key: cord-339444-f8lfob47 authors: Bigna, Jean Joel; Kamtchum-Tatuene, Joseph; Noubiap, Jean Jacques title: Claims about the safety and efficacy of early treatment of COVID-19 with hydroxychloroquine and azithromycin must be supported by real evidence date: 2020-09-25 journal: Travel Med Infect Dis DOI: 10.1016/j.tmaid.2020.101886 sha: doc_id: 339444 cord_uid: f8lfob47 nan Based on data from 1061 COVID-19 patients (94.7% symptomatic, 85.7% outpatients), the authors concluded that early treatment of COVID-19 (before complications occur) with the combination of hydroxychloroquine and azithromycin (HCQ+AZ) "is safe and associated with very low fatality rate" [1] . This claim is not supported by the data presented in their article and is, therefore, misleading. Indeed, in the absence of a control group of patients who did not receive HCQ+AZ, it is inappropriate to conclude that this therapeutic combination was associated with low fatality rate. This is a major concern, especially because spontaneous clearance of SARS-CoV-2 infection is more common in younger individuals who seem to represent the majority of the study cohort [2] . The authors do not provide a rationale for the decision not to consider a control group. This is a fundamental methodological point that would help to support the argument for a causal link between early HCQ+AZ therapy and good clinical outcomes. Furthermore, their safety data are also not generalizable because they are more likely to be driven by the specific demographic and clinical characteristics of the population. Indeed, with a mean age of only 43.6 years, the study population likely represents a selected group of individuals with low prevalence of comorbidities known to favor the occurrence of hydroxychloroquine-related cardiac adverse events. Moreover, the multivariable model used to explore factors associated with poor outcomes in this study presents some weaknesses. Although age, some co-medications, viral load at inclusion, and national early warning score for clinical severity were included in the multivariable model, it is surprising that important factors such as sex and chronic conditions were omitted. There is evidence supporting the association of male sex and several chronic conditions, including those reported in this study, with poor clinical outcomes [3] . This is more concerning in view of the significantly higher prevalence of chronic conditions in patients with poor clinical outcome in this study. A more effective handling of potential confounders could have been achieved through a multivariable model including all known clinically relevant prognostic factors or the use of propensity score matching. Finally, due to the retrospective nature of the study, it was not possible to consider blinding for outcome assessment. Except for death, clinical outcomes such as transfer to intensive care unit (ICU) or hospitalization for ten days may vary across physicians. For example, J o u r n a l P r e -p r o o f physicians may have a higher tendency to keep patients with chronic co-morbidities longer in hospital or to transfer them to ICU. It is worth noting that the discussion of the study limitations is focused on the explanation of missing data without highlighting the methodological concerns. In conclusion, given the many limitations raised above, it is advisable to be more careful about claims of association between HCQ+AZ and low fatality rates, especially when considering the latest evidence from a randomized controlled trial of HCQ as postexposure prophylaxis where HCQ did not prevent illness compatible with COVID-19 or confirmed infection [4] . In addition, recommendations for a universal test-and-treat strategy made by the authors should be tempered since this was not studied. Early treatment of COVID-19 patients with hydroxychloroquine and azithromycin: A retrospective analysis of 1061 cases in Marseille, France. Travel medicine and infectious disease Factors associated with hospital admission and critical illness among 5279 people with coronavirus disease Risk factors of critical & mortal COVID-19 cases: A systematic literature review and meta-analysis A Randomized Trial of Hydroxychloroquine as Postexposure Prophylaxis for Covid-19. The New England journal of medicine All authors contributed to writing this article and approved its final version. There is no funding source used for this paper. There is no competing interest.