the certainly of evidence derived from observational studies is low. We downgraded the level of evidence from low (observational data) to very low due to high risk of indication bias and low sample size of non-corticosteroids group. Our judgment result-ed in that sicker patients were more likely to receive corticosteroids than others. In addition, there was a substantial inconsistency among studies and in secondary outcomes. Therefore, overall certainty of evidence in all assessed outcomes is very low (Table 3).

The effects of corticosteroids administration on the outcomes of respiratory diseases caused by the coronaviruses COVID-19 related mortality
Five observational studies (7,16,20,22,23) had assessed the effects of corticosteroids administration on mortality of the COVID-19 patients (data from 430 patients). The analysis showed that administration of corticosteroids had no beneficial effect in reducing mortality following COVID-19 (OR = 1.08; 95% CI: 0.34 to 3.50; I 2 = 79.4%; P=0.001) (Fig. 2). No publication bias was observed in this section (P= 0.828) (Fig. 3).

Secondary outcomes in COVID-19
In addition to mortality rate, the efficacy of corticosteroids treatment on SpO2 improvement, need for oxygen therapy, duration of symptoms, and length of hospital stay were evaluated in two studies. First, corticosteroids therapy would shorten the need for oxygen therapy and accelerate SpO2 improvement (20). However, in another study, virus clearance time (HR = 1.26, 95% CI: 0.58-2.74.55), length of hospital stay (HR = 0.77, 95% CI: 0.33-1.78) and duration of symptoms (HR = 0.86, 95% CI: 0.40-1.83) would not be affected by corticosteroids administration (22).

Secondary outcomes in SARS-CoV
Length of hospital stay was another investigated outcome in the included studies. In the first study, Chen et al evaluated 401 patients and reported that corticosteroids administration would shorten the length of hospital stay (13). However, different doses of corticosteroids did not change the length of hospital stay (17). Finally, Wang et al. observed no relationships between corticosteroids administration and length of hospital stay (19). The disease complications are not related to corticosteroids administration (13). However, corticosteroids alleviated the disease symptoms and improved lung function (14).
In the only included clinical trial, Lee et al performed a double-blinded trial on 16 patients and reported that the virus clearance time, which was directly related to the length of hospital stay, was rather increased when using corticosteroids (15).

Secondary outcomes in SARS-CoV
The only evaluated secondary outcome in MERS-CoV section was virus clearance time. Corticosteroids therapy caused a delay in the clearance time in MERS-CoV infection (4).

Discussion
Current evidence showed that corticosteroids administration do not have beneficial effects in decreasing mortality rate following COVID-19 and SARS-CoV, while these drugs increase the risk of mortality in MERS-CoV patients. In addition, there were significant discrepancy among studies in evaluating the efficacy of corticosteroids in shortening the length of hospital stay, duration of symptom resolution and viral clearance time. Therefore, no evidences exist regarding safety and efficacy of corticosteroids on the treatment of respiratory infection caused by coronaviruses. These results confirmed the WHO conclusion that emphasis corticosteroids should not be used as a routine treatment for COVID-19 patients (6).
Only one double-blinded clinical trial was included in the current review, which did not have the good quality, when its risk of bias was assessed.
The study did not provide details of allocation concealment, blinding of participants and personnel. Also, considering virus clearance time as the main outcome and not paying attention to the mortality rate, length of hospital stay, or lung function resulted in the study have a high risk of bias in other bias item. All of the other included studies were cohort. Quality of evidence assessments for these studies showed that serious limitations exist regarding their research methodology. Hence, the findings of the included studies lie within the "very low level of evidence" range.  (27). Also, Zhou et al performed a meta-analysis to assess the effects of corticosteroids administration on the outcomes of influenza-related ARDS, reporting similar findings, and stating that prescription of these drugs not only increases the risk of mortality in influenza-related ARDS, but also increases the risk of acquiring secondary nosocomial infections (28). Although other meta-analyses exist, showing that corticosteroids can decrease mortality rate in ARDS patients (29) but the ARDS population included in these studies were mixed-population (all cause ARDS including trauma, contusion, post-surgery, bacterial and viral), preventing their results to be generalizable to viral respiratory infections. Finally, only one study was conducted on COVID-19, which performed a bivariate cox regression model, depicting that corticosteroids administration can decrease the mortality rate of ARDS patients following COVID-19. However, the analyses of this study were not adjusted for the potential confounders (7). This is a serious limitation, and has made the results doubtful.

Conclusion
Corticosteroids administration does not decrease the risk of mortality following COVID-19 and SARS-CoV, while it increases the mortality risk in patients with MERS-CoV. In addition, significant disagreement exists among the studies regarding the efficacy of corticosteroids in shortening the length of hospital stay, duration of symptom resolution and viral clearance time. In general, there is no evidence that corticosteroids are safe and effective in the treatment of respiratory infections caused by coronaviruses. Therefore, corticosteroids prescription in COVID-19 patients should be avoided, unless there are other indication.

Ethical considerations
Ethical issues (Including plagiarism, informed consent, misconduct, data fabrication and/or falsification, double publication and/or submission, redundancy, etc.) have been completely observed by the authors.