A nomogram for predicting delirium in critically ill cancer patients: A retrospective cohort study.

PURPOSE: This study aimed to develop and validate a nomogram for estimating delirium risk in critically ill cancer patients.

METHOD: A retrospective cohort study was conducted in the ICU of a Tianjin cancer hospital, enrolling 698 critically ill cancer patients from November 2023 to March 2025. Data were extracted from electronic medical records and critical care nursing systems, and delirium was assessed with the Intensive Care Delirium Screening Checklist (ICDSC). Predictors were screened via Least Absolute Shrinkage and Selection Operator (LASSO) regression, then entered into multivariable logistic regression to construct the nomogram. Model performance was evaluated by Hosmer-Lemeshow test, ROC curve, calibration plots and decision curve analysis (DCA); SHapley Additive exPlanations (SHAP) analysis was used to quantify variable contributions and improve interpretability.

RESULTS

The delirium incidence was 11.17%. Seven independent risk factors were included: male gender, primary tumor site, planned transfer, older age, elevated creatinine (Cr), lower potassium (K+), and prolonged ICU length of stay (ICU-LOS). The nomogram showed good discrimination, with AUC of 0.808 in the training set and 0.735 in the validation set. The calibration curve and Hosmer-Lemeshow test (P = 0.187) confirmed favorable consistency between predicted and observed probabilities, and DCA verified its clinical utility.

CONCLUSIONS

A practical delirium risk nomogram was developed and validated for critically ill cancer patients, with satisfactory discrimination, calibration and clinical applicability. It can assist clinicians in early risk identification and targeted interventions to optimize clinical outcomes.