Efficiency Assessment of Tunisian Public hospitals Using Data Envelopment Analysis (DEA)

In the recent past years, Tunisia pursued a national policy on health which was directed towards the performance. Although the lack of adequate resources presents the most important constraint, efficiency in the utilization of available resources is another challenge that cannot be overlooked. The objective of this study aims to assess the technical efficiency (TE) of a sample of Tunisian public hospital using the non-parametric approach of Data Envelopment Analysis (DEA). In this perspective, we started with measuring, comparing and analyzing the TE of the three categories of the Tunisian public hospitals, then to investigate the difference in the level of efficiency by district and finally to Guide the decision and the policy makers in their decision making process through the developed decision making tools. The data were gathered from a sample of 134 public hospitals throughout Tunisia. These would cover about 80% of the total number of Tunisian public hospitals. The model estimates the technical efficiency for the whole sample as well as for each hospital. The entire sample was operating on average at 0, 78 level of technical efficiency. Only 28% of the total hospitals were found to be technically efficient in relative term while the remaining were inefficient. The Public Health Establishment (PHE), the regional hospital (RH) and the District Hospital (DH) were operating on average at 0,9; 0,74 and 0,76 level of technical efficiency respectively. Only 45% of the PHE, 23% of the RH and 25% of the DH were technically efficient while the remaining were inefficient. The study identifies the inefficient hospitals and provides the magnitudes by which specific input per inefficient hospital ought to be more managed or to be reduced. It emphasizes also the disparity by districts in term of percentage of efficient and inefficient hospitals. Therefore, the highest percentage of efficient hospitals is also in the districts of North East (NE) and Center East (CE).