An abridged or Short Form PG-SGA score has recently been validated in the chemotherapy outpatient setting in place of the Malnutrition Screening Tool (MST) as a means of providing additional clinically relevant information during screening. The aim of this study was to identify the most relevant information contributing to the PG-SGA score to identify malnutrition with 80% sensitivity and 60% specificity.
A cross-sectional study recruited patients actively receiving treatment in the oncology day unit. Patients self-administered the MST. A dietitian blinded to the MST score administered the PG-SGA. Receiver operating characteristic curves were generated to determine the optimal cut off scores of PG-SGA sections with the greatest sensitivity and specificity for predicting malnutrition according to SGA category.
300 participants were recruited (96.2% response rate), 51.7% male, 58.6±13.3yrs. Participants represented both solid tumours and haematological diagnoses. Scores calculated from weight (box 1), dietary intake (box 2) and symptom data (box 3) with or without activity/function (box 4) data were comparable (AUC=0.85, 95% CI=0.80-0.89; AUC=0.85, 95% CI=0.81-0.89, respectively) and had higher diagnostic value than the MST (AUC=0.77, 95% CI=0.72-0.82) or box 3 data alone (AUC=0.78, 95% CI=0.73-0.83). Using boxes 1-3, a PG-SGA score of ≥2 was 90% sensitive and 67% specific at identifying those at risk of malnutrition.
Although PG-SGA scores from boxes 1-4 had been validated previously, our study, with a larger sample and more diverse diagnoses, determined additional information provided by the functional capacity question did not improve the overall discriminatory value of the PG-SGA.