Prognostic Role of Nutritional and Inflammatory Indices in Predicting Adverse Clinical Outcomes in Unplanned...

Background: Unplanned hospitalizations in patients with cancer are associated with adverse outcomes, including intensive care unit (ICU) transfer and in-hospital mortality. This study aimed to evaluate the predictive role of the prognostic nutritional index (PNI) and albumin-to-globulin ratio (AGR) for these outcomes in patients with unplanned hospitalization in a medical oncology ward. Methods: This retrospective, single-center study included patients aged ≥18 years with malignancy who had unplanned hospitalization between 1 January and 30 April 2025. PNI and AGR were calculated at admission. The primary outcome was ICU transfer or in-hospital mortality. Univariable and multivariable logistic regression analyses were performed, with AGR and PNI evaluated in separate models to avoid collinearity. Predictive performance was assessed using ROC analysis. Results: A total of 418 patients were included, with adverse clinical outcomes in 26.8%. Metastatic disease was present in 73.7%, and gastrointestinal (41.6%) and lung cancers (21.5%) were most common. In univariable analysis, metastatic disease (p < 0.001), Eastern Cooperative Oncology Group (ECOG) performance status (p < 0.001), cancer type (p = 0.030), reason for hospitalization (p = 0.001), AGR (p < 0.001), and PNI (p < 0.001) were significantly associated with adverse clinical outcomes. In multivariable analyses performed in separate models, ECOG ≥ 2 emerged as the strongest independent predictor of adverse clinical outcomes (AGR model: OR: 9.93; PNI model: OR: 11.14; both p < 0.001). Metastatic disease remained an independent risk factor, while higher AGR and PNI values were independently associated with a reduced risk (all p < 0.05). Among hospitalization reasons, only electrolyte imbalance/transfusion was associated with a lower risk, whereas most cancer type subgroups were not independently significant. Both indices showed moderate predictive performance, with PNI performing slightly better than AGR (AUC: 0.729 vs. 0.707). Conclusions: ECOG performance status, together with PNI and AGR, were identified as practical and accessible predictors of adverse clinical outcomes in patients with unplanned hospitalization in a medical oncology ward.