Market structure and the quality of hospital outpatient departments for diabetes patients under the single-payer system in Taiwan.

摘要

The main purpose of this study is to explore the effect of competition from various levels of hospitals on the practice of outpatient care for diabetes patients. The principal research hypothesis is that under price regulation (FFS), the probability of receiving the recommended care is higher for patients living in markets with more hospitals that have higher accreditation levels. The assumption of homogeneous hospitals in the earlier work by Gregory Pope (1989) was relaxed to develop a two-group hospital Cournot model which is applied to a market setting where hospitals are heterogeneous due to the Taiwanese accreditation system.; The quality expenditure is measured by a proxy: the probability of conducting recommended procedures during an individual patient visit. The procedures recommended for the care of diabetes patients are outlined in the Developing and Evaluating Performance Measures for Ambulatory Care Quality (DEMPAQ) project. Separate logistic regressions with proper specification are estimated for the probability of receiving recommended care as a function of patient characteristics, hospital characteristics, hospital market structure, and market area characteristics. Competition is measured by the number of hospitals at each accreditation level. The Instrumental Variable (IV) method is employed to identify the endogeneity and measurement error problem in measuring competition.; Empirical findings show that competition cannot create positive direct effects on medical centers, but brings a negative spillover effect on regional hospitals. For regional and district hospitals, competition brings a positive direct effect on themselves and creates a positive spillover effect on medical centers.; From the point of view of welfare economics, the positive spillover effect from regional and district hospitals suggests that there might need to be a mechanism to reflect this positive externality in the payment scheme. In the aspect of resource allocation, the positive spillover effect of regional hospitals on medical centers is greater than the negative spillover effect of medical centers on regional hospitals, implying the net spillover effect is positive. Therefore, policy makers may need to allocate more medical resources to regional and district hospitals to treat patients with chronic diseases, like diabetes.