Patient- and Nutrition-Derived Outcome Risk Assessment Score as a Predictor of Mortality in Critically Ill Surgical Patients: A Retrospective, Single-Center Observational Study

Background The Patient- And Nutrition-Derived Outcome Risk Assessment (PANDORA) was recently validated for predicting mortality in hospitalized patients; however, its utility in the intensive care unit (ICU) remains unknown. Methods We investigated whether PANDORA is associated with 30, 90, and 180 day mortality in critically ill surgical patients by performing logistic regressions, controlling for age, sex, race, body mass index, macronutrient deficit, and length of stay. The area under the receiver operating characteristic curves (AUC) of PANDORA vs Acute Physiology and Chronic Health Evaluation (APACHE) II scores for mortality at each time point were also compared. Results 312 patients comprised the analytic cohort. PANDORA was associated with mortality at 30 (OR 1.08; 95% CI 1.04-1.13; P < .001), 90 (OR 1.09; 95% CI 1.03-1.12; P < .001), and 180 days (OR 1.10; 95% CI 1.06-1.15; P < .001). PANDORA and APACHE II were comparable for mortality prediction at 30 (AUC: 0.69, 95% CI 0.62-0.76 vs 0.74, 95% CI 0.67-0.81; P = .29), 90 (AUC: 0.71, 95% CI 0.63-0.77 vs 0.74, 95% CI 0.67-0.80; P = .52), and 180 days (AUC: 0.73, 95% CI 0.67-0.79 vs 0.75, 95% CI 0.69-0.81; P = .66). Conclusion In surgical ICU patients, PANDORA was associated with mortality and was comparable with APACHE II for mortality prediction at 30, 90, and 180 days after initiation of care. Prospective studies are needed to assess whether nutrition support, stratified by PANDORA scores, can improve outcomes in surgical ICU patients.