Impact of caseload-threshold centralization of intensity-modulated radiotherapy on cancer care regions in Japan: a nationwide ecological geospatial study

Background In Japan, a national expert panel has designated intensity-modulated radiotherapy (IMRT) for centralization at the level of single or multiple cancer care areas to improve outcomes and sustain the delivery system, projecting demand growth through 2040. Radiotherapy requires repeated visits; thus, centralization trades off quality for geographic access. However, the impact of the geographic unit of centralization has not been quantified before implementation. We evaluated the impact of caseload-based centralization of IMRT facilities on patient access by centralization unit (prefecture versus secondary medical area [SMA]) and demand year (2025 and 2040). Methods We identified 611 IMRT-providing facilities from the FY2025 Bed Function Reports and derived a centralization threshold of 98 treatments annually using equipercentile correspondence. Accessibility was computed using an enhanced two-step floating catchment area method with 500-m mesh utilization-adjusted demand, comparing two scenarios reallocating the caseloads of withdrawing facilities to the nearest remaining facility. Outcomes were coverage (mesh- and demand-based), SMA accessibility score median (ASM), Gini coefficients, and cluster-based typology. Results Overall, 166 facilities (27.2%) were designated for centralization. Prefecture-level centralization increased the share of demand beyond 60 minutes from 4.46% to 5.67% and resulted in loss of accessibility in 19 SMAs. Restricting centralization to SMAs limited withdrawals to 125 facilities, limited the increase in demand beyond 60 minutes to 0.14 percentage points, and loss of accessibility to three SMAs. Conclusions Changes in accessibility depend not only on the caseload criterion but also on the geographic unit of centralization. Cancer centralization requires explicit consideration of geographic accessibility.