摘要：Introduction: Hospitals are regarded as the largest and most expensive operational units of the health system. Also, bed is one of the most valuable hospital resources. Due to the limited resources of the health system, paying attention to efficient resource allocation is necessary. This study aimed to allocate common beds to hospitals to optimize their overall efficiency. Methods: This is a cross-sectional and applied study. 70 Iranian hospitals affiliated to social security organizations were examined. The required data were collected from the statistical yearbook of this organization (2016). A centralized data envelopment analysis model with an input-orientation approach was used to assess the performance and bed reallocation to hospitals. The input data consist of the number of active beds, and the output data consist of bed occupancy rate, average patient stay, and survival of patients in thousand. Hospitals were clustered by k‐means clustering method for analysis. The data were analyzed by GAMS and SPSS softwares. Results: Hospitals were clustered into 4 groups of homogenous units. Based on this clustering, the number of hospitals in clusters 1 to 4 was 23, 2, 13, 32, and the overall efficiencies of them were 0.703, 1, 0.827, and 0.732, respectively. Before bed reallocation, 13(18.6%) hospitals were efficient. After bed reallocation by Centralized data envelopment analysis, 31(44.3%) hospitals became efficient. Conclusion: It seems that the active bed factor can be one of the inputs influencing the overall efficiency of hospitals. However, in this regard, inclusive attention to other resources such as physicians and nurses will be necessary to achieve desirable hospital efficiency.
