The study compared four treatment groups using analysis of variance or Kruskal–Wallis tests for continuous variables and chi-square or Fisher exact tests for categorical variables. Multivariable Cox regression estimated differences in mortality while controlling for demographics, preexisting conditions, and disease severity, and Kaplan–Meier curves summarized survival. Investigators also calculated propensity scores from the Cox-model patient characteristics and performed 1:1 matching between patients who received hydroxychloroquine, alone or with azithromycin, and patients who did not receive hydroxychloroquine. Matched-group status was then evaluated as a mortality predictor in a separate Cox model, with Kaplan–Meier curves comparing the matched groups.