Diagnostic test accuracy of preoperative nutritional screening tools in adults for malnutrition: a systematic review and network meta-analysis

Background: Good nutritional screening tests can triage malnourished patients for further assessment and management by dietitians before surgery to reduce the risk of postoperative complications. The authors assessed the diagnostic test accuracy of common nutritional screening tools for preoperative malnutrition in adults undergoing surgery and determined which test had the highest accuracy. Methods: MEDLINE, EMBASE, CINAHL, and Web of Science were searched for relevant titles with no language restriction from inception till 1 January 2023. Studies reporting on the diagnostic test accuracy of preoperative malnutrition in adults using one or more of the following index nutritional screening tools were included: Malnutrition Screening Tool (MST), Malnutrition Universal Screening Tool (MUST), Mini Nutritional Assessment (MNA), short-form Mini Nutritional Assessment (MNA-SF), Nutritional Risk Index (NRI), Nutrition Risk Screening Tool 2002 (NRS-2002), and Preoperative Nutrition Screening (PONS). The reference standard was the Subjective Global Assessment (SGA) before surgery. Random-effects bivariate binomial model meta-analyses, meta-regressions, and a network meta-analysis were used to estimate the pooled and relative sensitivities and specificities. Results: Of the 16 included studies (5695 participants with an 11 957 index and 11 957 SGA tests), all were conducted after hospital admission before surgery. Eleven studies (n=3896) were at high risk of bias using the Quality Assessment of Diagnostic Accuracy Studies tool due to a lack of blinded assessments. MUST had the highest overall test accuracy performance (sensitivity 86%, 95% CI: 75-93%; specificity 89%, 95% CI: 83-93%). Network meta-analysis showed NRI had similar relative sensitivity (0.93, 95% CI: 0.77-1.13) but lower relative specificity (0.75, 95% CI: 0.61-0.92) than MUST. Conclusions: Of all easy-to-use tests applicable at the bedside, MUST had the highest test accuracy performance for screening preoperative malnutrition. However, its predictive accuracy is likely insufficient to justify the application of nutritional optimization interventions without additional assessments.