The current issue and full text archive of this journal is available on Emerald Insight at:

https://www.emerald.com/insight/1463-5771.htm

Measuring the operational and The efficiency

in Egyptian
financial efficiency using a hospitals

Malmquist data envelopment

analysis: a case of 2521
Egyptian hospitals Received 27 January 2020
Revised 15 April 2020
Accepted 5 July 2020
Ahmed Mohamed Habib
Department of Accounting, Faculty of Commerce, Zagazig University,
Zagazig, Egypt, and
Tamer Mohamed Shahwan
Department of Management, Faculty of Commerce, Zagazig University,
Zagazig, Egypt

Abstract
Purpose – The efficient use of organizational resources is integral to the existence of prime firms. This study,
using Malmquist data envelopment analysis (DEA), aims to assess in the level of operational and financial
efficiency and its determinants for ensuring and sustaining excellent performance in 33 Egyptian private
hospitals.
Design/methodology/approach – This study adopted a Malmquist DEA approach to assess the changes in
operational and financial efficiency in Egyptian hospitals. Tobit regression was also used to identify the
significant variables affecting their efficiency. In addition, a sensitivity analysis is carried out for model
validations.
Findings – Out of 33 hospitals, 17 were found inefficient due to the decline in their technical efficiency.
Moreover, the total value of the software programs and operational expenses and the total number of
employees are common factors affecting both operational and financial efficiency. In addition, the number of
physicians significantly affects the hospital’s financial efficiency.
Practical implications – The study sheds light on the value of using DEA to assess efficiency. DEA in the
context of emerging economy such as Egypt’s can be a useful tool for decision-makers and practitioners in
identifying and addressing performance weaknesses and thus supports continuous improvement in
performance.
Originality/value – Several studies have adopted the DEA approach to assess the overall efficiency of
hospitals in Europe and the United States. However, in the MENA region, these studies are uncommon. This
study is thought to be one of the earliest attempts to assess hospitals’ efficiency in Egypt.
Keywords Data envelopment analysis, Decision-making units, Malmquist index, Technical efficiency,
Operational efficiency, Financial efficiency
Paper type Case study

1. Introduction
Organizations currently face many challenges in the world of business where performance
appraisal and continuous improvement, in general, have become key to overcoming such
challenges; they identify and address deficiencies and make best use of the limited resources
available to ensure performance excellence. In this context, the quality of performance of
health-care services has lately become one of the most important general concerns for every
member of society. The availability and quality of such services affect the performance of
Benchmarking: An International
both individuals and society, thus affecting even the state’s plans of sustainable economic Journal
development. Vol. 27 No. 9, 2020
pp. 2521-2536
Numerous efforts have been made worldwide to ensure that access to health service of © Emerald Publishing Limited
1463-5771
acceptable quality is confirmed to be an inherent human right and a necessary part of DOI 10.1108/BIJ-01-2020-0041
BIJ development plans. The 2030 Sustainable Development Agenda, applicable to all nations, is
27,9 of unprecedented scope and ambition and goes far beyond the Millennium Development
Goals (MDGs). Health is a priority among the Sustainable Development Goals (SDGs), which
comprise a broad range of economic, social and environmental objectives. Yet the 2018 report
of the World Health Organization (WHO) indicates weaknesses in the health system and lack
of proper health financing in many countries, which leaves major gaps in the national,
regional and global defenses against outbreaks of infectious diseases.
2522 In Egypt as an emerging economy, article 18 of the Constitution in 2014 stipulated that
citizens have the right to integrated health care according to quality standards. Moreover, not
less than 3% of the annual GNP should be directed to its health care. In addition, the annual
report of the Health Services Statistics for 2016 issued by the Egyptian Central Agency for
Public Mobilization and Statistics (CAPMAS) indicates an increase of 29.2% in the number of
patients during the period from 2007 to 2016, as well as an increase of 59.21% in the costs
incurred by the state during the same period. Such an increase is a burden on state resources,
which requires appropriate measures to be taken to reveal any deficiencies or deviations in
the performance of activities and operations. But this can be done only through a proper
assessment of the performance of hospitals and health-care units. Furthermore, the report
issued by the Ministry of Finance on the State General Budget for the fiscal year 2019–2020
also indicates a gradual increase in the value of spending on the health sector, reaching 43.9bn
Egyptian pounds for the fiscal year 2015–2016, 54.1bn Egyptian pounds for the fiscal year
2016–2017, LE 60.8bn for the fiscal year 2017–2018, LE 61.8bn for the fiscal year 2018–2019
and LE 73.1bn in the draft budget for the fiscal year 2019–2020. This progressive increase in
the governmental spending on the health sector places more responsibilities and burdens on
organizations working in this sector to ensure the efficient use of these resources, thus,
achieving the target goals of sustainable development strategy, that is, Egypt’s vision 2030.
Despite such an increase in the amount spent by the Egyptian health sector, the waiting
lists are still long. In addition, the indicators of Egypt’s health system according to WHO refer
to an increase in the mortality rate and a decrease in the density of physicians and nurses per
1,000 population. These indicators raise any doubts regarding the efficiency level of the
Egyptian health-care sector. Accordingly, health organizations try hard to check whether
their resources are being effectively used and to get reliable information about the quality of
their health-care systems. Moreover, many stakeholders associated with health services (e.g.
potential investors, patients, suppliers, creditors, governmental agencies and financial
institutions) are very keen to differentiate between efficient and inefficient health-care
providers and to determine the main drivers behind inefficiency, thus, boosting performance.
Accordingly, many approaches have been adopted in assessing how well decision-making
units (DMUs) accommodate and convert health sector inputs into outputs. One of the first
approaches to assessing DMUs’ efficiency was the use of ratio analysis as a parametric
method (Fixler et al., 2014). An alternative approach to assessing corporate performance is the
use of regression analysis to tackle the problems affecting ratio analysis (Thanassoulis, 1993).
However, the parametric nature of regression analysis restricts the usefulness of this
approach in capturing the actual function of the phenomenon under investigation (Sexton,
1986; Thanassoulis, 1993; Shahwan and Hassan, 2013). Moreover, the estimate in regression
analysis is based on average performance, not best performance (Shahwan and Kabe, 2013;
Shahwan and Hassan, 2013). Another approach is the use of stochastic frontier analysis
(SFA) where, the production function is predefined as it is in regression analysis (Fixler et al.,
2014). Unlike the models aforementioned, with their obvious drawbacks, DEA has been
widely proposed as an alternative approach for assessing the relative efficiency of DMUs
against the best observed practices (Asanduluia et al., 2014; Lo Storto, 2013; Khan et al., 2013;
Bhagavath, 2006; Vitaliano, 1998). For instance, Bowlin et al. (1984) and Thanassoulis (1993)
examine the ability of DEA to estimate the performance of hypothetical hospitals based on a
set of diverse inputs and outputs. Their empirical findings confirm the superiority of DEA to The efficiency
regression analysis for the accuracy of its estimates and for capturing the relationship. in Egyptian
Simultaneously, Arsad et al. (2011); Tuskan and Stojanovic (2016); Mousa (2015); Yu et al.
(2014); and Saeed et al. (2013) confirm the beneficial role of DEA in benchmarking DMUs for
hospitals
identifying the units that show best practice, thus ensuring continued improvement.
In the context of Egypt, the Ministry of Health and Population has adopted only
conventional methods, such as issuing guidelines and ratios for assessing hospitals’
efficiency, but still hospitals and health-care units can identify their strengths, weaknesses, 2523
threats and opportunities. Previous studies (e.g. Kohl et al., 2019; Fiallos et al., 2017; Wang
et al., 2016; Fixler et al., 2014) have indicated that, from a comprehensive perspective, such an
approach is less accurate in assessing performance, particularly the efficiency dimension.
Based on all the aforementioned, this study’s target is to address the following research
questions:
RQ1. Are there significant differences between hospitals’ operational efficiency scores
over the duration of research?
RQ2. Are there significant differences between hospitals’ financial efficiency scores over
the duration of research?
RQ3. Do the input variables under investigation have a significant impact on the
operational efficiency scores of Egyptian hospitals over the duration of research?
RQ4. Do the input variables under investigation have a significant impact on the
financial efficiency scores of Egyptian hospitals over the duration of research?
The remainder of this paper is organized as follows: the brief “Literature Review” section
reports on the current documents on assessing hospital efficiency, while the “Data and
Methodology” section reviews the sample of the study, DEA models for assessing the
operational and financial efficiency of hospitals and their relevant extensions. The “Results
and discussion” section considers the Malmquist index analysis, efficiency determinants
using tobit regression analysis, validation and sensitivity analysis (robustness tests) and
other required statistical tests, such as the Sign test, Wilcoxon test, Mann–Whitney test and
Kruskal–Wallis test. Finally, the “Conclusion” section draws some inferences about the
models and tests results and suggests some future research directions.

2. Literature review
Since excellent performance has become an integral element of hospital survival, numerous
practitioners, academicians and policymakers have been motivated to assess hospital
efficiency. According to Ozgen and Ya (2004) and Kohl et al. (2019), efficiency has been
considered a source of competitive advantage. At the same time, assessing hospital efficiency
is one of the key drivers in developing the best strategies and policies to ensure continuous
improvement (e.g. Kim and Kim, 2019; Kohl et al., 2019; Plasschaert, 2018; Sahin and Ilgun,
2019; Arsad et al., 2011; Fiallos et al., 2017; Jia and Yuan, 2017; Valdmanis et al., 2017; Sun et al.,
2016; Tuskan and Stojanovic, 2016; Nakata et al., 2016; Wang et al., 2016; Fixler et al., 2014;
Mousa, 2015; Chatfield, 2014; De Souza et al., 2014; Yu et al., 2014; Saeed et al., 2013).
Many techniques have been adopted to assess the efficiency of DMUs in accommodating
and converting their inputs into outputs. One of the first approaches to be adopted in
assessing DMUs’ efficiency was the use of ratio analysis as a parametric method (Fixler et al.,
2014). With this approach, DMUs are considered an efficient unit when the ratio equals or
exceeds 1. However, the use of either single or multiple ratios is not sufficient to determining
overall efficiency (Athanassopoulos and Ballantine, 1995; Collbert et al., 2000; Fixler et al.,
2014). Another approach to assessing corporate performance is the use of regression analysis
BIJ in order to overcome the obstacles affecting ratio analysis (Thanassoulis, 1993). However, the
27,9 parametric nature of the regression analysis restricts its usefulness in capturing the full
function of the phenomenon in question (Sexton, 1986; Thanassoulis, 1993; Shahwan and
Hassan, 2013). Moreover, the coefficient estimate in regression analysis is based on average
performance, not best performance (Shahwan and Kaba, 2013; Shahwan and Hassan, 2013). A
different approach is the use of SFA where the production function is predefined as it is in
regression analysis (Fixler et al., 2014). Unlike the previous models, with their obvious
2524 drawbacks, DEA has been widely proposed as an alternative approach for assessing the
relative efficiency of the DMUs compared with the best practices observed in general
(Asanduluia et al., 2014; Lo Storto, 2013; Khan et al., 2013; Bhagavath, 2006; Vitaliano, 1998).
For instance, Bowlin et al. (1984) and Thanassoulis (1993) examine the ability of DEA to
assess the performance of hypothetical hospitals based on sets of multiple inputs and
outputs. These authors’ empirical findings confirm the superiority of DEA to regression
analysis in its power to estimate accuracy and to capture existing relationships. At the same
time, Arsad et al. (2011); Tuskan and Stojanovic (2016); Mousa (2015); Yu et al. (2014); and
Saeed et al. (2013) confirm the beneficial role of DEA in benchmarking the DMUs for
identifying the units that engage in best practices, thus ensuring the continuity of the
improvement process.
Similarly, Kohl et al. (2019) review and analyze 262 papers on the use of DEA in health care
with a focus on hospitals during the period 2005–2016. They point out the reliability of the
DEA model in assessing hospital efficiency. Using DEA models, they ascertain the most
common variables used as inputs (e.g. beds, medical staff, nurses, nonmedical staff, overall
staff and operational expenses) and outputs (outpatients, inpatients, services and surgery).
Gandhi and Sharma (2018) apply both DEA models and a Malmquist DEA approach to assess
the financially technical efficiency of the private Indian hospitals during the period 2010–
2014. Out of 37 hospitals, 14 and 20 are financially efficient under the CCR model (Charnes et
al., 1978) and the BCC model (Banker et al., 1984) of DEA, respectively. Moreover, the
improvement of the overall productivity in these hospitals is due to technological progress in
this industry. Plasschaert (2018) states that increasing technical efficiency was the crucial
determinant of total efficiency increase in 50 hospitals in Belgium during the period 2012–
2014. Fiallos et al. (2017) use DEA for assessing the performance of physicians in a pediatric
emergency department. The results confirm that the absence of trainees increases the
average efficiency of physicians. Jia and Yuan (2017) apply DEA window analysis for
assessing a Chinese hospital’s operational efficiencies. The results indicate that the
improvement of technical efficiency was attributable to pure technical efficiency rather than
scale efficiency. Sun et al. (2016) also clarify the important role of DEA in supporting
managers and decision-makers via the optimal allocation and proper restructuring of hospital
resources. Wang et al. (2016) deploy DEA with tobit regression as a hybrid model to assess the
efficiency score of 32 hospitals in China and to identify the determinants of efficiency or
inefficiency in them. They indicate also that 40.6% of the samples were inefficient. The
results of the tobit regression also made clear the positive impact of the total annual income,
the number of discharged patients, the number of outpatients and that of emergency visits on
the estimated efficiency scores. However, the total expenditure and number of open beds
negatively affected the efficiency scores. In addition, there was no correlation between the
number of health-care workers and the technical efficiency scores. Fixler et al. (2014)
confirmed the use of DEA as a relevant analytical model for assessing the efficiency of
Canadian acute care hospitals. The results confirm the inefficiency of most of the hospitals
under investigation, which demonstrated decreasing returns to scale (DRS). Chatfield (2014)
analyzed the efficiency of 388 hospitals in the United States by means of DEA. The study
demonstrates that only 33 hospitals were efficient, the average efficiency score being equal to
0.683. De Souza et al. (2014) apply DEA analysis for assessing the financially technical
efficiency of 20 public and voluntary hospitals in Brazil for the period from 2008 to 2010. The The efficiency
findings show statistically significant differences between public and voluntary hospitals in Egyptian
with respect to financial management efficiency. As claimed by Schuhmann (2008),
measuring hospitals’ financial management efficiency would ensure the optimal allocation of
hospitals
hospitals’ resources.
Very few studies have been on DEA as applied in the hospital sector of the MENA region;
Marrakchi and Essid (2019) assess the technical efficiency of 134 Tunisian public hospitals
using DEA. Their empirical findings show that only 28% of the selected samples were 2525
technically efficient. Dhaoui (2019) assessed the technical efficiency, allocative efficiency and
the determinants of 18 health systems in MENA countries for the years 1997, 2005 and 2014.
Under an input-oriented approach, the average efficiency scores were 79, 83.6 and 78.7% in
1997, 2005 and 2014, respectively. The tobit regression results also show that private
expenditure and the control of corruption positively affect efficiency scores. Mousa and
Aldehayyat (2018) evaluate the relative efficiency of 13 Saudi regions concerning health-care
services. The Riyadh region was placed in the top rank, both the public and private sectors
being efficient. In the context of Egypt as an emerging economy, the present study to the best
of our knowledge could be considered one of the first attempts to assess the relative efficiency
of Egyptian hospitals along two dimensions: one operational and the other financial. Thus,
the study would attract the attention of managers and decision-makers who could apply the
DEA model in hospitals, where it would be more helpful for decision-makers and managers in
identifying performance weaknesses. To this end, the present study aims to test the following
hypotheses:
H1. The differences between hospitals’ operational efficiency scores over the duration of
the research are statistically significant.
H2. The differences between hospitals’ financial efficiency scores over the duration of the
research are statistically significant.
H3. All the input variables under investigation have no statistically insignificant impact
on the operational efficiency scores of Egyptian hospitals over the duration of the
research.
H4. All the input variables under investigation have no statistically insignificant impact
on the financial efficiency scores of Egyptian hospitals over the duration of the
research.

3. Data and methodology

The empirical part of the current study includes data collection, identifying the study
variables, developing the relationship between the selected inputs and outputs using the
Malmquist DEA approach in order to assess the change in total factor productivity and
finally identifying the significant factors affecting the efficiency of the selected sample, using
tobit regression.
We identified 33 private hospitals working in Egypt based on data obtained through a
series of personal visits to the CAPMAS. The collected data were coded without reference to
any identifying details of these 33 hospitals. Therefore, we refer further to the selected
hospitals as H1, H2 and so on. The collected data set is used for assessing the efficiency scores
of the selected hospitals as DMUs over the period 2014–2016.
Among the wealth of available tools for assessing the efficiency of DMUs, DEA with
Malmquist’s index was selected for assessing changes in the level of overall operational and
financial efficiency in Egyptian hospitals and exploring whether their performance had
improved over the three years from 2014 to 2016. DEA as a nonparametric method is
BIJ deployed for assessing the hospitals’ efficiency due to its distinguishing features. First, as
27,9 argued by Arsad et al. (2011), Tuskan and Stojanovic (2016), Mousa (2015), Yu et al. (2014),
Zhu (2014), Shahwan and Hassan (2013) and Saeed et al. (2013), DEA as a nonparametric
approach is highly flexible in capturing the relationship between specific inputs and outputs,
since it does not require any prior assumptions regarding the basic functional form of the
selected variables or their residuals. Moreover, DEA does not require any prior information
about the weighting that should be given to the selected input and output factors. Second,
2526 DEA is a very powerful technique for benchmarking; thus, it helps to support managers and
policymakers in the process of continuous improvement (Arsad et al., 2011; Tuskan and
Stojanovic, 2016; Mousa, 2015; Yu et al., 2014; and Saeed et al., 2013). Finally, as argued by
Cooper et al. (2000), the use of DEA is realistic, since it targets best practice among DMUs
instead of depending on average practice (the measure of central tendencies) adopted in
regression analysis.
Productivity change can be mathematically described using the Malmquist DEA
approach, as follows (Coelli, 1996; Ng, 2008; Desta, 2016):
!12
d0t ðXtþ1 ; Ytþ1 Þ d0tþ1 ðXtþ1 ; Ytþ1 Þ
M0 ðXtþ1 ; Ytþ1 ; Xt ; Yt Þ ¼ 3 tþ1 (1)
d0t ðXt ; Yt Þ d0 ðXt ; Yt Þ

where the value of m refers to the difference in productivity between two points of production
at time t and tþ1. Xt and Yt refer to the input and the output at time t, Xtþ1 and Ytþ1 refer
respectively to the input and the output at time tþ1 and d0t is the distance function assessing
the efficiency of obtaining output Y at time t in the light of input X available at time t.
It is worth noting, as stated by Coelli (1996); Ng (2008); and Desta (2016), that the calculated
score <1 refers to a decline in efficiency, whereas the calculated score >1 refers to an increase
in efficiency, while the calculated score 5 1 does not reflect any changes in efficiency.
According to Desta (2016), productivity change can also be divided into two subtypes: change
in technological efficiency and managerial (technical) change. This can be expressed in
mathematical terms as follows:
! !12
d0tþ1 ðXtþ1 ; Ytþ1 Þ d0t ðXtþ1 ; Ytþ1 Þ d0t ðXt ; Yt Þ
M0 ðXtþ1 ; Ytþ1 ; Xt ; Yt Þ ¼ 3 tþ1 3 (2)
d0t ðXt ; Yt Þ d0 ðXtþ1 ; Ytþ1 Þ d0tþ1 ðXt ; Yt Þ
! !12
d0tþ1 ðXtþ1 ;Ytþ1 Þ d0t ðXtþ1 ;Ytþ1 Þ dt ðX ;Y Þ
where d0t ðXt ;Yt Þ
refers to the change in technical efficiency. d0tþ1 ðXtþ1 ;Ytþ1 Þ
3dtþ1
0 t t
ðX ;Y Þ
0 t t

defines the change in technological efficiency.

Regarding the identification and selection of DEA inputs and outputs, Table 1
summarizes the factors that were deployed in assessing the hospital’s efficiency.
Following the literature (see aforementioned), the DEA models in this study adopt seven
common multiple inputs for assessing both operational and financial efficiency. These input
variables comprise the total numbers for medical doctors, nurses, other personnel, software
programs beds, assets and operational expenses. To assess operational efficiency, namely in
model 1, four multiple outputs were chosen: totals of inpatients, outpatients, surgical
procedures and revenues. To assess financial efficiency, namely in model 2, the return on
assets (ROA) and the return on equity (ROE) were chosen as output indicators.
When a hospital’s efficiency score and efficiency changes had been estimated using the
Malmquist DEA technique, we conducted a tobit regression in the second stage to identify
the critical factors affecting the efficiency scores. Previous studies support the superiority of
the tobit regression as a useful tool for examining the relationship between efficiency scores
and their drivers (Tobin, 1958; Plasschaert, 2018; Wang et al., 2016). Accordingly, the overall
Category Variable Definition Studies
The efficiency
in Egyptian
Inputs (used in Medical Total number of medical doctors (e.g. Kim and Kim, 2019; Kohl hospitals
both model 1 and doctors at the end of the year et al., 2019; Plasschaert, 2018;
model 2) Nurses Total number of nurses at the end Sahin and Ilgun, 2019; Sun et al.,
of the year 2016; Nakata et al., 2016)
Other Total number of employees at the
personnel end of the year, excluding doctors 2527
and nurses
Software Total value of software programs
programs at the end of the year
Beds Number of available beds at the
end of the year
Assets Total value of assets at the end of
the year
Operational Total of operational expenses by
expenses the end of the year
Outputs (model 1) Inpatients Number of patients discharged (e.g. Kim and Kim, 2019; Kohl
for assessing after hospitalization (for various et al., 2019; Plasschaert, 2018;
operational reasons) who remained alive until Sahin and Ilgun, 2019; Sun et al.,
efficiency the end of the year 2016; Nakata et al., 2016)
Outpatients Number of patients attending for
outpatient appointments,
diagnostic services and hospital
emergencies in the course of the
year
Surgery Total number of surgical
operations in the course of the
year
Revenues Total revenues by the end of the
year
Outputs (model 2) ROA Net profits at the end of the year (e.g. Kohl et al., 2019; De Souza
for assessing divided by total assets at the end et al., 2014)
financial efficiency of the same year
ROE Net profits at the end of the year Table 1.
divided by the total owners’ Inputs and outputs in
equity at the end of the same year DEA models

relative efficiency scores computed by the DEA model were applied to the set of independent
variables (input variables of DEA models), that is, the total numbers of medical doctors,
nurses, other personnel, software programs, beds, assets and operational expenses.
Accordingly, model 1 and model 2 defined by equations (3) and (4) were adopted to test H3
and H4, respectively. Mathematically, these could be rewritten as follows:
yi ¼ β0 þ β1 x1 þ β2 x2 þ β3 x3 þ β4 x4 þ β5 x5 þ β6 x6 þ β7 x7 þ εi (3)

where the value of yi refers to the operational efficiency score according to the Malmquist
DEA model. β0 refers to the constant of the tobit regression model; βi represents the tobit
regression coefficients of the explanatory variables; X1 ; X2 ; X3 ; X4 ; X5 ; X6 ; and X7 refer to the
total numbers of medical doctors, nurses, other personnel, software programs, beds, assets
and operational expenses, respectively, and εi refers to error value.
zi ¼ β0 þ β1 x1 þ β2 x2 þ β3 x3 þ β4 x4 þ β5 x5 þ β6 x6 þ β7 x7 þ εi (4)
BIJ where the value of zi refers to the financial efficiency score according to the Malmquist DEA
27,9 model. All the variables were as described earlier.
Table 2 displays the descriptive statistics of the data set showing the mean, minimum,
maximum and standard deviation.
In the data set, the average of medical doctors was about 25 in a range of 3–60.
Simultaneously, the number of medical doctors and nurses had a standard deviation of 13.179
physicians and 8.134 nurses, respectively. Thus, the selected sample for the two variables
2528 was homogeneous. Moreover, the average of other employees was about 10 and the standard
deviation of this variable was about 4%. This also supports the homogeneity of the data set.
The average for software programs was about 494,778 (£) in a range of 39,000–4307,000 (£).
The average number of beds was about 30 in a range of 9–59. The assets average was about
4,874,829 (£) in a range of 769,000–24,641,000 (£). The operational expenses average was
about 2,694,647 (£) in a range of 150,000–24,757,000 (£). The inpatients’ average was about
2,240 in a range of 151–9032. The outpatients’ average was about 12,221 in a range of 2277–
45,911. The average of surgical procedures was about 4,946 in a range of 519–34,624. The
revenues average was about 3,124,172 (£) in a range of 322,000–22,089,000 (£). The ROA
average was about 0.06 in a range of 0.027–0.26. The ROE average was about 0.15 in a range
of 0.7–0.42. Accordingly, the sample of 33 private hospitals can be considered homogeneous.

4. Results and discussion

4.1 Results of the DEA models
Table 3, panel A, based on model (1), reveals the Malmquist index results of the annual mean
of the efficiency changes in respect to operational efficiency for 2015, 2016 and the overall
period (2014–2016). The Malmquist Index results were estimated using DEAP software
version 2.1. The findings show an increase in operational efficiency for the first period (2014–
2015), with a decline in efficiency for the second period (2015–2016), while the overall results
for the period (2014–2016) indicate a decline in efficiency. It should be noted that the
Malmquist index does not apply to the first year (2014) because it was used as the base year
for the second year (2015). The Malmquist index annual means for year 2 (2014–2015) indicate
that technological changes were the key source of the increasing efficiency of the total factor
productivity (TFP) index. However, reducing the efficiency of the TFP of the selected
hospitals during (2015–2016) and (2014–2016) can be attributed to retrogressive changes in
the technical or managerial efficiency. The findings are compatible with those in previous
empirical evidence (e.g. Plasschaert, 2018; Sahin and Ilgun, 2019; Fiallos et al., 2017; Jia and
Yuan, 2017; Valdmanis et al., 2017; Sun et al., 2016; Nakata et al., 2016; Wang et al., 2016; Fixler

Variables Min Max Mean Std. Dev

Medical doctors 3 60 25.09 13.179

Nurses 2 41 13.33 8.134
Other personnel 2 23 9.76 4.091
Software programs (£) 39,000 4,307,000 494,777.8 712,431.1
Beds 9 59 29.59 11.516
Assets (£) 769,000 24,641,000 4,874,828.8 4,967,142.7
Table 2. Operational expenses (£) 150,000 24,757,000 2,694,646.5 4,472,076.9
Descriptive summary Inpatients 151 9032 2239.7 1,604.994
statistics of variables Outpatients 2277 45,911 12,221.4 8463.812
used to assess the Surgery 519 34,624 4945.8 5012.981
efficiency of the Revenues (£) 322,000 22,089,000 3,124,171.7 3,945,099.7
Egyptian private ROA 0.02651 0.25962 0.06388 0.03566
hospitals ROE 0.06864 0.42155 0.15433 0.07665
Panel A: Malmquist index summary of annual means (model 1)
The efficiency
Year effch techch tfpch in Egyptian
(2014–2015) 0.998 1.004 1.002 hospitals
(2015–2016) 0.934 1.067 0.996
(2014–2016) 0.965 1.035 0.999
Panel B: Top ten hospitals with tfpch >1 in rank sequence
DMUs effch techch tfpch Rank 2529
H7 1.000 1.186 1.186 6
H9 1.038 1.243 1.290 5
H10 1.111 0.987 1.097 10
H12 1.964 0.880 1.729 1
H13 1.213 1.066 1.293 4
H14 1.065 1.038 1.106 7
H17 1.073 1.229 1.318 3
H21 1.116 0.989 1.103 8
H28 1.061 1.037 1.101 9
H30 1.789 0.855 1.530 2
Panel C: Test results of the difference using Wilcoxon and sign tests
Table 3.
Total factor change Wilcoxon test Sign test The Malmquist index
Z-statistic p-value Z-statistic p-value summary, Wilcoxon
(tfpch2014/2015 vs tfpch2015/2016) 0.295 0.768 0.348 0.728 and sign test results for
Note(s): effch 5 Technical efficiency change; techch 5 Technological change; and tfpch 5 Total factor operational efficiency
productivity (TFP) change changes

et al., 2014; Chatfield, 2014). They suggest that “on average” DMUs do not operate at the
optimum level of operational TE. Further, all the previous empirical evidence clarifies the
important role of DEA in supporting managers and decision-makers through the optimum
allocation and restructuring of resources.
Table 3, panel B, shows the top ten hospitals for operational efficiency with (tfpch > 1)
over the period (2014–2016). The DMU 12 achieved the best results in terms of improvement
(1.729), followed by DMU 30 with DMU 17 and so on. The increases in the TFP of these top ten
hospitals can be linked to technical changes or changes in the managerial efficiency of
converting inputs to outputs as opposed to changes in technological efficiency. To reinforce
the aforementioned results, Table 3, panel C, shows some complementary statistical tests
based on the Wilcoxon test and sign test (using IBM SPSS Version 25) that were carried out to
verify the statistically significant differences between hospitals’ scores for operational
efficiency over the period 2014–2015. No statistically significant evidence of improvement in
operational performance at the 5% significance level was found.
Table 4, panel A, based on model 2 reveals the Malmquist index results of the annual mean
of the efficiency changes in respect to financial efficiency for 2015, 2016 and the overall period
(2014–2016).
The Malmquist index results show a decline in the financial efficiency of hospitals for the
first period (2014–2015) as well as the second period (2015–2016), and the overall results for
the period 2014–2016 also indicate a decline in the financial efficiency. The Malmquist index
annual means for the second year (2014–2015), the third year (2015–2016) and the overall
period (2014–2016) indicate that technical or managerial efficiency changes were the key
source of the declining efficiency of the TFP. The findings are compatible with previous
empirical evidence (e.g. Kim and Kim, 2019; Plasschaert, 2018; Arsad et al., 2011; Sun et al.,
2016; Tuskan and Stojanovic, 2016; Nakata et al., 2016; Wang et al., 2016; Fixler et al., 2014;
Mousa, 2015; Yu et al., 2014; Saeed et al., 2013). It suggests that “on average” DMUs do not
operate at the optimum level of operational TE. Further, all the previous empirical evidence
BIJ Panel A: Malmquist index summary of annual means (model 2)
27,9 Year effch techch tfpch
(2014–2015) 0.941 1.056 0.993
(2015–2016) 0.879 1.132 0.995
(2014–2016) 0.909 1.093 0.994
Panel B: Top ten hospitals with tfpch >1 in rank sequence
2530 DMUs effch techch tfpch Rank
H7 0.963 1.122 1.080 10
H9 1.000 1.222 1.222 4
H10 1.075 1.160 1.247 2
H12 0.955 1.133 1.082 9
H13 1.274 1.083 1.379 1
H14 1.000 1.173 1.173 6
H17 1.000 1.146 1.146 7
H21 1.020 1.160 1.183 5
H28 1.224 1.012 1.239 3
H30 1.027 1.077 1.106 8
Panel C: Test results of the difference using Wilcoxon and sign tests
Table 4.
The Malmquist index Total factor change Wilcoxon test Sign test
summary, Wilcoxon Z-statistic p-value Z-statistic p-value
and sign test results for (tfpch 2014/2015 vs tfpch 2015/2016) 0.679 0.497 1.044 0.296
financial efficiency Note(s): effch 5 Technical efficiency change; techch 5 Technological change; and tfpch 5 Total factor
changes productivity (TFP) change

clarifies the important role of DEA in supporting managers and decision-makers through the
optimum allocation and restructuring of resources.
Table 4, panel B, shows the financial efficiency of the top ten hospitals with (tfpch > 1)
over the period 2014–2016. DMU 12 achieved the best results in terms of improvement (1.379),
followed by DMU 10 with DMU 30 and so on. The increases in the TFP of these top ten
hospitals can be linked to technological efficiency changes as opposed to changes in technical
or managerial efficiency in converting the inputs to outputs. Table 4, panel C, shows the
findings of the Wilcoxon test and sign test, which indicate insignificant evidence of
improvement in the financial performance of Egyptian hospitals at the 5% significance level.
4.2 Efficiency drivers
The tobit regression analysis was applied to identify the critical factors affecting the
operational and financial efficiency scores. Table 5 reveals the tobit regression analysis to
identify the influence of the input variables mentioned earlier on both Egyptian private
hospitals’ operational and financial efficiency scores during the period (2014–2016). On the
one hand, Table 5, panel A, summarizes the operational efficiency drivers of these hospitals
using tobit regression analysis. The overall operational efficiency scores of these hospitals
are negatively associated with the total employees’ number after the exclusion of physicians
and nurses (other personnel) and the total value of software variables at 0.01 significance
level. On the other side, operational expenses have a significant positive influence on the
operational efficiency scores. Simultaneously, the other variables including the number of
both medical doctors and nurses, the number of available beds and the total value of assets
have insignificant effects on the overall operational efficiency scores of the data set.
Accordingly, H3 is partially supported. The findings are compatible with the previous
empirical evidence (e.g. Plasschaert, 2018; Wang et al., 2016). They point that identifying the
most significant variables may greatly assist managers and policymakers in taking
appropriate corrective decisions to improve efficiency.
Panel A: Operational efficiency model Panel B: Financial efficiency model
The efficiency
(model 1) (model 2) in Egyptian
Variables Coef Std. Err z p>z Coef Std. Err z p>z hospitals
Medical doctors 0.00067 0.00568 1.18 0.242 0.01447 0.00447 3.24 0.002***
Nurses 0.01379 0.00919 1.50 0.137 0.00228 0.00749 0.30 0.762
Other personnel 0.03196 0.01131 2.83 0.006*** 0.02314 0.00927 2.49 0.014**
Software 0.00075 0.00014 5.52 0.000*** 0.00045 0.00011 4.19 0.000*** 2531
programs
Beds 0.00048 0.00372 0.13 0.898 0.00289 0.00279 1.04 0.303
Assets 0.00002 0.00002 1.50 0.137 0.00001 0.00001 0.96 0.340
Operational 0.00018 0.00003 5.40 0.000*** 0.00009 0.00002 4.52 0.000***
expenses
Constant 1.36620 0.11248 12.15 0.000*** 1.24222 0.08759 14.18 0.000***
Table 5.
Pseudo R2 0.4942 0.6639 The tobit regression
χ2 61.29 0.0000*** 73.33 0.0000*** results for operational
N 99 99 and financial efficiency
Note(s): *, ** and *** are significant at the 10, 5 and 1% levels, respectively models

Regarding the overall financial efficiency drivers, panel B of Table 5 shows that the number
of medical doctors and the total value of software programs have a significant impact on the
overall financial efficiency scores of the selected sample at the 0.01 significance level.
Moreover, the total number of employees (all personnel except physicians and nurses)
negatively affects the overall financial efficiency scores at the 0.05 significance level, while
the total amount of operational expenses positively affects the financial efficiency scores at
the 0.01 significance level. The other variables, including the number of nurses, the number of
available beds and the total value of assets for each hospital, are not significant. Accordingly,
H4 is also partially supported. The findings are compatible with the previous empirical
evidence (e.g. Kohl et al., 2019; Wang et al., 2016; De Souza et al., 2014). They point out that
identifying the most significant determinants of hospitals’ financial efficiency is critical to
managers and policymakers in order to develop the best strategies and policies concerning
with the continuous improvement as an essential pillar of total quality management.

4.3 Robustness tests

A sensitivity analysis was carried out to examine the internal and external validity of the
findings (Fixler et al., 2014; Parkin and Hollingsworth, 1997). The internal validity test was
designed to compare the results obtained with the modified models using different choices of
variables. In the current study, the output variables of the basic models were removed from
the model in sequence, for study purposes. The correlation and coherence between the basic
model and other modified models are summarized in Table 6.
Using the Mann–Whitney U test, the resulting DEA scores were compared with the
original scores to verify whether the removal of the variable resulted in a significant
difference in efficiency scores. The Spearman rank correlations were also calculated. Table 6,
panel A, shows that the removal of outpatients from the operational efficiency model had a
significant impact on the results at the 5% level. Moreover, the general distribution of the
efficiency scores went down from 0.792 to 0.709 and the percentage of efficient DMUs from 45
to 33%. In addition, the high Spearman rank correlation suggests that the hospital rankings
were not significantly altered.
Table 6, panel B, indicates that the removal of ROA and ROE from the financial efficiency
model had a significant impact on the results at the 1 and 10% levels, respectively. Moreover,
the general distribution of the efficiency scores went down from 0.549 to 0.474 and 0.498
BIJ Panel A: Operational efficiency model (model 1)
27,9 Variable Average Efficient Asymp. Asymp. Spearman rank
removed efficiency score hospitals (%) sig. (Mann– sig. (Kruskal– correlation sig
Whitney U) Wallis H)
None 0.792 %45 – – –
Inpatients 0.736 %39 0.174 – 0.924***
Outpatients 0.709 %33 0.024** – 0.885***
2532 Surgical ops 0.790 %43 0.882 – 0.989***
Revenues 0.769 %38 0.426 – 0.954***
– – – – 0.138 –
Panel A: Financial efficiency model (model 2)
Variable Average Efficient p-value Asymp. Spearman rank
Removed efficiency score hospitals (%) (Mann– Sig. (Kruskal– correlation sig
Whitney) Wallis H)
None 0.549 %23 – – –
ROA 0.474 %16 0.009*** – 0.908***
Table 6. ROE 0.498 %21 0.052* – 0.922***
Sensitivity analysis of – – – – 0.028**
the two DEA models Note(s): *, ** and *** are significant at the 10, 5 and 1% levels, respectively

respectively and the percentage of efficient DMUs from 23 to 16% and 21% respectively. In
addition, the high Spearman rank correlation suggests that the hospital rankings were not
significantly altered. Finally, the external validity was assessed by looking at the consistency
of the results over time, as illustrated in Table 7.
In this analysis, the efficiency scores of 2014 were compared to the efficiency scores in
2005 and 2016. Table 7, panels A and B indicate that the distributions of both operational and
financial efficiency scores were equal because the values of the Mann–Whitney U test and the
Kruskal–Wallis H test were insignificant during the period of study. Moreover, the high
Spearman rank correlation suggests that the hospital rankings did not significantly alter
during the period 2014–2015. Thus, the general distribution of the efficiency scores does not
appear to vary much from year to year, either.

Asymp. sig. (Mann– Asymp. sig. (Kruskal– Spearman rank correlation

Year Whitney U) Wallis H) sig

Panel A: Operational efficiency model (model 1)

(2014–2015) 0.941 – 0.986***
(2015–2016) 0.560 – 0.435**
(2014–2016) 0.514 – 0.458**
(2014, 2015 and – 0.771 –
2016)
Panel B: Financial efficiency model (model 2)
(2014–2015) 0.670 – 0.986***
Table 7. (2015–2016) 0.974 – 0.929***
The distribution (2014–2016) 0.704 – 0.909***
variance of efficiency (2014, 2015 and – 0.897 –
scores over the study 2016)
period Note(s): *, ** and *** are significant at the 10, 5 and 1% levels, respectively
5. Conclusions The efficiency
This study focused on employing DEA methodology to assess the operational and financial in Egyptian
efficiency of 33 private hospitals in Egypt, using data provided by CAPMAS. Given the rarity
of implementing DEA methodology to assess efficiency in the Egyptian context in general
hospitals
and in the hospital sector in particular, the aim of the study was to consider DEA as an
analytical tool with which hospitals could assess their current efficiency levels and also to
illustrate the usefulness of DEA in these circumstances. The hospital operational and
financial efficiency model inputs included totals of medical doctors, nurses, other personnel, 2533
software programs, beds, assets and operational expenses. Outputs measured by the
operational efficiency model included inpatients, outpatients, surgical procedures and
revenues. Outputs measured by the financial efficiency model included ROA and ROE.
The findings of Malmquist analysis showed that hospitals were declining in operational
and financial efficiency during the period of the study because of retrogression in technical or
managerial efficiency. Moreover, the Wilcoxon test and sign test results did not detect any
statistically significant evidence of improvement in operational and financial efficiency. The
models were able to identify the best hospitals regarding operational and financial efficiency,
which is ideal for the establishment of best practices. The tobit regression analysis showed
that software programs, operational expenses and other personnel were the most significant
variables affecting the operational efficiency model. Moreover, the operational expenses,
software programs, medical doctors and other personnel respectively were the most
significant variables affecting the financial efficiency model. Extensive model validation and
sensitivity analyses were also carried out.
The results generally showed that the ineffective performance of the hospitals resulted
from declining technical or managerial efficiency, which may result from the excessive use of
inputs or the inability from current inputs to maximize outputs to their fullest extent. This
requires managers and policymakers in Egyptian hospitals to take the appropriate steps to
prevent the waste of resources and to maximize outputs, which positively reflects the
efficiency in hospitals. Accordingly, the results of this study are of great importance for
managers and policymakers in Egyptian hospitals, if they wish to raise the level of efficiency
and engage in best practices. Moreover, DEA is particularly useful for improving the current
level of efficiency in hospitals where such change is needed, as well as in successful hospitals
that may be prime candidates for exemplifying guidelines in good practice.
Future researchers would do well to aim at applying the DEA approach in other fields,
such as banking, industry, tourism, education and so on. In particular, it should bring to the
attention of policymakers, managers and analysts in health care the advantages of DEA as a
tool for performance assessing and benchmarking. In our opinion, there are three conditions
for achieving this. The first is that the model variables (causes) must be very closely related to
the objectives of the evaluation (results) and defined precisely according to the theories and
the previous studies. The second that it must be seen as an attempt to make a comparative
study between the results of DEA as a nonparametric technique for assessing performance
and other approaches such as ratio approach, regression, the analysis of stochastic frontiers,
artificial neural networks and so on. The third is close communication with hospital
managers, particularly after analyses have been undertaken.

References
Arsad, R., Shaari, S.N.M. and Isa, Z. (2011), “Comparative study on DuPont analysis and DEA models
for measuring stock performance using financial ratio”, Proceedings of the 13th IMT-GT
International Conference on Mathematics, Statistics and their Applications (ICMSA2017),
Universiti Utara Malaysia, Sintok, Kedah.
BIJ Asanduluia, L., . . . Roman, M. and Fatulescu, P. (2014), “The efficiency of healthcare systems in
Europe: a data envelopment analysis approach”, Procedia Economics and Finance, Vol. 10,
27,9 pp. 261-268.
Athanassopoulos, A.D. and Ballantine, J.A. (1195), “Ratio and frontier analysis for assessing corporate
performance: evidence from the grocery industry in the UK”, Journal of the Operational
Research Society, Vol. 46 No. 4, pp. 427-440.
Banker, R., Charnes, A. and Cooper, W.W. (1984), “Some models for estimating technical and scale
2534 inefficiencies in data envelopment analysis”, Management Science, Vol. 30 No. 9, pp. 1078-1092.
Bhagavath, V. (2006), “Technical efficiency measurement by data envelopment analysis: an
application in transportation”, Alliance Journal of Business Research, Vol. 2 No. 1, pp. 60-72.
Bowlin, W.F., Charnes, A., Cooper, W.W. and Sherman, H.D. (1984), “Data envelopment analysis and
regression approaches to efficiency estimation and evaluation”, Annals of Operations Research,
Vol. 2 No. 1, pp. 113-138.
Charnes, A., Cooper, W.W. and Rhodes, E. (1978), “Measuring the efficiency of decision-making units”,
European Journal of Operational Research, Vol. 2 No. 6, pp. 429-444.
Chatfield, J.S. (2014), “Data envelopment analysis comparison of hospital efficiency and quality”,
International Journal of Management Accounting Research, Vol. 4 No. 1, pp. 93-109.
Coelli, T. (1996), “A guide to DEAP Version 2.1, a data envelopment analysis (computer) program”,
University of New England, CEPA Working Paper No.08.
Collbert, A., Levary, R.R. and Shaner, M.C. (2000), “Determining the relative efficiency of MBA
programs using DEA”, European Journal of Operational Research, Vol. 125 No. 3, pp. 656-669.
Cooper, W.W., Seiford, L.M. and Tone, K. (2000), Data Envelopment Analysis: A Comprehensive Text
with Models, Applications, References and DEA-Solver Software, Kluwer Academic Publishers,
Boston.
De Souza, A.A., Moreira, D.R., Avelar, E.A., Marques, A.M.F. and Lara, A.L. (2014), “Data envelopment
analysis of efficiency in hospital organisations”, International Journal of Business Innovation
and Research, Vol. 8 No. 3, pp. 316-332.
Desta, T.S. (2016), “Are the best African banks really the best? A Malmquist data envelopment
analysis”, Meditari Accountancy Research, Vol. 24 No. 4, pp. 588-610.
Dhaoui, I. (2019), “Healthcare system efficiency and its determinants: a two-stage Data Envelopment
Analysis (DEA) from MENA countries”, Economic Research Forum, pp. 1-48, Working Paper
No. 1320.
Fiallos, J., Patrick, J., Michalowski, W. and Farion, K. (2017), “Using data envelopment analysis for
assessing the performance of pediatric emergency department physicians”, Health Care
Management Science, Vol. 20 No. 1, pp. 129-140.
Fixler, T., Paradi, J.C. and Yang, X. (2014), “A data envelopment analysis approach for measuring the
efficiency of Canadian acute care hospitals”, Health Services Management Research, Vol. 27
Nos 3-4, pp. 57-69.
Gandhi, A.V. and Sharma, D. (2018), “Technical efficiency of private sector hospitals in India using
data envelopment analysis”, Benchmarking: An International Journal, Vol. 25 No. 9,
pp. 3570-3591.
Jia, T. and Yuan, H. (2017), “The application of DEA (Data Envelopment Analysis) window analysis in
the assessment of influence on operational efficiencies after the establishment of branched
hospitals”, BMC Health Services Research, Vol. 17 No. 265, pp. 1-8.
Khan, K., Turner, J.R. and Maqsood, T. (2013), “Factors that influence the success of public sector
projects in Pakistan”, IRNOP Conference, June 17-19, BI Norwegian Business School, Oslo.
Kim, C. and Kim, H.J. (2019), “A study on healthcare supply chain management efficiency: using
bootstrap data envelopment analysis”, Health Care Management Science, Vol. 22 No. 3,
pp. 534-548.
Kohl, S., Schoenfelder, J., Fugener, A. and Brunner, J.O. (2019), “The use of Data Envelopment The efficiency
Analysis (DEA) in healthcare with a focus on hospitals”, Health Care Management Science,
Vol. 22 No. 2, pp. 245-286. in Egyptian
Lo Storto, C. (2013), “Evaluating technical efficiency of Italian major municipalities: a data
hospitals
envelopment analysis model”, Procedia - Social and Behavioral Sciences, Vol. 81, pp. 346-350.
Marrakchi, M.M. and Essid, H. (2019), “Efficiency assessment of Tunisian public hospitals using data
envelopment analysis (DEA)”, Economic Research Forum, pp. 245-286, Working Paper
No. 1291. 2535
Mousa, W. and Aldehayyat, J.S. (2018), “Regional efficiency of healthcare services in Saudi Arabia”,
Middle East Development Journal, Vol. 10 No. 1, pp. 152-174.
Mousa, G. (2015), “Financial ratios versus data envelopment analysis: the efficiency assessment of
banking sector in Bahrain bourse”, International Journal of Business and Statistical Analysis,
Vol. 2 No. 2, pp. 75-84.
Nakata, Y., Yoshimura, T., Watanabe, Y., Otake, H., Oiso, G. and Sawa, T. (2016), “Efficiency of
inpatient orthopedic surgery in Japan: a medical claims database Analysis”, International
Journal of Health Care Quality Assurance, Vol. 30 No. 6, pp. 506-515.
Ng, Y.C. (2008), “The productive efficiency of the health care sector of China”, The Review of Regional
Studies, Vol. 38 No. 3, pp. 381-393.
Ozgen, H. and Ozcan, Y.A. (2004), “Longitudinal analysis of efficiency in multiple output dialysis
markets”, Health Care Management Science, Vol. 7 No. 4, pp. 253-261.
Parkin, D. and Hollingsworth, B. (1997), “Measuring production efficiency of acute hospitals in
Scotland, 1991-1994: validity issues in data envelopment analysis”, Applied Economics, Vol. 29
No. 11, pp. 1425-1434.
Plasschaert, B. (2018), “Efficiency in Flemish Hospitals: A DEA Approach”, Faculty of Economics and
Business Administration, Gent University, Gent, Master’s thesis.
Saeed, S., Ali, F., Adeeb, B. and Hamid, M. (2013), “Examining efficiency of islamic and conventional
banks in Pakistan: using data envelopment analysis”, Global Journal of Management and
Business Research Finance, Vol. 13 No. 10, pp. 24-34.
Sahin, B. and Ilgun, G. (2019), “Assessment of the impact of public hospital associations (PHAs) on the
efficiency of hospitals under the ministry of health in Turkey with data envelopment analysis”,
Health Care Management Science, Vol. 22 No. 3, pp. 437-446.
Schuhmann, T.M. (2008), “Hospital financial performance: trends to watch”, Healthcare Financial
Management, Vol. 62 No. 7, pp. 59-66.
Sexton, T.R. (1986), “The methodology of data envelopment analysis”, in Silkman, R.H. (Ed.),
Measuring Efficiency: An Assessment of Data Envelopment Analysis, New Directions for
Program Evaluation no. 52, Jossey-Bass, San Francisco, Vol. 1986, pp. 7-29, No. 32.
Shahwan, T.M. and Hassan, Y.M. (2013), “Efficiency analysis of UAE banks using data envelopment
analysis”, Journal of Economic and Administrative Sciences, Vol. 29 No. 1, pp. 4-20.
Shahwan, T.M. and Kaba, A. (2013), “Efficiency analysis of GCC academic libraries: an application of
data envelopment analysis”, Performance Measurement and Metrics, Vol. 14 No. 3, pp. 197-210.
Sun, B., Zhang, L., Yang, W., Zhang, J., Luo, D. and Han, C. (2016), “Data envelopment analysis on
evaluating the efficiency of public hospitals in tianjin, China”, Transactions of Tianjin
University, Vol. 22 No. 2, pp. 182-188.
Thanassoulis, E. (1993), “A comparison of regression analysis and data envelopment analysis as
alternative methods for performance assessments”, Journal of the Operational Research Society,
Vol. 44 No. 11, pp. 1129-1144.
Tobin, J. (1958), “Estimation of relationships for limited dependent variables”, Econometrica, Vol. 26
No. 1, pp. 24-36.
BIJ Tuskan, B. and Stojanovic, A. (2016), “Measurement of cost efficiency in the European banking
industry”, Croatian Operational Research Review, Vol. 7 No. 1, pp. 47-66.
27,9
Valdmanis, V., Rosko, M., Mancuso, P., Tavakoli, M. and Farrar, S. (2017), “Measuring performance
change in Scottish hospitals: a Malmquist and times-series approach”, Health Services and
Outcomes Research Methodology, Vol. 17 No. 2, pp. 113-126.
Vitaliano, D.F. (1998), “Assessing public library efficiency using data envelopment analysis”, Annals
of Public and Cooperative Economics, Vol. 69 No. 1, pp. 107-122.
2536
Wang, x., Luo, H., Qin, X., Feng, J., Gao, H. and Feng, Q. (2016), “Evaluation of performance and
impacts of maternal and child health hospital services using Data Envelopment Analysis in
Guangxi Zhuang Autonomous Region, China: a comparison study among poverty and non-
poverty county level hospitals”, International Journal for Equity in Health, Vol. 15
No. 131, pp. 1-6.
Yu, Y., Barros, A., Tsai, C. and Liao, K. (2014), “A comparison of ratios and data envelopment analysis:
efficiency assessment of taiwan public listed companies”, International Journal of Academic
Research in Accounting, Finance and Management Sciences, Vol. 4 No. 1, pp. 212-219.
Zhu, J. (2014), Quantitative Models for Performance Evaluation and Benchmarking: Data Envelopment
Analysis with Spreadsheets, 3rd ed., Springer, New York.

Further reading
World Health Organization (WHO) (2018a), Framework for Health Information Systems and Core
Indicators for Monitoring Health Situation and Health System Performance, WHO Press, WHO
Regional Office for the Eastern Mediterranean, Cairo.
World Health Organization (WHO) (2018b), Global Reference List of 100 Core Health Indicators - Plus
Health-Related SDGs, WHO Press, World Health Organization Press, Geneva.
World Health Organization (WHO) (2018c), World Health Statistics 2018: Monitoring Health for the
SDGs, WHO Press, World Health Organization Press, Geneva.

Corresponding author
Tamer Mohamed Shahwan can be contacted at: [email protected]

For instructions on how to order reprints of this article, please visit our website:
www.emeraldgrouppublishing.com/licensing/reprints.htm
Or contact us for further details: [email protected]