Diagnostic accuracy of the Malnutrition Universal Screening Tool and Patient-Generated Subjective Global Assessment Short Form for detecting malnutrition in oncology outpatients

Le résumé fourni par la source

BACKGROUND & AIMS: Malnutrition is common in oncology and is associated with poorer treatment tolerance and outcomes yet remains under-recognised in outpatient care. This quality improvement project compared the diagnostic accuracy of the Malnutrition Universal Screening Tool (MUST) and the Patient-Generated Subjective Global Assessment Short Form (PG-SGA SF) for detecting malnutrition in oncology outpatients. METHODS: Adult patients attending a regional oncology day therapies unit for systemic anti-cancer therapy between 631 January 2025 were invited to complete a patient questionnaire incorporating PG-SGA SF items. Dietitians calculated PG-SGA SF and MUST scores. Malnutrition was diagnosed using the Global Leadership Initiative on Malnutrition (GLIM) criteria as the reference standard. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), 95% confidence intervals, and Youden's Index were calculated for MUST cut-offs (≥1, ≥2) and PG-SGA SF cut-offs (≥2-≥9). Differences in classification distributions between tools were assessed using Fisher's Exact Test and McNemar's tests. RESULTS: Of 189 patients who returned questionnaires, 140 were included in the final analysis. Median age was 63 years; 77.1% were female and 58.6% had metastatic disease. GLIM criteria classified 46/140 (32.9%) patients as malnourished. MUST ≥1 demonstrated sensitivity 71.7% and specificity 88.3% (PPV 75.0%, NPV 86.5%), yielding the highest Youden's Index (0.600). MUST ≥2 showed very high specificity (98.9%) but poor sensitivity (21.7%). PG-SGA SF ≥ 2 achieved the highest sensitivity (89.1%) but low specificity (48.9%), while PG-SGA SF ≥ 4 provided the most balanced performance (sensitivity 76.1%, specificity 68.1%; Youden's Index 0.442). Overall classification distributions differed significantly between MUST and PG-SGA SF (Fisher's Exact Test, p = 0.0001) CONCLUSIONS: In this outpatient oncology cohort, MUST ≥1 demonstrated the best overall discrimination for detecting GLIM-defined malnutrition, while PG-SGA SF identified a greater proportion of patients at potential nutritional risk at lower cut-offs. PG-SGA SF thresholds around ≥4-6 may represent a pragmatic balance between sensitivity and specificity for routine outpatient screening. Limited agreement between tools indicates that they capture different dimensions of nutritional compromise, with the PG-SGA SF incorporating patient-reported symptoms and functional status that may identify risk not detected by weight-based screening alone. These aspects should be considered when selecting malnutrition screening tools and cut-off thresholds within oncology outpatient services.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.