Impact of distribution of facemasks on community incidence and outcomes of COVID-19: A cluster randomised trial in India

Background: Randomised evidence on the effectiveness of delivering facemasks for reducing the health impact of COVID-19 remains limited. Methods: We conducted a cluster randomised trial in Telangana, India, in August–November 2020 to investigate whether distribution and promotion of facemasks in villages reduces the incidence and adverse outcomes of COVID-19. We randomised 20 villages from the ongoing APCAPS surveillance study (1:1 ratio) to village-wide distribution of 3-layer cloth facemasks along with promotional messaging, or no intervention. Outcomes were incidence of notified COVID-19 (primary), hospitalised COVID-19 and all-cause mortality (secondary), collected through household surveillance by village health workers. Mask wearing was assessed through standardised observations in village centres. Data were analysed by multilevel Poisson regression. Findings: Use of the study facemasks and any face coverings in public spaces was higher in intervention villages (19% and 59%, respectively) than control villages (0% and 38%). In the 10 intervention (N=16,741 adults) and 10 control villages (N=15,278 adults), respectively, the crude incidence per thousand person months (number of events) of notified COVID-19 was 2.15 (n=144) and 2.45 (n=150), of hospitalised COVID-19 was 0.07 (n=5) and 0.21 (n=13), and of all-cause mortality was 0.91 (n=61) and 1.10 (n=67). In models accounting for age, sex and pre-intervention COVID-19 rate, rate ratios in intervention versus control villages were 0.96 (95% confidence interval 0.57–1.63) for COVID-19 cases, 0.36 (0.12–1.05) for COVID-19 hospitalisations, and 0.84 (0.55–1.29) for all-cause mortality. No adverse effects were reported. Interpretation: We are unable to draw firm conclusions about the effect of village-wide distribution and promotion of facemasks on COVID-19 incidence from these data due to a low number of events leading to imprecise effect estimates. Nonetheless, our findings are consistent with the modest protective effect on incident cases seen in previous randomised trials, extending these to adverse outcomes for the first time. Registration: The trial was pre-registered on the Clinical Trials Registry of India on 25/07/2020 (CTRI/2020/07/026796).