BACKGROUND & AIMS: The Geriatric Nutritional Risk Index (GNRI) is a screening tool for nutrition-related risk that correlates with mortality rate in hospitalized older patients and is simple, objective, and readily available to clinicians. In this study, we aimed to validate the performance of the GNRI in predicting short-term hospital mortality in older patients with sepsis. METHODS: This observational study enrolled 401 older patients presenting with infection and systemic inflammatory response syndrome in an emergency department. Demographic, physiological, and laboratory data were collected. The GNRI score was categorized into five classes. The primary outcome was 28-day hospital mortality. Univariate and multivariate analyses were performed to identify clinical predictors of outcome. A logistic regression model was used. RESULTS: 51 patients (12.7%) died in the hospital within 28 days. Co-morbid metastatic cancer, heart rate, respiratory rate, temperature, serum creatinine, total lymphocyte count, and GNRI (<87) were independently related to the outcome in the multivariable logistic regression analysis. CONCLUSIONS: The GNRI is a prognostic factor for short-term hospital mortality in older patients with sepsis. A GNRI below 87 can be suggested as an indicator of nutritional support need in an acute-care setting.
BACKGROUND & AIMS: The Geriatric Nutritional Risk Index (GNRI) is a screening tool for nutrition-related risk that correlates with mortality rate in hospitalized older patients and is simple, objective, and readily available to clinicians. In this study, we aimed to validate the performance of the GNRI in predicting short-term hospital mortality in older patients with sepsis. METHODS: This observational study enrolled 401 older patients presenting with infection and systemic inflammatory response syndrome in an emergency department. Demographic, physiological, and laboratory data were collected. The GNRI score was categorized into five classes. The primary outcome was 28-day hospital mortality. Univariate and multivariate analyses were performed to identify clinical predictors of outcome. A logistic regression model was used. RESULTS: 51 patients (12.7%) died in the hospital within 28 days. Co-morbid metastatic cancer, heart rate, respiratory rate, temperature, serum creatinine, total lymphocyte count, and GNRI (<87) were independently related to the outcome in the multivariable logistic regression analysis. CONCLUSIONS: The GNRI is a prognostic factor for short-term hospital mortality in older patients with sepsis. A GNRI below 87 can be suggested as an indicator of nutritional support need in an acute-care setting.