The Impact of Floating Point Value under Global Budget System on the Health Care Utilization and Quality for the Dialysis Treatment in Taiwan

This study plans to analyze the impact of floating point value under the global budget system on the health care utilization and quality for dialysis treatment of ESRD patients in Taiwan. The Aim 1 of this project focuses on the dimension of the “revenue，，of a dialysis center. We hypothesizes that a dialysis center will provide more hemodialysis treatment for a patient in a given period of time in order to compensate its potential loss caused by floating point value system. We use two levels of estimation for testing this hypothesis. The first one focuses on the patient level. Two types of estimation are planning to perform, the first one focuses on the number of hemodialysis in a month for a patients, and the second one focuses on the probability for a patient that his/her total amount of dialysis treatment in a month exceed the regulated maximum (12). Lastly, we analysis the probability of being transfer to a higher ranked dialysis facility, such as hospital after the policy. The second level estimation use provider as unit. We hypothesizes that in order to compensate the potential loss due to floating point values, a dialysis might “steal” costumers from other dialysis providers. The aim 2 of this project analyzes whether the floating point value system increases the possibilities of being infected while taking hemodialysis. The hypothesis emphasizes that the floating point value system decreases the profit level for a dialysis center. In order to compensate this potential loss, one might use less inputs for its treatment service so that increase the possibility of being infected for a patients while taking hemodialysis treatment. The NHRI data is used for this project, and a series of difference-in-difference estimation is going to be performed. In addition, due to the potential selection bias for patients in our treatment and control group, a PSM method is used to tackle this potential econometric problem. Furthermore, the statistic module developed by Norton et al. (2004) is used to deal with the problem of DID policy effect under nonlinear estimation.

Project ID:PF10408-0720
External Project ID:MOST104-2410-H182-030