Abstract
BACKGROUND: The Geriatric Nutrition Risk Index (GNRI) serves as a straightforward screening tool for predicting the likelihood of postoperative complications and mortality in elderly patients. This study aimed to investigate the association between GNRI and postoperative outcomes in elderly patients undergoing pancreaticoduodenectomy (PD). METHODS: We enrolled patients aged 65 years or older who underwent PD between January 2018 and March 2023. Patients were dichotomized into at-risk (GNRI ≤98) or no-risk (GNRI >98) groups. Propensity score matching was used to compare the baseline differences and postoperative outcomes between these two groups. Logistic regression analysis was conducted to assess the association between GNRI and major morbidity (Clavien-Dindo scale with ≥ grade 3b complications). RESULTS: There were 392 eligible patients included. Following propensity score matching, patients in the at-risk GNRI group had higher rates of major complication (13.2% vs. 5.4%, P=0.03) and pulmonary infection (17.8% vs. 9.3%, P=0.046). Based on the multivariate analysis, at-risk group was an independent prognostic factor for major morbidity (odds ratio =2.698, 95% CI: 1.062-6.856, P=0.04). Subgroup analysis revealed that high-risk patients (GNRI <82) exhibited higher rates of preoperative jaundice and longer operative time in comparison to the moderate (GNRI =82-91) and low-risk (GNRI =92-98) groups. However, there were no statistically significant disparities in morbidity and mortality among these three groups. CONCLUSIONS: We demonstrated that nutritional status evaluated by GNRI predicts postoperative complications after PD in elderly patients.