Evaluation of the DPC-based inclusive payment system in Japan for cataract operations by a new model

We propose a new model for analyzing the length of hospital stay when variances are heterogeneous. This model is an alternative to the conventional models such as Cox's proportional hazard model and can be used to address various problems of survival analysis. The model is a heteroskedastic version of the Box–Cox transformation model and is estimated by the tobit maximum likelihood method, and its estimation can be easily done using a standard statistical package program. Using the proposed model, we analyze the effects of the diagnosis procedure combination/per diem payment system (DPC/PDPS) on the length of hospital stay following cataract operations in Japan. Data collected from six general hospitals before and after the introduction of the system are analyzed. The number of patients is 2677.