30-day mortality prediction in acute upper gastrointestinal bleeding: Incremental value of the prognostic nutritional index with ABC and Rockall scores

Mortality risk among patients admitted to the emergency department (ED) with acute upper gastrointestinal (GI) bleeding is heterogeneous, underscoring the importance of early identification of high-risk individuals. This study aimed to evaluate the prognostic performance of the prognostic nutritional index (PNI) in predicting 30-day mortality and to determine whether incorporating PNI into established risk markers enhances prognostic accuracy. In this retrospective cohort study, we analyzed data from 619 patients with acute upper GI bleeding who presented to a tertiary university hospital between January 1, 2018, and December 31, 2024. Demographic, clinical, and laboratory data were extracted from medical records. PNI was calculated using serum albumin and lymphocyte count at the time of admission, with the primary outcome being 30-day mortality. Predictors of mortality were examined using univariable and multivariable logistic regression analyses. The incremental prognostic value of PNI was evaluated through receiver operating characteristic (ROC) analysis and the DeLong test. The median age of participants was 74.0 years (interquartile range: 63.0-81.0), and 38% of the patients were female. The observed 30-day mortality rate was 7.9%. Non-survivors displayed significantly lower PNI levels compared to survivors (37.6 vs. 43.6; p < 0.001). In multivariable analysis, PNI (odds ratio [OR]: 0.847 [0.765-0.938]), lactate level (OR: 1.225 [1.047-1.434]), and the ABC score (OR: 1.201 [1.053-1.370]) were identified as independent predictors of mortality. The risk of mortality increased substantially when low PNI was combined with a high ABC score or elevated lactate level. Incorporating PNI into a baseline model resulted in a modest increase in the area under the receiver operating characteristic curve (AUROC) from 0.708 to 0.774 (p = 0.049). In conclusion, PNI serves as an independent predictor of 30-day mortality in patients with acute upper GI bleeding. Its integration with existing risk scores may enhance prognostic discrimination and facilitate early risk stratification in the ED.