Steroids use in Severe/ Life Threatening COVID-19 Pneumonia: Does the Type or Dose Matter?


  • Abdul Rehman Arshad
  • Bilal Saeed
  • Sagheer Hussain



coronavirus, glucocorticoids, outcome, Pakistan, pandemic, survival


Objective: To compare the effect of different doses of methylprednisolone and dexamethasone on in-hospital mortality in severe COVID-19 pneumonia Methods: This retrospective chart review was done by reviewing old medical reports of patients with severe disease admitted to COVID-19 Intensive Care and High Dependency Unit from October 2020 to September 2021. Those with suspected COVID-19 infection (suggestive radiological findings but negative PCR for SARSCoV- 2 on at least two occasions) were excluded. Patients requiring high flow oxygen (>6 liters per minute) or higher levels of respiratory support were classified as having severe disease. We recorded the type of steroids used and the doses. Methylprednisolone in doses up to 40mg per day, or other steroids in equivalent doses, were considered low dose. Primary outcome of interest was in- hospital mortality. Results: There were 279 patients aged 52.53± 11.31 years, including 216 (77.42%) males. Mean hospital stay was 10.18± 3.13 days. During hospital stay, 96 (34.41%) patients died. Amongst patients receiving dexamethasone, 70 (44.87%) expired, whereas 26 (21.14%) out of 123 patients who received methylprednisolone expired (p<0.001; hazard ratio 3.037). With high dose steroids, 52 (41.27%) out of 126 patients expired, whereas 44 (28.76%) out of 153 patients treated with low dose steroids expired (p=0.029; hazard ratio 1.741). In multivariate binary logistic regression, in-hospital mortality was related to the type of steroid but not the steroid dose. Conclusion: Methylprednisolone is superior to dexamethasone for treatment of severe COVID-19 pneumonia.



How to Cite

Arshad, A. R. ., Saeed, B. ., & Hussain, S. . (2023). Steroids use in Severe/ Life Threatening COVID-19 Pneumonia: Does the Type or Dose Matter?. Annals of King Edward Medical University, 28(4), 423-427.