Changes in General Anesthesia Utilization Disparity among Patients with Autism Spectrum Disorder Before and After Health Insurance Coverage Expansion: A Retrospective Analysis of a Single Regional Specialized Dental Center, 2021-2025

Background: Patients in Korea with autism spectrum disorder (ASD) rely heavily on general anesthesia (GA) during dental treatments at specialized dental centers.In February 2022, the expanded National Health Insurance (NHI) coverage reduced out-ofpocket expenses for patients undergoing dental GA by approximately 67%.However, it is unclear whether insurance-type disparities in GA access changed after NHI expanded.Methods: Administrative data (2021-2025) and electronic medical records of 526 patients with ASD (2023-2025) from a specialized dental center were retrospectively analyzed.Insurance-type disparities were assessed using Fisher' s exact tests, differencein-differences, and multivariable logistic regression.Interrupted time series models examined monthly GA counts across all disability types, adjusting for resident resignation.Results: Patients with ASD showed the highest GA rate among all disability types across all study years (26.2% in 2021 and 15.5% in 2025).In 2021, GA rates were higher among NHI patients than among medical aid (MA) patients (28.6% vs. 3.4%; odds ratio [OR]=11.20,p=0.002).The NHI-MA OR had converged toward unity (0.93) by 2025 and the difference-in-differences analyses confirmed this narrowing (2021 vs. 2025: OR=12.07,p=0.022).The interrupted time series regression analysis, adjusted for the nationwide resignation of resident physicians, revealed a significant post-policy increase in monthly GA procedures among MA patients (+0.534/ month, p=0.001).Behavioral uncooperativeness (adjusted OR=23.01) and treatment complexity (adjusted OR=89.20) were the principal GA predictors, but insurance type was not an independently associated predictor (adjusted OR=1.38, p=0.423).Conclusion: A substantial insurance-type disparity in GA access existed before the 2022 coverage expansion; however, this disparity was no longer statistically significant after coverage expansion.This pattern is consistent with policy implementation, although the single-center design and concurrent operational disruptions preclude causal inferences.There is an urgent need for national analyses that link health insurance reviews and assessment service claims to disability registration data.