Regional variation in long-term care spending in Japan

Background Health inequalities are widening in Japan, and thus, it is important to understand whether (and to what extent) there is a regional variation in long-term care (LTC) spending across municipalities. This study assesses regional variation in LTC spending and identifies the drivers of such variation. Methods We conducted a cross-sectional study using publicly available municipality-level data across Japan in 2019, in which the unit of analysis was municipality. The outcome of interest was per-capita LTC spending, which was estimated by dividing total LTC spending in a municipality by the number of older adults (people aged >= 65). To further identify drivers of regional variation in LTC spending, we conducted linear regression of per-capita spending against a series of demand, supply, and structural factors. Shapley decomposition approach was used to highlight the contribution of each independent variable to the goodness of fit of the regression model. Results In Fiscal 2019, per-capita LTC spending varied from 133.1 to 549.9 thousand yen (max/min ratio 4.1) across the 1460 municipalities analyzed, showing considerable regional variation. The included covariates explained 84.0% of the total variance in LTC spending, and demand-determined variance was remarkably high, which contributed more than 85.7% of the overall R-2. Specifically, the highest contributing factor was the proportion of severe care-need level and care level certification rate. Conclusions Our results demonstrate that, even after adjusting for different municipalities' age and sex distribution, there is a large variation in LTC spending. Furthermore, our findings highlight that, to reduce the spending gap between municipalities, the issues underlying large variations in LTC spending across municipalities must be identified and addressed.