Abstract
This paper estimates the technical efficiency of hospitals using the DPC (Diagnosis Procedure Combination) data. This DPC data is a balanced panel of 127 hospitals in a three-year period between FY2005 and FY2007.
By using the DPC data, this study has the following three new points. First, in the efficiency estimation, the main output is measured by a number of inpatients weighted by case mix index from the DPC. This adjustment would improve the accuracy of the output measurement.
Second, we adopt “Hospital Standardized Mortality Ratio” as a quality control variable. Previous studies suffer from the lack of a quality variable and cannot control the quality of a medical service provided by hospitals.
Third, we adopted the “true fixed effect model” in Greene (2005) which distinguishes between inefficiency and heterogeneity of hospitals. This heterogeneity will be captured by a value of the fixed effect variable in the model.
As a result of efficiency estimation, the mean level of the efficiency is around 59%- 61%. This level is relatively low compared to previous studies. This would be caused by controlling the quality of medical service.
In addition, the value of the fixed effect variable has a 0.9 mean level and single peaked distribution. The results point out that fixed effect values would be a kind of indicator on the production or cost structure of hospitals.
The policy implication from these results is that the value of the fixed effect variable would be available to improve fairness among hospitals in the reimbursement system. This idea can apply to all kinds of Case Mix Groups and would require further examination.
