Background: Delirium, an acute neuropsychiatric syndrome marked by fluctuating disturbances in attention, cognition, and consciousness, is prevalent among older hospitalized cancer patients. The Mayo Delirium Prediction (MDP) tool, designed to predict delirium risk in older adults using electronic health record (EHR) data, may help identify individuals at increased risk, thus offering valuable insights for targeted interventions and resource allocation (Pagali, 2021). This study aims to validate the Mayo Delirium Prediction (MDP) tool in cancer patients.
Methods: This IRB-approved, retrospective validation study included 930 hospitalized cancer patients aged 50 years or older, admitted to MD Anderson Cancer Center from September 1, 2022, to September 1, 2023. Patients with acute substance use disorder were excluded. MDP variables recorded within 24 hours of admission, including age, medical history, and laboratory values, were extracted from the electronic Health Record. The MDP tool’s predictive accuracy was assessed using receiver operating characteristic (ROC) curves, sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for patients admitted to medical and surgical services.
Results: The cohort comprised 767 medical and 163 surgical patients (median age, 67 years). The MDP tool demonstrated areas under the ROC curve (95% CI) of 0.8047 (0.7585–0.8509) for the entire cohort, 0.7867 (0.7371–0.8363) for medical patients, and 0.9488 (0.8795–1.0000) for surgical patients. Sensitivity, specificity, PPV, and NPV differed by risk cutoffs. At a cutoff of ≥0.3, sensitivity was 42.6% (95% CI: 32.6%–52.5%) and specificity was 94.6% (95% CI: 93.1%–96.1%). At a cutoff of >0.05, sensitivity was 88.3% (95% CI: 81.8%–94.8%), but specificity was 50.5% (95% CI: 47.1%–53.9%).
Conclusions: The MDP tool demonstrates robust capability for predicting delirium risk in hospitalized cancer patients, particularly in surgical settings. Performance differs by risk cutoffs, highlighting the need for tailored application in clinical practice. Further validation and refinement could enhance the tool’s utility in preventing and managing delirium in this vulnerable population.