Outpatients prescribed with fluvoxamine around the time of COVID-19 diagnosis are not at a reduced risk of subsequent hospitalization and death compared to their non-prescribed peers: population-based matched cohort study.

ABSTRACT

PURPOSE: To assess the effect of exposure to fluvoxamine around the COVID-19 diagnosis on subsequent hospitalizations and mortality in COVID-19 outpatients in a real-life setting. METHODS: Using nationwide administrative data, we identified adult COVID-19 outpatients diagnosed up to August 15, 2021 and conducted two cohort studies. Study 1 included subjects prescribed fluvoxamine around the index COVID-19 diagnosis (Cohort A), their peers suffering similar psychiatric difficulties but not prescribed fluvoxamine (Cohort B) and those free of psychiatric difficulties/treatments (Cohort C). Study 2 included subjects prescribed fluvoxamine (Cohort Fluvoxamine) and their peers prescribed paroxetine (Cohort Paroxetine). Cohorts were mutually exactly matched and incidence of COVID-19-related hospitalization, 30-day all-cause hospitalization and of COVID-19-related mortality was estimated. RESULTS: Of the 416,030 first-episode outpatients, Study 1 included 1016 Cohort A, 95,984 Cohort B and 275,804 Cohort C patients. Matched Cohort A (n = 749) vs. Cohort B (n = 31,336) relative risks (95%CI/CrI), frequentist and Bayes with skeptical, otpimistic and pesimistic priors, were COVID-related hospitalization 1.37 (0.56-3.33), 1.15 (0.55-2.11), 1.03 (0.56.1.96) and 1.43 (0.63-2.94), respectively; 30-day all-cause hospitalization 1.88 (0.76-4.67), 1.76 (1.39-2.25), 1.76 (1.39-2.24) and 1.86 (1.43-2.38), respectively; COVID-19-related mortality 0.73 (0.35-1.55), 0.93 (0.53-1.76), 0.79 (0.40-1.54) and 0.88 (0.37-2.11), respectively. Matched Cohort A vs. C (866 vs. 222,792) comparison yielded similar estimates, as did the matched Cohort Fluvoxamine vs. Paroxetine comparison in Study 2 (344 of 994 matched to 535 of 1796 patients). CONSLUSION Outpatients prescribed fluvoxamine around the time of COVID-19 diagnosis were not at a reduced risk of hospitalizations and mortality compared to their non-prescribed peers.