Steroids use in non-oxygen requiring COVID-19 patients: a systematic review and meta-analysis

Summary Background Corticosteroids have become the mainstay treatment in severe COVID-19. However, its role is mild disease is controversial due to lack of robust scientific evidence. This systematic review and meta-analysis were conducted to assess effect of steroids in mild COVID-19 patients. Methods PubMed, EMBASE, Web of Science and medRxiv were searched from 31 December 2019 to 14 May 2021 for studies that reported effectiveness of steroids in non-oxygen requiring COVID-19 patients in terms of progressing to severe disease, mortality, duration of fever, duration of viral clearance and length of hospital stay (LOHS). Studies on inhalational steroids, case reports and reviews were excluded. Risk of bias (ROB) was assessed by the Cochrane's ROB tool and ROBANS tool. Quantitative data synthesis was done using the generic inverse variance method. Results A total of 6411 studies were identified, 2990 articles were screened after exclusion. Seven studies which fit the criteria (involving 2214 non-oxygen requiring COVID-19 patients) were included and analysed. Overall odds of progression to severe disease among the non-oxygen requiring COVID-19 patients receiving steroids was 5.97 [95% confidence interval (CI): 1.27–27.99, I2 = 0%] and odds of death (OR: 1.35, 95% CI: 1.01–1.79; I2 = 0%) as compared to the patients not receiving steroids. Mean duration of fever (7.4 days), duration to viral clearance (18.9 days) and LOHS (20.8 days) were significantly higher in the steroid arm, as compared to that in no-steroid arm (6.7, 16.5 and 15.2 days, respectively). Conclusion Steroids in non-oxygen requiring COVID-19 patients can be more detrimental than beneficial. Protocol registration The study was prospectively registered in PROSPERO (CRD 42021254951).