Measuring Technical Efficiency of Hospitals affiliated with Shahrekord University of Medical Sciences, Using a Combination Method Data Envelopment Analysis (DEA) – Principle Component Analysis (PCA)

Background: Measuring the efficiency of hospitals due to the high proportion of budget allocated to them on the one hand, and the need to ensure the best practices regarding the use of scarce resources on the other hand, is of particular importance. The purpose of this study is to evaluate the technical efficiency of the affiliated hospitals of Shahrekord University of Medical Sciences by using a combination of Principal Component Analysis and (PCA) & Data Envelopment Analysis (DEA). Methods: This was an analytical and cross-sectional study measuring the technical efficiency of all 8 hospitals affiliated to Shahrekord University of Medical Sciences. The required information was collected from the medical records unit of each hospital. For better differentiation between efficient and inefficient units, and the increase of research accuracy and further differentiation between hospitals in terms of efficiency, at first, 17 indicators were selected to assess and adjust these parameters to 3 components proportional to the number of the hospitals by using PCA and SPSS 16 software. After doing the PCA, 7 studied input variables became 7 principal components among which the first input component reflecting the 83 % of scattering data was selected as principal input component, and for being more influenced by human resource variables, it was named as a human resource index. Furthermore, among the output variables, the first 2 output components, which represented 76% of the variance of the data, were selected as the 2 principal components of the output for the study, which were mostly affected by these variables, respectively, the number of admissions and length of stay. Then, the modified input and output components were entered into the software Windeap 2.1 and the technical efficiency of hospitals and their rank were calculated by assuming constant and variable efficiency with respect to the scale. In order to evaluate the effect of using the combined method instead of the conventional method of efficiency measurement, the results of the PCA - DEA method were compared with the results of the conventional DEA method. Results: The result of DEA on the selected components showed the capacity to upgrade the Technical Efficiency (TE) of hospitals is 15 % (TE: 0.852). Moreover, out of 8 hospitals, 1 hospital was increasing return to scale, 3 decreasing returns to scale and 4 constant returns to scale. The technical efficiency of 3 hospitals was 1 (TE = 1), 2 hospitals had the technical efficiency between 0.80 to 1 (1 > TE > 0.80) and that in 3 hospitals was less than 0.80 (TE < 0.80).  The scale efficiency for 50 % of hospitals and the management efficiency for 62/5 % of them were equal 1. Conclusion: The average of total technical efficiency, management efficiency and scale efficiency were calculated to be 0.999, 1 and 0.999, respectively based on the usual comprehensive analysis method; while using the combined method, the average total technical efficiency, management efficiency and scale efficiency were 0.852, 0.947 and 0.902 respectively. The results confirm that the use of PCA method, due to its important role in reducing alignments, increases research accuracy and better differentiates between hospitals in terms of efficiency.