This study analysed national and prefecture‑level claims data from Japan's National Database of Health Insurance Claims (NDB Open Data) covering fiscal years 2014 through 2024. The investigators used only fully aggregated public datasets, combined with population estimates and other publicly accessible sources listed in the study protocol and replication package. The analysis focused on health insurance billings specifically coded for ongoing post‑implantation deep brain stimulation (DBS) management.
Annual billings for DBS management were expressed per million population to standardise activity across prefectures. Between‑prefecture inequality in DBS management billing rates was summarised using the Gini coefficient, calculated annually to characterise the degree of concentration or unevenness. The authors evaluated longitudinal trajectories for each prefecture relative to the national billing rate and examined how baseline billing rates related to subsequent trends. All data sources and analysis code are reported as publicly available in the replication package cited by the authors.
At the national level, billings for ongoing DBS management increased over the study interval. Annual national DBS management billings rose from 18,659 in 2014 to 27,233 in 2024. Across the entire period, the dataset contained a cumulative total of 258,408 DBS management billings. The overall trend indicates growth in billed DBS management activity at the national scale during the decade examined.
Despite national growth, geographic distribution of DBS management activity remained markedly uneven. The between‑prefecture Gini coefficient of prefecture DBS management billing rates ranged from 0.54 to 0.40 over the study period, indicating high inequality that narrowed only slowly. The pace of narrowing in the Gini coefficient decelerated over time rather than showing rapid convergence.
By 2024 the disparity across prefectures remained substantial: the prefecture at the 90th percentile of billing rate billed at 9.1 times the rate of the prefecture at the 10th percentile. This ratio highlights persistent concentration of ongoing DBS management activity in higher‑billing prefectures relative to lower‑billing ones.
The authors classified prefectures by baseline billing rates and followed their relative trajectories. Of 23 prefectures identified as having low baseline DBS management billing rates, nine showed a subsequent decline in relative billing rate during the follow‑up period, indicating further divergence for some low‑activity areas. However, across the full sample the analysis found that lower baseline billing rates were associated with a more positive relative rate of change overall, meaning some low‑baseline prefectures improved relative to the national rate even as many remained low in absolute terms.
The authors interpret these findings to mean that, in Japan, sustained post‑implantation DBS management activity is geographically concentrated and that this concentration persisted across a decade despite increasing national activity and a universal health coverage system. They emphasise that national growth in aggregate DBS management billing and universal coverage alone did not produce equitable long‑term access to specialist DBS care across prefectures.
The nationwide, longitudinal description is presented as a potential evidence base to inform policies aimed at reducing geographic barriers to long‑term specialist follow‑up for DBS patients across healthcare systems. The study does not make clinical care recommendations; it provides descriptive evidence on distributional patterns of billed DBS management activity.
All input data used in the analysis were publicly available; the authors report that the source manifest, processed analytic data, and complete analysis code are available in a replication package. The authors declared no funding for the study and no competing interests. The report is posted as a preprint and has not been certified by peer review. The authors state that only aggregated public data were used and that no individual‑level or identifiable information was accessed.