ObjectivesIn this retrospective cohort study, we aimed to assess clinical effectiveness and viral clearance following the use of molnupiravir, azvudine and paxlovid in hospitalized patients with COVID-19 in China dominated by the omicron BA.5.2 and BF.7 subvariant of SARS-CoV-2. MethodsEnrolled patients were assigned to the molnupiravir group or the azvudine group or the paxlovid group or the control group (not taking any antiviral drugs). The primary outcome of the cohort study was viral clearance and viral burden rebound after treatment and the secondary outcome was 28-day all-cause mortality. The four groups were propensity score-matched (1:1). We plotted viral load trends for each antiviral drug intervention using locally weighted regression (LOWESS) smoothed data. Multivariate logistic regression (stepwise algorithm) models were used to determine any risk factors for 28-day mortality. ResultsOf the 1537 patients receiving any treatment, 886 (57.6 %) received molnupiravir, 390 (25.4 %) received azvudine, 94 (6.1 %) received paxlovid, and 167 (10.9 %) did not use any antiviral drugs. Our data analysis showed that age (OR = 1.05, 95 % CI: 1.03-1.07, P < 0.001), Charlson comorbidty index (OR = 1.32, 95 % CI: 1.18-1.48, P < 0.001), severity of COVID-19 (P < 0.001), gamma globulin (OR = 2.04, 95 % CI: 1.03-3.99, P = 0.039) and corticosteroids use (OR = 2.3, 95 % CI: 1.19-4.69, P = 0.017) were independent prognostic factors for 28-day mortality in COVID-19 patients. After propensity score matching (PSM), the paxlovid recipients (OR = 0.22, 95 % CI: 0.05–0.83, P = 0.036) or azvudine recipients (OR = 0.27, 95 % CI: 0.07-0.91, P = 0.046) had lower 28-day mortality compared to their matched controls. Viral rebound occurred in the control group around days 9-16, while no viral rebound was found in any of the three oral antiviral groups. We found that molnupiravir group performed comparably in terms of the rate of nucleic acid conversion negative compared with the paxlovid group, while azvudine group performed slightly worse compared with the paxlovid group or molnupiravir group. ConclusionsIn our retrospective cohort of hospitalized patients with COVID-19 during the wave of omicron strain, the molnupiravir, paxlovid and azvudine recipients showed a faster and more stable decrease in viral load and rare virus rebound in response to antiviral treatments when compared to the controls. The study supported that initiation treatment with paxlovid and azvudine was associated with significantly lower risk of all-cause death within 28 days.