The Use of Stroke-Related Early Tracheostomy Score (SET) as a Predictor for the Need for Tracheostomy in Stroke Patients Admitted to ICU

Akram Henein Guirgis¹*, Mohamed Badawy¹, Amr Aboulela², Muaz Abdullatif³, Sherif  Abdelhamid⁴, Farah Soukieh⁵, Alaa Abuqata⁶, Rajesh Quadros⁶, Alaa A Ibrahim⁶ and Nabil Joseph^7
1ICU Consultant, Al Kuwait Hospital, Dubai/EHS, UAE
²ICU Consultant, Al-Qassimi Hospital/EHS, UAE
³Neurology Consultant, Al-Qassimi hospital/EHS, UAE
⁴Anesthesia Consultant Kuwait Hospital Dubai/EHS, UAE 
⁵Internal Medicine Resident Al-Qassimi Hospital/EHS, UAE
⁶ICU GP Kuwait Hospital, Dubai/EHS, UAE
⁷Professor of Community Medicine, Mansoura University, Faculty of Medicine, Egypt
*Corresponding Author: Akram Henein Guirgis, ICU Consultant, Al Kuwait Hospital, Dubai, UAE.

Published: September 25, 2025

DOI: 10.55162/MCMS.09.316

View Pdf

Abstract  

Background: Stroke patients who are in the intensive care units (ICUs) experience problems because they tend to have a long time of weaning time, and they have a continued high occurrence of extubation failure, and a high probability of the need to be placed on a tracheostomy. A proposed tool to describe predictive scores to assist the choice of early tracheostomy in neurocritical care is the Stroke-Related Early Tracheostomy (SET) score, which has not been externally validated.

Objectives: To determine the accuracy of predicting the requirement of tracheostomy, prolonged ICU stay, and 28-day mortality among stroke patients under ventilation using the SET score.

Methods: In this prospective observational study, 114 stroke patients who were on mechanical ventilation were admitted to the two tertiary intensive care units in the United Arab Emirates. The SETs were done in less than 24 hours after admission. ROC curve analysis and logistic regression were conducted as statistical analyses used to examine predictive validity.

Results: 78.9% of patients were under tracheostomy. The SET score presented ideal predictive ability of the tracheostomy requirement (AUC = 0.897), with the best cutoff value of 8 (sensitivity: 82.22 percent, specificity: 91.67 percent). It was also a moderate predictor of prolonged stay in ICU (AUC = 0.767) but not good at discriminating mortality (AUC = 0.516).

Conclusion: SET score is a dependable aspect of predicting the requirement for tracheostomy amongst stroke patients. The use of early may help in better management of the airway and ICU resource planning.

Keywords: Stroke; Tracheostomy; Intensive Care Units; Mechanical Ventilation; Prognostic Score; Neurologic Critical Care