Rule-based decision tree for surgical priorization in otolaringology

Abstract

Effective waitlist management for surgery in otolaryngology (ENT), as in most public health care programs, remains a significant issue. In the present study, using the Classification and Regression Tree (CART) algorithm, a decision tree model based on biopsychosocial factors was developed to classify 205 patients into three priority classes (High, Medium, and Low) based on a continuous score assigned by the attending physicians. The last tree considered severity (Sev), urgency (Urg), and being able to attend the study (Dest) as the main predictors, yielding five validated clinical rules. With an overall accuracy of 82,3%, the model shows its predictive power and practical relevance. Its transparency and clinical reasoning underpin its trustworthiness as a decision aid. The next steps will include interfacing with dynamic prioritization methods.