Abstract:
Output-orientated Data Envelopment Analysis was used to measure the technical efficiency of 25 regional hospitals in Thailand from 2007-2008 as 50 decision making units (DMUs). The results revealed, there were 31 efficient DMUs from 50 DMUs for overall technical efficiency scores and a minimum was 0.810. There were 36 efficient DMUs from 50 DMUs for pure technical efficiency scores and a minimum was 0.817. For scale efficiency scores, there were 32 efficient DMUs from 50 DMUs and a minimum was 0.889. In addition, medians of all three scores were 1.000. Most of patterns of scale inefficiency were the increasing return to scale (irs:drs = 14:5). In medical education services, the results found teaching hospitals were more efficient than non-teaching hospitals and a combined undergraduate and postgraduate teaching hospital was the most efficient group. The next step was to identify the determinants of hospital efficiency with regression analysis using ordinary least squares (OLS). The results revealed if beds-physician ratio and other personnel-physician ratio decreased one unit, pure technical efficiency scores tended to increase 0.029290 and 0.008336 units respectively. If nurses-physician ratio and trained internsphysician staff ratio increased one unit, pure technical efficiency scores tended to increase 0.023639 and 0.208326 units respectively. And the most influential explanatory variable of pure technical efficiency scores was a trained interns-physician staff ratio. For scale efficiency scores, if in-patient visits adjusted with relative weight of DRG per physician increased one unit, scale efficiency scores tended to increase 0.000110 units. All above information could be used for policy makers in health sector and hospital managers improve the inefficient regional hospitals in proper direction such as most of patterns of scale inefficiency were the increasing returns to scale which can be improved through up-sizing and should supported medical education in regional hospitals which have competency. In addition, the details of each inefficient hospital should be explored and analyzed with the information from DEA and regression analyses
