Abstract:
The objective of this study was to measure the university hospitals efficiency and identified the determinants of the efficiency. DEA technique specified with inputs (number of bed and number of physician) and outputs (OPD visits, IPD bed days and number of medical student year 6th) was used in the first step.
29 data from 5 university hospitals since 2001 to 2007 were analyzed. The analysis found that efficiency scores were ranged from 0.525 to 1, average was about 0.887. 72.4% of decision making units (DMUs) were found inefficiency in scale, while about 31.0% were inefficiency in technique. Among the scale inefficiency hospitals, about 95.2% were operated with decreasing returns to scale pattern. Also found that, UC implementation tended to increase technical efficiency level. Next step, regression analysis was done and the results showed that bed-physician ratio and pharmacist-physician ratio related to scale efficiency score significantly. Increasing in the ratios by 1 unit will make the score decreased by 0.08 and increased by 1.18 in sequences. For technical efficiency score was related to occupancy rate, out-patient visit-physician ratio and number of medical student year 6th-bed ratio significantly. Increasing in the factors by 1 units will make the score increased 0.01, increased 2.98E-05 and decreased 0.69 in sequences. Moreover, the study also found that hospitals under ministry of education control had higher technical efficiency level than was not. The reasons for this finding might be because of economy of scope and scale, management system and image of hospital. For policy makers, this study shows the evidence that most university hospitals were running in a decreasing return to scale pattern, so downsizing of the hospitals should be done to meet the most efficiency scale at constant return to scale pattern. Focusing on bed utilization at the maximal capacity or decreasing number of bed should be one solution to be considered because from the study shows that bed-physician ratio and occupancy rate highly significantly related to technical efficiency score.
