The geriatric nutrition risk index versus the mini-nutritional assessment short form in predicting postoperative delirium and hospital length of stay among older non-cardiac surgical patients: a prospective cohort study

Backgrounds: Malnutrition has been shown to be associated with poor prognosis in older surgical patients. Several tools are available for detecting malnutrition. But little is known about their ability to assess risks of postoperative adverse outcomes. The study aimed to compare the ability of the Geriatric Nutritional Risk Index (GNRI) and the Mini-Nutritional Assessment Short Form (MNA-SF) in predicting postoperative delirium (POD) and length of stay (LOS) among older non-cardiac surgical patients.

Methods: Prospective study of 288 older non-cardiac surgical patients from the West China Hospital of Sichuan University. Preoperative nutritional status was assessed using the GNRI and MNA-SF, and patients were followed for the occurrence of POD and LOS. Multivariable logistic regression and linear regression analyses were used to identify predictors of these outcomes. The relative performance of the GNRI and MNA-SF as predictors of these outcomes were determined by Receiver Operating Characteristic curves (ROC) analyses and the area under the curve (AUC).

Results: Multivariable analysis revealed that preoperative malnutrition by the MNA-SF was significantly associated with POD. Linear regression analysis showed that preoperative low/high nutritional risk of the GNRI and malnutrition by the MNA-SF were independent predictors of prolonged LOS. Moreover, the area under the curve (AUC) of MNA-SF scores for POD was better than GNRI scores (AUC = 0.718, 95%CI: 0.64-0.80, P < 0.001 vs AUC = 0.606, 95%CI: 0.52-0.69, P = 0.019; Delong's test, P = 0.006), but the AUC of GNRI scores and MNA-SF scores have no significant difference when predicting prolonged LOS (AUC = 0.611, 95%CI: 0.54-0.69, P = 0.006 vs AUC = 0.533, 95%CI: 0.45-0.62, P = 0.421; Delong's test, P = 0.079).

Conclusion: The MNA-SF was more effective than the GNRI at predicting the development of POD, but the two nutrition screening methods have similar performance in predicting prolonged LOS among older non-cardiac surgical patients.