Network data envelopment analysis in healthcare: A systematic literature review

Healthcare performance assessment using nonparametric benchmarking approaches has risen in the literature in recent years. Several studies have been published employing data envelopment analysis (DEA) to evaluate healthcare performance. Among them, network-DEA use has caught special attention due to its advantages in considering more complex systems with departments interacting with each other. Accordingly, this systematic literature review aims to understand and analyze what has been done in the last years using network-DEA to study healthcare and provide helpful findings to those interested in the subject. Following the PRISMA guidelines, a systematic literature review was performed, selecting 40 relevant documents after 1,554 studies were collected and evaluated. All documents analyzing healthcare using network-DEA available in Scopus or Web of Science have been included until February 2025. From them, helpful information such as the framework used, the variables selected, formulation's characteristics, and other metadata have been collected and summarized. The findings suggest that there are topics clearly more studied (e.g. hospitals) than others (e.g., pharmacies). Moreover, the frameworks present several differences, albeit some similarities have been detected within the same case study. As for the variables, an extensive list has been identified, albeit most studies employ the number of beds, physicians, and nurses as well as operating expenses as inputs, inpatient discharges and outpatient visits as outputs, and the overall patient treated as links. By providing this information, this review aims to assist researchers in using network-DEA models to study healthcare by offering essential tools and highlighting gaps in the literature.