Geriatric nutritional risk index and outcomes following a major surgical procedure: A national cohort study

BACKGROUND: Despite a rising incidence among older adults, the effect of malnutrition on perioperative morbidity remains unclear. We assessed the prognostic value of the geriatric nutritional risk index for outcomes after major surgery. METHODS: The nationwide Epic COSMOS electronic health record database identified adults undergoing coronary artery bypass grafting, lung resection, abdominal aortic aneurysm repair, or major gastrointestinal resection. The geriatric nutritional risk index was calculated from serum albumin and anthropometric measurements and categorized using established cutoffs as major risk (<82), moderate risk (82-91), low risk (92-98), and no risk (>98). Associations with 30-day mortality, complications, readmission, and nonhome discharge were evaluated using entropy-balanced weighted logistic regression adjusted for patient and operative factors. RESULTS: Among 506,040 patients undergoing major surgery, mean age was 66.9 years and 46.2% were female. Patients with a major geriatric nutritional risk index were older than patients with no geriatric nutritional risk index (median, 71 vs 68 years), more often had a Charlson Comorbidity Index ≥1 (81.1% vs 75.2%), and were Black (16.4% vs 11.5%) (all P < .001). Outcomes worsened with increasing geriatric nutritional risk index: mortality (no risk, 1.7%; low risk, 4.1%; moderate risk, 6.6%; and major risk, 13.3%), readmission (no risk, 12.3%; low risk, 17.0%; moderate risk, 19.0%; and major risk, 23.5%), and nonhome discharge (no risk, 33.1%; low risk, 46.1%; moderate risk, 53.3%; and major risk, 64.3%) (all P < .001). Compared with no geriatric nutritional risk index, low and major geriatric nutritional risk index were independently associated with higher odds of any 30-day complication (low risk: adjusted odds ratio, 1.79; 95% confidence interval, 1.75-1.82; major risk: adjusted odds ratio, 2.47; 95% confidence interval, 2.40-2.54) and mortality (low risk: adjusted odds ratio, 3.35; 95% confidence interval, 3.20-3.50; major risk: adjusted odds ratio, 8.03; 95% confidence interval, 7.70-8.38) (all P < .001). The geriatric nutritional risk index demonstrated a strong inverse association with 30-day mortality, with the highest risk observed among patients with a major geriatric nutritional risk index. CONCLUSION: Greater geriatric nutritional risk index-defined nutrition-related risk was independently associated with higher mortality, major complications, and readmission compared with no geriatric nutritional risk index. Geriatric nutritional risk index-based assessment may identify nutritionally vulnerable patients for preoperative optimization, early intervention, and tailored postoperative discharge planning.