Public Hospital Performance Evaluation Method And Application In The 13 Hospitals

Objective:Reasonable evaluation index system of public hospital performance and the methods to determine the index weights are searched for. Through the evaluation for the performances of 13 level-3 general hospitals in a city for six years from 2004 to 2009, the advantages and disadvantages of different evaluation methods are compared. The more scientific, more rigorous, and more reasonable evaluation methods of public hospital performance are found, providing the reference for the achievement of the new medical reform of the public hospitals.Materials and methods:Data come from the hospital management data report from 2004 to 2009 collected by a provincial health department, including nine parts, such as hospital management, basic conditions of hospitals, the quantity of medical service, the quality of medical service, the index of nursing quality, the indicators of controlling the quality, the medical service, the medical security and the economic benefit, and total 117 indexes. 13 level-3 general hospitals are chosen as evaluated objects and another two special hospitals as references.The indexes with good representation and high sensitivity are chosen from the 117 indicators which can fully reflect the performance of the public hospitals. The stratified analysis is used to determine the first-order index weights and the entropy method to determine the second-order index weights. We also employ the weighted TOPSIS method, the weighted rank sum method and the principal component scores for the comprehensive evaluation. The statistical software SAS9.2 is used to complete the data cleaning, sorting and statistical analysis. Results:Public general hospital comprehensive performance evaluation were done from the medical quality, diagnosis quality, patients angle, hospital internal quality management, financial operations, workload and single disease index, total 7 ways. In six years, the comprehensive performance of 13 general hospitals shows ascendant trend. From the Euclidean distance of 2009 is 2 times of 2008. But there was not a huge gap reflected by the weighted RSR method evaluation. Every hospital evaluation result shows bigger difference in 2009. TOPSIS evaluation results show that H6 is the optimal ( C ? 0.8117), while H5 is the second ( C ? 0.3083), the worst is H3 ( C ? 0.0631), then is H1. The result of Weighted RSR method is used to classify 13 hospitals by linear regression. H5 and H6 are in first level. H2, H11, H9 H13 and H8 are in second level. H1, H4, H12, H10 and H7 are in third level. H3 is in last level. Principal component scoring method of analysis results show that the optimal H6 (f=0.600), H5 is second (f=0.377). The difference of score indicates it is a great distance between H6 and H5. This is consistent with results from TOPSIS analysis. From the above results, the results of the weighted TOPSIS method, the weighted RSR method and principal component scoring method are similar. Observe the optimal and the worst, hospitals with higher ranked at medical quality, finance, found operation, workload are bad at diagnosis of quality, patient angle, and internal quality management and single disease index, such as H1.Conclusion:The overall development trend of the city's general hospitals is good. 13 hospitals have the significant differences in every aspect. The structural differences exist between the specialized hospital and the general hospital, as a result, they are not suitable for the joining investigation without reliable and effective calibration standard before. Analytic hierarchy process (AHP) and combination entropy weight method are combined to determine the weights of index system with the condition of not deviating the research topics and applying the objective method to establish the reasonable weights, achieving the unity of theory and practice. Meantime the weights can reflect the regular changes. Although all aspects of investigation are fully covered by the comprehensive evaluation index system of the performance determined by this study, the unbalanced characteristics among the hospital patients caused the partiality of the evaluation. Therefore, we need to further adjust index system in order to make the result more reliable.