Prospective evaluation of the CASST score to guide elective tracheostomy following oral cancer surgery: challenges and clinical misalignment

The Clinical Assessment Scoring System for Tracheostomy (CASST) was developed to predict the need for elective tracheostomy in head and neck cancer surgery, but evidence on its external performance remains limited. This prospective study evaluated the predictive accuracy of CASST and its concordance with clinical decision-making in patients undergoing major oral cancer surgery. A secondary objective was to describe short-term postoperative airway complications stratified by CASST category. All patients undergoing major oral cancer resections at a tertiary care centre between September 2021 and January 2024 were enrolled and followed prospectively. CASST scores were computed preoperatively but did not guide airway management decisions. Among 258 patients, 126 (48.8%) underwent elective tracheostomy. Only 31.7% of these procedures aligned with CASST score recommendations (score ≥7). The CASST score demonstrated a sensitivity of 31.7%, specificity of 91.7%, positive predictive value of 78.4%, and negative predictive value of 58.5%. The area under the receiver operating characteristic curve was 0.617, demonstrating poor overall discrimination. Short-term postoperative complications did not differ significantly between CASST risk groups. While CASST demonstrates high specificity, its low sensitivity and frequent misalignment with clinical practice limit its standalone utility. Further refinement and integration with intraoperative findings may improve its applicability in airway management planning.