A taxonomy of Chinese hospitals and application to medical dispute resolutions.

Medical disputes can be viewed as a negative indicator of health care quality and patient satisfaction. However, dispute prevention from the perspective of systematic supervision is unexplored. This study examines hospital clustering based on diagnosis-related group (DRG) indicators and explores the association between hospital clusters and medical disputes. Health administrative data from Sichuan Province in 2017 were used. A twostep cluster analysis was performed to cluster hospitals based on DRG indicators. A multiple regression analysis was conducted to evaluate the relationship between clusters and the incidence/number of medical disputes. The 1660 hospitals were grouped into three DRG clusters: basic (62.5%, n = 1038), diverse (31.0%, n = 515), and lengthy (6.4%, n = 107). After adjusting for covariates, the diverse hospitals were associated with an increased probability of having medical disputes (OR 5.24, 95% CI 2.97-9.26), while the diverse and lengthy hospitals were associated with a greater number of medical disputes (IRR 10.67, 95% CI 6.58-17.32; IRR 4.06, 95% CI 1.22-13.54). Our findings highlighted that the cluster-level performance of hospitals can be monitored. Future studies could examine this relationship using a longitudinal design and explore ways to reduce medical disputes in hospitals.