PP166 - NRS-2002 OUTPERFORMS MST IN PREDICTING OUTCOME FOR GENERAL SURGERY PATIENTS

PP166

NRS-2002 OUTPERFORMS MST IN PREDICTING OUTCOME FOR GENERAL SURGERY PATIENTS

S. C. Brudeseth^1,*, J. Moen Johansen^1,2, T. Hjetland Løland^3,4, A. B. Kildal^3,5, M. Hagve^1,2

¹Metabolism and Clinical Nutrition Research Group, Department of Clinical Medicine, Faculty of Health Sciences, UIT – The Arctic University of Norway, ²Department of Gastrointestinal Surgery, University Hospital of North Norway, ³Anesthesia and Critical Care Research Group, Department of Clinical Medicine, Faculty of Health Sciences, UIT – The Arctic University of Norway, ⁴Clinical Nutrition Centre, ⁵Department of Anesthesiology and Intensive Care, University Hospital of North Norway, Tromsø, Norway

Rationale: Nutritional screening of surgical patients must be accurate, as nutritional risk may affect outcomes. Hence, a tool with high sensitivity that accounts for inflammatory burden is needed. We aimed to compare the agreement and predictive performance of two recommended screening tools: the Malnutrition Screening Tool (MST), valued for its simplicity and feasibility, and Nutritional Risk Screening 2002 (NRS-2002), which has stronger evidence for predicting outcomes.

Methods: In this prospective single-center observational study, 598 patients were screened with both MST and NRS-2002. Length of stay (LOS), adverse outcomes, and nutritional interventions were recorded. Agreement was assessed with McNemar’s test, and clinical outcomes were analyzed using multivariable regression.

Results: Nutritional risk was identified in 33% of patients using NRS-2002 and 27% by MST, but agreement between the tools was low, (58% and 69%, respectively, p < 0.005). The difference was greater in acutely admitted patients, as NRS-2002 identified more patients (40% vs 29%), with only 53% agreement with MST (p < 0.0001). Only NRS-2002 predicted overall adverse outcome (NRS-2002; OR 2.1, 95% CI 1.3–3.3, p < 0.0001, MST: OR 1.2, 95% CI 0.8–2.0, n.s.), including surgical complications (Clavien–Dindo > 3b, p < 0.001). Nutritional risk identified with both tools were associated with prolonged LOS (p < 0.005), but NRS-2002 estimated an increase in four days compared to two days using MST. Independent of which tool was used, nutritional risk was associated with onset of a nutritional intervention.

Conclusion: MST and NRS-2002 identify largely different subsets of surgical patients, despite classifying a similar overall proportion. NRS-2002 outperforms MST in predicting outcome, suggesting it remains a more reliable tool for nutritional risk assessment in general surgery patients.

Disclosure of Interest: None declared