Ethics and regulation:
This is a retrospective, observational cohort study. All patients diagnosed with COVID-19 by PCR identification of SARS-CoV-2 were invited to attend either the outpatient clinic or our hospital for evaluation and treatment. Data which emerged from epidemiological interviews and clinical and biological assessments were recorded in the hospital information system (HIS). For the purposes of this study, data from patients hospitalised between 3 March 2020 and 13 March 2021 were extracted from the HIS. To comply with the European General Data Protection Regulation, all patients were informed that their personal and medical data might be used for research purposes unless they refused. The investigators’ declaration to comply with reference method MR 004 was filed prior to the onset of this study and was the subject of a declaration in the GDPR/APHM Register No.2020-152.
Data collection: We first selected all patients hospitalised in our Institute within the one-year study period. To avoid bias or the use of inappropriate data, we excluded those who were immunocompromised (3), those mis-diagnosed as having COVID-19 (3), those treated with ivermectin alone (48), one minor patient (under the age of 18), 273 patients who were hospitalised for less than three days, 95 patients for whom treatment started more than four days after admission, and 17 patients for whom the timescale between admission and the first PCR test was more than 15 days (Figure 1 ). We then retained those for whom a positive PCR was obtained between D0-1 and D0+1. For patients who were treated with HCQ and/or AZT, “D0” was defined as the date treatment started (first treatment received). For patients who were not treated with HCQ or AZT, “D0” was defined as the date of admission to the IHU. Finally, we included those who had a second (positive or negative) PCR test between D0+1 and D0+10. All qPCR tests were performed in the same laboratory. When the CT of the qPCR was over 35 cycles, it was considered to be negative timescale between the onset (18). Explicative variables such as, age, initial viral load, date of onset of symptoms, treatment, and death were extracted from the HIS in compliance with the provisions of the GDPR. We identified four treatment groups: those who were treated with the HCQ regimen; those that did not receive HCQ; those treated with a combination of HCQ and AZ; and those treated only with AZ. We conducted an initial analysis of treatment with HCQ (with or without AZ) compared to treatment without HCQ (AZ alone or nothing), and a second analysis comparing patients treated with HCQ and AZ to those receiving AZ alone (excluding those receiving HCQ alone).