Evaluating Patient Safety Drivers using Decision Trees

Abstract

This study aims to identify the drivers of patient safety challenges in healthcare. Hospital-level aggregate survey data from the UK hospital is used to identify what drives the number of reported incidents affecting patient safety. Leveraging a decision tree algorithm, our results suggest that the role of teamwork and safety culture in incident reporting and investigation are the most critical dimensions influencing the number of reported patient safety challenges. The decision tree algorithm can be useful for hospitals to enhance patient safety through data-driven approaches and direct resources towards service improvements.