Investigating the impact of age, gender, and comorbid conditions on the prolonged length of stay after endarterectomy

Abstract

Endarterectomy is a commonly performed surgical procedure for reducing long-term stroke risks. Due to the prolonged Length of Stay (LOS) experienced by patients undergoing endarterectomy, predicting this parameter has become increasingly important for both costs savings and the improvement of the management of beds. This study aims to develop a prediction model of LOS value starting from the clinical data related to patients undergoing endarterectomy, exploiting the potential of several Machine Learning algorithms. Data extracted from the information system of the “San Giovanni di Dio and Ruggi d'Aragona” University Hospital (Salerno, Italy) were considered to perform the analysis. The proposed prediction model shows promising outcomes in estimating the LOS and therefore it can be a significant tool for enhancing the planning of endarterectomy procedures.