Relation

Statistical analysis for evaluating 28-day mortality and other clinical outcomes in a randomized controlled trial of Dexamethasone (DEX) among hospitalized COVID-19 patients in RECOVERY trial

  • All rate and risk ratios given 95% confidence intervals

  • All p-values 2-sided

  • Intention-to-treat assumption

  • 28-day mortality: evaluated with Hazard ratios (HR) from Cox regression (see 28-day mortality methods node for details) and tracked with a Kaplan-Meier curve

  • Time to discharge: modeled with Cox regression; those that died during hospitalization were right censored

  • Invasive mechanical ventilation or death (composite): No date information, so log-binomial regression used to estimate risk ratio

0

1

Updated 2020-06-26

Tags

SARS-CoV-2 (COVID-19)

Biomedical Sciences