AbstractObjectiveWe compared the use of various care services and institutional deaths in older adults among these facility types.DesignThis was a retrospective cohort study.MethodsWe used administrative claim data from April 2014 to March 2017. The study participants comprised Fukuoka Prefecture residents aged 75 and older with certified care needs of level 3 or more in April 2014 and who received home care services during the study period. Participants were divided into 4 groups according to the facility type from which they received home care services: General Clinics, Home Care Support Clinics and hospitals (HCSCs), Enhanced HCSCs with beds and Enhanced HCSCs without beds. The outcomes were utilization of medical and long-term care services and the incidence of institutional deaths. We constructed generalized linear regression models. The evaluated potential risk factors were sex, age, care needs levels, and Charlson comorbidity index (CCI) scores.ResultsThe numbers of inpatient care days were 54.3 days, 70.0 days, 64.7 days, and 75.1 days for users of enhanced HCSCs with beds, enhanced HCSCs without beds, HCSCs, and general clinics, respectively. While the number of home care days were 63.8 days, 50.9 days, 57.8 days, and 29.0 days, respectively. The odds of institutional death in general clinic users were 2.32 times higher (P<0.001) than users of enhanced HCSCs with beds.ConclusionsParticipants who used enhanced HCSCs with beds had reduced inpatient care utilization, increased home care utilization, increased home-based end-of-life care utilization, and fewer institutional deaths. These findings suggest that hospitalizations and institutional deaths could be reduced by further expanding the role of enhanced HCSCs with beds. Our study provides useful information for further investigations of home care for older adults as part of community-based integrated care.Strengths and limitations of this studyThis was a retrospective cohort study including 18,347 participants.We followed up participants for 3 years.We considered the level of care needs and Charlson comorbidity index as confounders. Despite that, the inclusion of these variables did not provide detailed information about living conditions that reflect family structure and characteristics of living.We calculated the number of years that participants lived during the study period, and the annual utilization rates per person-year of observation were estimated.There were no clinical data for individual participants because this study focused on the types of healthcare facilities that provide home care.