Peer Review Report For: Does the Home-death Rate Reflect Medically Attended Community Deaths in Japan? [version 1; peer review: 1 approved]

Background The home-death rate is widely used in Japan as an indicator of community-based end-of-life care; however, this rate includes deaths without professional involvement. This study examines whether the home-death rate reflects medically attended home-based deaths across municipalities. Methods An ecological panel study was conducted using data from 1,741 municipalities obtained in 2017, 2020, and 2023 (5,222 municipality-year observations). National datasets from the Ministry of Health, Labor, and Welfare include data on medical institutions, home-visit nursing services, and clinic-certified home deaths. The dependent variable was the proportion of deaths occurring at home. Explanatory variables included home deaths certified by home-care support clinics and general clinics as well as deaths under home-visit nursing reimbursed by medical or long-term care insurance. Fixed-effects panel regression models with municipality and year effects were applied, using cluster-robust standard errors. Finally, correlations between the model residuals and prefectural rates of coroner-investigated deaths were examined to explore the influence of non-medical deaths. Results Home deaths certified by general clinics and deaths under medical-insurance-reimbursed visiting nursing were significantly associated with higher home-death rates (β=4.06, 95% confidence interval (95%CI), 1.28–6.85; β=31.75, 95%CI 8.14–55.37, respectively). The associations between home-care support clinics and long-term care insurance-reimbursed visiting nursing were not significant. No correlation was found between residuals and prefectural rates of coroner-investigated deaths (r=0.09, 95%CI, –0.20–0.37). Conclusions Japan’s home-death rate partly reflects medically attended home deaths, particularly those involving general clinics and medical-insurance-reimbursed visiting nursing. However, because this rate also includes deaths without professional care, it should be interpreted as a contextual rather than a quality indicator of end-of-life care and complemented by measures capturing care involvement and coordination.