Sex-Specific Validity of the PRE-DELIRIC Model and the E-PRE-DELIRIC Model for Predicting Delirium in Patients After Cardiac Surgery.

Background: Several risk prediction models are used in intensive care units (ICUs) to detect delirium risk.

Objective: The aim of this study was to compare the sex-specific validity of the Prediction of Delirium in ICU patients (PRE-DELIRIC) and Early Prediction of Delirium in ICU patients (E-PRE-DELIRIC) models for predicting delirium in ICU patients after cardiac surgery.

Methods: Retrospective data from electronic medical records of ICU patients who underwent cardiac surgery between January 2019 and December 2022 were examined. Both models were evaluated using sensitivity, specificity, and positive and negative predictive values based on the highest Youden index for female and male patients. Receiver operating characteristic curve analysis was performed.

Results: Delirium incidence was 30.7% and 27.5% in female and male patients, respectively. In the PRE-DELIRIC model, the highest Youden index scores for female and male patients were 67.00 and 57.54, respectively, with sensitivities and specificities of 61% and 88%, and 72% and 79%, respectively. In the E-PRE-DELIRIC model, the highest scores for female and male patients were 14.25 and 14.05, with sensitivities and specificities of 78% and 50%, and 70% and 53%, respectively. The area under the curve of the PRE-DELIRIC model for female and male patients was 0.77 and 0.78, respectively, and that of the E-PRE-DELIRIC model was 0.67 and 0.63, respectively.

Conclusions: The PRE-DELIRIC model showed higher specificity and was more reliable for male patients, whereas the E-PRE-DELIRIC model demonstrated better sensitivity for female patients due to its inclusion of more female-sensitive factors. A combined approach is recommended for improved reliability and validity in delirium risk assessment in clinical settings.