Visualisation of Data Envelopment Analysis in primary health services

Abstract

.

Benchmark efciency analysis in public health typically focuses on hospitals rather than primary care providers. Data Envelopment Analysis (DEA) is widely used to assess resource efciency among decision-making units (DMUs). However, traditional DEA struggles to diferentiate between efcient units and is sensitive to the selection of inputs and outputs. Methods like super-efciency and cross-efciency address some of these limitations but often exclude outliers and may overlook efciency related to specialisation. DEA Visualisation integrates DEA with multivariate statistical methods allowing for the identifcation of inefciency sources and specialisation patterns without losing discriminatory power or removing extreme cases from the sample. This study analyses 82 public primary health centres in Madrid serving senior citizens in 2018. The fndings reveal inefciencies such as a preference for prescribing specifc rather than generic drugs, increasing public health costs. Additionally, two extreme cases (outliers or mavericks) were identifed as having high infrastructure costs and disproportionate stafng. Redistributing patients from overcrowded centres could enhance efciency, while centres focused on preventive care showed greater cost-efectiveness, particularly in reducing prescription costs