In our study, a duration of SARS-CoV-2 RNA shedding of >23 days (above the median) was defined as prolonged viral shedding. The main characteristics for patients with prolonged viral shedding and those without are shown in table 1. Prolonged RNA shedding was associated with a delayed recovery of radiological manifestations (median 20 days versus 13.5 days; p<0.001) and a delayed recovery of the body temperature (median 12 days versus 7.5 days; p<0.001). In the final multivariate logistic model that included all 120 patients, older age (OR 1.03, 95% CI 1.00–1.05; p=0.03) and the lack of LPV/r treatment (OR 2.42, 95% CI 1.10–5.36; p=0.029) were independent risk factors associated with prolonged SARS-CoV-2 shedding (table 2 and supplementary table E2). However, comorbidities (including current smoking, hypertension, diabetes and cardiac disease) and the administration of systemic corticosteroids were not associated with prolonged viral shedding (table 2 and supplementary table E2). Similar risk factors for prolonged SARS-CoV-2 RNA clearance were observed when the data were analysed using a Cox proportional hazards model (supplementary table E3, and supplementary figures E1 and E2).