Comparative effectiveness of sotrovimab and molnupiravir for prevention of severe COVID-19 outcomes in non-hospitalised patients: an observational cohort study using the OpenSAFELY platform

Abstract Objective To compare the effectiveness of sotrovimab (a neutralising monoclonal antibody) vs. molnupiravir (an antiviral) in preventing severe COVID-19 outcomes in non-hospitalised high-risk COVID-19 adult patients. Design With the approval of NHS England, we conducted a real-world cohort study using the OpenSAFELY-TPP platform. Setting Patient-level electronic health record data were obtained from 24 million people registered with a general practice in England that uses TPP software. The primary care data were securely linked with data on COVID-19 infection and therapeutics, hospital admission, and death within the OpenSAFELY-TPP platform, covering a period where both medications were frequently prescribed in community settings. Participants Non-hospitalised adult COVID-19 patients at high risk of severe outcomes treated with sotrovimab or molnupiravir since December 16, 2021. Interventions Sotrovimab or molnupiravir administered in the community by COVID-19 Medicine Delivery Units. Main outcome measure COVID-19 related hospitalisation or COVID-19 related death within 28 days after treatment initiation. Results Between December 16, 2021 and February 10, 2022, 3331 and 2689 patients were treated with sotrovimab and molnupiravir, with no substantial differences in their baseline characteristics. The mean age of all 6020 patients was 52 (SD=16) years; 59% were female, 89% White and 88% had three or more COVID-19 vaccinations. Within 28 days after treatment initiation, 87 (1.4%) COVID-19 related hospitalisations/deaths were observed (32 treated with sotrovimab and 55 with molnupiravir). Cox proportional hazards models stratified by area showed that after adjusting for demographics, high-risk cohort categories, vaccination status, calendar time, body mass index and other comorbidities, treatment with sotrovimab was associated with a substantially lower risk than treatment with molnupiravir (hazard ratio, HR=0.54, 95% CI: 0.33 to 0.88; P=0.014). Consistent results were obtained from propensity score weighted Cox models (HR=0.50, 95% CI: 0.31 to 0.81; P=0.005) and when restricted to fully vaccinated people (HR=0.53, 95% CI: 0.31 to 0.90; P=0.019). No substantial effect modifications by other characteristics were detected (all P values for interaction>0.10). Findings were similar in an exploratory analysis of patients treated between February 16 and May 1, 2022 when the Omicron BA.2 variant was dominant in England. Conclusion In routine care of non-hospitalised high-risk adult patients with COVID-19 in England, those who received sotrovimab were at lower risk of severe COVID-19 outcomes than those receiving molnupiravir.