Objective To estimate the comparative effectiveness of combination therapy with hydroxychloroquine (HCQ) and azithromycin for coronavirus disease 2019 (COVID-19)- related death based on a large monocentric cohort independent of investigators’ putative biases in a real-world setting.
Design Retrospective monocentric cohort study, with comprehensive data collection authenticated by an external bailiff and death reports from a national database (French National Death Registry).
Setting Institut Hospitalo-Universitaire Méditerranée Infection Center in Marseille, France.
Participants All adults older than 18 years with PCR-proven COVID-19 who were treated directly in our centre between 2 March 2020 and 31 December 2021 and did not refuse the use of their data.
Interventions HCQ and azithromycin (HCQ-AZ) as a reference treatment were compared to other regimens containing HCQ, ivermectin and azithromycin alone, combined, or none of these three drugs. The effect of vaccination was also evaluated.
Main outcome measures 6-week all-cause mortality. Multivariable logistic regression estimated treatment effectiveness with adjustments for age, sex, comorbidities, vaccination, period of infection or virus variant, and outpatient or inpatient care.
Results Total 30,423 COVID-19 patients were analysed (86 refused the analysis of their data) including 30,202 with available treatment data, and 535 died (1.77%). All-cause mortality was very low among patients < 50 years (8/15,925 (0.05%)) and among outpatients treated with HCQ-AZ (21 deaths out of 21,135 (0.1%), never exceeding 0.2% regardless of epidemic period). HCQ-AZ treatment was associated with a significantly lower mortality rate than no HCQ-AZ after adjustment for sex, age, period and patient care setting (adjusted OR (aOR) 95% confidence interval (CI) 0.55, 0.45-0.68). The effect was greater among outpatients (71% death protection rate) than among inpatients (45%). In a subset of 16,063 patients with available comorbidities and vaccinations status, obesity (2.01, 1.23-3.29), chronic respiratory disease (2.93, 1.29-6.64), and immunodeficiency (4.01, 1.69-9.50), on the one hand, and vaccination (0.29, 0.12-0.67) and HCQ-AZ treatment (0.47, 0.29-0.76), on the other hand, were independent factors associated with mortality. HCQ, alone or in any association, was associated with significant protection from death among outpatients (0.41, 0.21-0.79) and inpatients (0.59, 0.47-0.73).
Conclusions HCQ prescribed early or late protects in part from COVID-19-related death. During pandemic health crises, financial stakes are enormous. Authentication of the data by an independent external judicial officer should be required. Public sharing of anonymized databases, ensuring their verifiability, should be mandatory in this context to avoid fake publications.