The characteristics of coronavirus disease 2019 (COVID-19) clusters in medical and social welfare facilities and the factors associated with cluster size are still not yet fully understood. We reviewed COVID-19 cases in Japan identified from January 15 to April 30, 2020 and analyzed the factors associated with cluster size in medical and social welfare facilities. In this study, COVID-19 clusters were identified in 56 medical and 34 social welfare facilities. The number of cases in those facilities peaked after the peak of the general population. The duration of occurrence of new cases in clusters was positively correlated with the number of cases in both types of facilities (rho = 0.44, P < 0.001; and rho = 0.69, P < 0.001, respectively). However, the number of days between the first case in a prefecture and the onset of clusters was negatively correlated with the number of cases only in clusters in social welfare facilities (rho = - 0.4, P = 0.004). Our results suggest that COVID-19 cases in those facilities were prevalent in the latter phase of the disease's community transmission, although the underlying mechanisms for such a trend could differ between medical and social welfare facilities.
Keywords: COVID-19; cluster-based approach; medical facilities; social welfare facilities.