How efficient are referral hospitals in Uganda? A data envelopment analysis and tobit regression approach

Mujasi, Paschal N.; Asbu, Eyob Z.; Puig i Junoy, Jaume

doi:http://dx.doi.org/10.1186/s12913-016-1472-9

How efficient are referral hospitals in Uganda? A data envelopment analysis and tobit regression approach

License

Citation

Mujasi PN, Asbu EZ, Puig-Junoy J. How efficient are referral hospitals in Uganda? A data envelopment analysis and tobit regression approach. BMC Health Services Research. 2016;16:230. DOI: 10.1186/s12913-016-1472-9

Abstract

Background: Hospitals represent a significant proportion of health expenditures in Uganda, accounting for about 26 % of total health expenditure. Improving the technical efficiency of hospitals in Uganda can result in large savings which can be devoted to expand access to services and improve quality of care. This paper explores the technical efficiency of referral hospitals in Uganda during the 2012/2013 financial year. Methods: This was a cross sectional study using secondary data. Input and output data were obtained from the Uganda Ministry of Health annual health sector performance report for the period July 1, 2012 to June 30, 2013 for the 14 public sector regional referral and 4 large private not for profit hospitals. We assumed an output-oriented model with Variable Returns to Scale to estimate the efficiency score for each hospital using Data Envelopment Analysis (DEA) with STATA13. Using a Tobit model DEA, efficiency scores were regressed against selected institutional and contextual/environmental factors to estimate their impacts on efficiency. Results: The average variable returns to scale (Pure) technical efficiency score was 91.4 % and the average scale efficiency score was 87.1 % while the average constant returns to scale technical efficiency score was 79.4 %. Technically inefficient hospitals could have become more efficient by increasing the outpatient department visits by 45,943; and inpatient days by 31,425 without changing the total number of inputs. Alternatively, they would achieve efficiency by for example transferring the excess 216 medical staff and 454 beds to other levels of the health system without changing the total number of outputs. Tobit regression indicates that significant factors in explaining hospital efficiency are: hospital size (p < 0.01); bed occupancy rate (p < 0.01) and outpatient visits as a proportion of inpatient days (p < 0.05). Conclusions: Hospitals identified at the high and low extremes of efficiency should be investigated further to determine how and why production processes are operating differently at these hospitals. As policy makers gain insight into mechanisms promoting hospital services utilization in hospitals with high efficiency they can develop context-appropriate strategies for supporting hospitals with low efficiency to improve their service and thereby better address unmet needs for hospital services in Uganda.