Impact of elevated pooling levels in basic medical insurance on insured individuals' inpatient service utilization-An...

China has continuously promoted the transition of basic medical insurance from municipal-level pooling to provincial-level pooling to strengthen fund risk resilience, mutual-aid capacity, and benefit equalization. Although existing studies suggest that higher pooling levels may influence healthcare utilization, limited micro-level evidence exists on how provincial-level pooling of the Basic Medical Insurance Scheme for Urban and Rural Residents (BMISURR) affects inpatient service utilization, hospitalization costs, and income-related equity. This study examines the impact of provincial-level pooling on inpatient care among BMISURR enrollees from both utilization and equity perspectives.

This study used seven waves of data from the China Family Panel Studies (CFPS) conducted between 2010 and 2022. After checking questionnaire skip patterns and excluding observations with missing key variables, the final analytical sample comprised 4,895 person-wave observations of BMISURR enrollees. A staggered difference-in-differences (DID) model with province and year fixed effects was employed to estimate the effects of provincial-level pooling on inpatient service utilization and hospitalization costs. An interaction specification was further used to compare the effects of two pooling models: the unified revenue-and-expenditure model and the provincial risk adjustment fund model. In addition, concentration indices (CI) and Wagstaff decomposition were applied to assess income-related inequality and identify the contribution of policy and socioeconomic factors to observed inequality in inpatient care.

Provincial-level pooling of BMISURR significantly increased both inpatient service utilization and hospitalization costs. The implementation of provincial-level pooling raised the probability of inpatient service utilization by 24.3% points and increased the average hospitalization costs per enrollee (including zeros for non-users) by approximately 445 yuan. Heterogeneity analysis by pooling model showed that both pooling models promoted inpatient utilization and expenditure, while the provincial risk adjustment fund model produced a stronger expansionary effect than the unified revenue-and-expenditure model. Concentration index (CI) results indicated that inpatient service utilization and hospitalization costs were concentrated among higher-income groups. Decomposition analysis further showed that income and healthcare resources were the dominant contributors to pro-rich inequality. Provincial-level pooling contributed to the pro-rich inequality pattern, accounting for 10.31% of inequality in inpatient service utilization and 12.78% in hospitalization costs. The interaction between pooling models contributed more substantially to inequality in hospitalization costs than to inequality in inpatient utilization, suggesting that institutional design differences are more closely related to expenditure disparities than to basic access disparities.

Provincial-level pooling of BMISURR expanded inpatient service utilization and increased hospitalization costs among insured residents, with stronger effects observed under the provincial risk adjustment fund model. However, expanded pooling did not automatically translate into improved equity. Instead, inpatient care remained concentrated among higher-income groups, largely due to income stratification and unequal healthcare resource distribution. Future reforms should strengthen provincial fund governance, improve expenditure control under different pooling models, introduce targeted protections for low-income and medically vulnerable groups, promote more balanced healthcare resource allocation, and integrate provincial-level pooling with hierarchical diagnosis and treatment and with payment reform.