Several systematic reviews have been conducted on hospital efficiency worldwide [20-21]. For example, a 2018 study reviewed 57 articles using DEA [20] and a 2014 study reviewed 23 articles using DEA, SFA, and balanced scorecard [21]. To our knowledge, this is the first attempt to measure hospital efficiency using meta-analysis in Eastern Mediterranean countries. There was a growing trend in recent years to measure the efficiency of hospitals using different methods. In this study, we reviewed studies that measured the TE of hospitals in EMR countries. Fifty articles which calculated hospital efficiency were eligible for inclusion in the meta-analysis.
It must be noted that the vast majority of studies on hospital efficiency were conducted in Iran. This may partly be due to the Iranian Ministry of Health and Medical Education’s attempt at reducing hospital costs. In addition, efficiency and strategies for improving it have become a key priority for the Iranian government.
A mean TE of 0.829 ± 0.026 was obtained for Eastern Mediterranean countries. This finding is consistent with the results of previous studies in other countries [6, 22- 23]. In the study conducted by Du (2017) on Chinese hospitals, a mean efficiency of 0.74, 0.902, and 0.805 was obtained for hospitals in the Central, Eastern, and Western regions of the country respectively [22]. Blatnik et al. (2017) examined hospital efficiency in Slovenia and reported a mean TE of 0.936 [23]. In general, hospital efficiency varies by country.
Mean hospital efficiency varied in high-income countries such as Saudi Arabia, Oman, the United Arab Emirates, and Bahrain. For example, Oman and Saudi Arabia had the highest mean hospital efficiency rates and the United Arab Emirates had the highest mean TE. According to the 2017 WHO report on “Eastern Mediterranean Region Framework for health information systems and core indicators for monitoring health situation and health system performance” Bahrain and Oman had the highest general government expenditure on health as percentage of general government expenditure (10.5% and 6.8% respectively) among the four countries [24]. In addition, mean hospital efficiency varied in low- and middle-income countries such as Pakistan, Afghanistan, Iran, Jordan, Tunisia, Palestine, and Iraq. For example, among these countries, Iraq and Jordan had the highest mean TE and Pakistan had the lowest mean TE. The 2017 WHO report showed that among these four countries, Iran had the highest and Pakistan the lowest general government expenditure on health as percentage of general government expenditure (17.5% and 4.7% respectively) [24]. Therefore, hospital managers and policymakers must focus on improving efficiency and reducing healthcare costs in regions that have lower rates of hospital efficiency.
The development of outpatient care, the strengthening of hospital management, and modeling based on efficient hospitals are recommended as effective strategies to increase hospital efficiency. In addition, hospitals can serve as productive business entities through health system structure reform at the macro level, proper implementation of healthcare stratification, and responsiveness of insurance companies. This allows hospitals to increase patient satisfaction and provide safe, high-quality care.
Although there are many studies on hospital efficiency in the EMR, their results have been highly heterogeneous. One reason for the observed inconsistencies may be the different methods used to measure hospital efficiency, leading to different results.
Most studies on hospital efficiency have employed DEA, Pabón Lasso Analysis and SFA. Here, about 46% of the reviewed studies used the DEA approach. This is in line with previous international findings [25-26]. For example, Rasool et al. (2014) investigated hospital efficiency in Pakistan and obtained a mean hospital efficiency of 0.703 using DEA [25]. Also, about 14% of the articles had used SFA to measure hospital efficiency. For example, Goudarzi et al. (2014) estimated a mean hospital efficiency of 0.684 in Iranian hospitals using the SFA approach [26]. In addition, about 18% of the studies employed Pabón Lasso Analysis. For example, Younsi (2014) used this approach in his study on 40 Tunisian hospitals and reported that 27.5% had a high level of efficiency (high BOR and high BTR) [16]. In the present review, mean TE values obtained through DEA, DEA plus Pabón Lasso Analysis, Pabón Lasso Analysis, and SFA were 0.947, 0.960, 0.860, and 0.713 respectively. Therefore, based on the approach used to measure hospital efficiency, the calculation can yield different results. However, differences in measurement may not only be due due to differences in methodology. Indeed, methodology can affect results, and inputs and outputs can as well. As an example, a study in Iran showed that DEA was the dominant method of measurement of efficiency Iranian hospitals. The results of this articles demonstrated that the ability to handle multiple inputs and outputs in different units of measurement was the main explanation for using DEA as the dominant method of measurement [27].
The measurement of hospital efficiency is done through a set of input and output variables. The present findings show that the most commonly used input variables used in studies on hospital efficiency in the EMR are the number of employees and the number of beds, while the most commonly used output variables are the number of inpatient admissions, the number of inpatient days, BOR, BTR, and ALS. For example, in a study on hospital efficiency in Oman, Ramanathan (2005) used out‐patient visits, in‐patient services, and surgical operations as outputs, and the number of beds and manpower as inputs [10]. In addition, some studies have used other inputs such as work hours, non-labor costs (i.e. equipment, food, drugs), and the area of the hospital in cubic meters [28-29], and outputs such as mortality rate, number of nursing students, number of medical students, number of nursing and medical training weeks, and number of scientific publications [30-31]. Researchers must use more input and output variables when measuring hospital efficiency to increase the accuracy of their findings.
In some countries, mean efficiency has increased significantly in recent years. For example, Helal et al. (2017) showed a significant improvement in the average efficiency of Saudi hospitals in 2014 compared to 2006, with hospital efficiency reaching 92.3% in 2014 [32].
The present review showed that, on average, small hospitals [33] and government and teaching hospitals [34] did not have a desirable level of efficiency. For example, Chaabouni and Abednnadher (2016) examined Tunisian public hospitals and showed that cost-effectiveness decreases with hospital size. They found that mean cost-effectiveness was 0.995 in large hospitals compared to 0.875 in small hospitals [33]. In a study on Iranian hospitals, Ketabi (2011) showed that CCUs in 83.3% of teaching hospitals and 60% of private hospitals perform inefficiently. This was attributed to the excess of medical equipment as well as personnel and technological capabilities. Teaching hospitals were less efficient because of bureaucratic processes and private hospitals had lower BORs [35]. There is a larger demand for care in public hospitals than private hospitals, and thus public hospital managers in particular must make optimal use of their resources [36].
The finding of the meta-regression showed that there were significant differences in measurement efficiency scores among different methods. This finding was in line with previous studies [36]. As an example, Kontodimopoulos et al. [37] showed lower efficiency scores for DEA compared with SFA, while Gannon [7] reported the opposite.
The present review showed that hospital efficiency decreases by 0.002% as the publication date increases by one year. In other words, the time sequence of studies on hospital efficiency indicates lower levels of efficiency in recent years compared to previous years.
The results of this article should, however, be cautiously interpreted. Hospital efficiency has only been studied in a limited number of Eastern Mediterranean countries. This gap in the literature indicates that the reviewed studies are not comprehensive in terms of coverage and methodology. Other variables, such as ownership or type of hospital, that can impact the measurement of hospital efficiency, but a small sample size restricts control of this variable.