Technical efficiency of selected hospitals in Eastern Ethiopia

Abstract

This study examines the relative technical efficiency of 12 hospitals in Eastern Ethiopia. Using six-year-round panel data for the period between 2007/08 and 2012/13, this study examines the technical efficiency, total factor productivity, and determinants of the technical inefficiency of hospitals. Data envelopment analysis (DEA) and DEA- based Malmquist productivity index used to estimate relative technical efficiency, scale efficiency, and total factor productivity index of hospitals. Tobit model used to examine the determinants of the technical inefficiency of hospitals. The DEA Variable Returns to Scale (VRS) estimate indicated that 6 (50%), 5 (42%), 3 (25%), 3 (25%), 4 (33%), and 3 (25%) of the hospitals were technically inefficient while 9 (75%), 9 (75%), 7 (58%), 7 (58%), 7 (58%) and 8 (67%) of hospitals were scale inefficient between 2007/08 and 2012/13, respectively. On average, Malmquist Total Factor Productivity (MTFP) of the hospitals decreased by 3.6% over the panel period. The Tobit model shows that teaching hospital is less efficiency than other hospitals. The Tobit regression model further shows that medical doctor to total staff ratio, the proportion of outpatient visit to inpatient days, and the proportion of inpatients treated per medical doctor were negatively related with technical inefficiency of hospitals. Hence, policy interventions that help utilize excess capacity of hospitals, increase doctor to other staff ratio, and standardize number of inpatients treated per doctor would contribute to the improvement of the technical efficiency of hospitals.

Measuring Technical Efficiency of Hospitals Using Data Envelopment Analysis: A Review

“…The number of DUMs among the included studies was different. In some studies, A study conducted in Ethiopia to measure technical efficiency for 12 public and private hospitals was employed BCC, VRS, output-oriented and the total combination number of inputs and outputs was 6 (Ali, Debela, & Bamud, 2017). Another study conducted in China to measure technical efficiency for 12 county level hospitals was used CCR, CRS, input-oriented and the total combination number of inputs and outputs was 7 (N. Li, Wang, Ni, & Wang, 2017), A study conducted in India to measure technical efficiency for 20 public hospitals was employed CCR and BCC, CRS and VRS, inputoriented and the total combination number of inputs and outputs was 4 (Surat, Dalbir, & Kamlesh, 2017).…”

“…Several studies were used the DEA model to estimate hospitals efficiency. A study was conducted in Eastern Ethiopia by (Ali, Debela, & Bamud, 2017) to measure technical efficiency for 12 hospitals for six rounded years from 2007/08 to 2012/13. In Iran, a study was conducted by (Sheikhzadeh, Roudsari, Vahidi, Emrouznejad, & Dastgiri, 2012) to estimate technical efficiency for 11 hospitals for the year of 2004.…”

“…In regard to return to scale, there were studies used VRS these studies were (Ali et al, 2017), (Kakeman et al, 2016), (Kalhor et al, 2016), (Mujasi et al, 2016), (H. Li & Dong, 2015), (Applanaidu et al, 2014), (J. P. Harrison & Meyer, 2014), (Jehu-Appiah et al, 2014), (Jat & Sebastian, 2013), (Yusefzadeh, Ghaderi, Bagherzade, & Barouni, 2013). For CRS model, the studies were shared in this model type were (N. Li, Wang, Ni, & Wang, 2017), , (Cheng et al, 2016).…”

“…In term of model orientation, there were some studies applied out-oriented. These studies were (Ali et al, 2017), (Cheng et al, 2016), (Mahate & Hamidi, 2016), (Mujasi et al, 2016), (H. Li & Dong, 2015), (Jehu-Appiah et al, 2014), (Kirigia & Asbu, 2013).…”

“…The studies with combined number of 5 inputs and outputs were (Cheng et al, 2015), (Yusefzadeh, Ghaderi, Bagherzade, & Barouni, 2013). The studies with combined number of 6 inputs and outputs were (Ali et al, 2017), , (Cheng et al, 2016), (J. P. Harrison & Meyer, 2014).…”