Diffusion and geographic variation in robot-assisted colon resection after reimbursement in Japan: a nationwide...

This study aimed to quantify how rapidly robot-assisted colon resection billing diffused after reimbursement in Japan and whether geographic variation narrowed as uptake expanded. We analyzed National Database (NDB) Open Data for fiscal years (FYs) 2022-2024. The primary outcome was the robot-assisted share of two laparoscopic billing categories for malignant colon tumor resection (robot-assisted and conventional). Variation across treating-institution prefectures was summarized using the interquartile range (IQR) for absolute spread and the coefficient of variation (CV) and Gini coefficient for relative dispersion. Longitudinal analyses included prefectures with unmasked values in all relevant years; FY2024 permitted a complete 47-prefecture cross-section. The national robot-assisted share rose from 3.38% in FY2022 to 11.40% in FY2023 and 19.55% in FY2024. In the common 33-prefecture panel, the IQR widened from 5.73 to 6.61 and 11.79 percentage points, whereas the CV fell from 1.086 to 0.630 and 0.482 and the Gini coefficient from 0.565 to 0.347 and 0.269. Across all 47 prefectures in FY2024, shares ranged from 0.00% to 41.47%; the complete cross-section was more heterogeneous than the restricted longitudinal panel, and the 14 prefectures excluded because of earlier masking had a lower median share (6.83% versus 18.05%). Rapid national growth did not translate into geographically uniform billing-based uptake. Relative dispersion decreased while absolute percentage-point spread widened, so apparent geographic convergence depended on the scale used. Post-reimbursement surveillance should report both dimensions and account for selection introduced by disclosure-related masking.