This retrospective, nonrandomized observational study evaluated hospitalized COVID-19 patients treated with hydroxychloroquine alone, hydroxychloroquine plus azithromycin, azithromycin alone, or neither drug. Under the health system's protocol-based treatment algorithm, hydroxychloroquine alone and hydroxychloroquine plus azithromycin were associated with lower in-hospital mortality than receiving neither drug. Because treatment was not randomly assigned, the reported association does not by itself establish a causal treatment effect.