Projected inpatient demand in Japan through 2050 using municipal population projections; a national wide repeated cross-sectional study

Background Japan is one of the most rapidly ageing societies in the world and has experienced sustained population decline since 2008. Understanding how these demographic changes will influence future inpatient demand is critical for planning hospital capacity, workforce allocation, and sustainable health care delivery. However, many existing projections rely on static assumptions regarding inpatient utilization and may not fully capture dynamic changes over time or regional heterogeneity. Methods This repeated cross-sectional study used nationwide aggregated data from Japan’s Patient Survey (1999–2023) and population data from the national census and official projections. Age-, sex-, and prefecture-specific inpatient utilization rates were modelled using three approaches: (1) age–period–cohort models; (2) statistical trend models, including linear, log-linear, and spline specifications; and (3) conservative scenario-based models, including constant and exponential decay assumptions. Projected utilization rates were applied to municipal population projections using an indirect standardization approach, and estimates were aggregated in a bottom-up manner to prefectural and national levels for 2025–2050. Spatial patterns were examined using geographic information systems. Results Between 1999 and 2023, inpatient utilization rates declined across most age groups despite rapid population ageing, accompanied by an overall reduction in the number of inpatients. Models that closely reproduced historical trends consistently projected further declines in utilization through 2050. When combined with population projections, total inpatient demand was projected to decrease in most scenarios. Conservative assumptions suggested temporary stabilization or modest increases around 2030, followed by long-term decline. Spatial analyses indicated increasing concentration of inpatient demand in metropolitan and regional core areas, whereas many peripheral municipalities were projected to experience very low inpatient volumes. Model comparisons demonstrated that assumptions regarding future utilization trends substantially influenced projected demand trajectories. Conclusion In Japan, inpatient demand is likely to decline and become increasingly concentrated in urban areas despite continued population ageing. These findings suggest that population ageing alone does not determine future hospital demand. Instead, dynamic changes in utilization patterns may offset demographic pressures. Policy responses should emphasize region-specific strategies to optimize hospital capacity, support vulnerable facilities in low-demand areas, and adapt health systems to population decline.