RESUMO
BACKGROUND AND AIMS: Malnutrition is a modifiable risk factor for morbidity and mortality in cirrhosis. Nutrition risk screening is recommended in cirrhosis nutrition guidelines, but is not routinely completed in practice. The patient-generated subjective global assessment short form (PG-SGA SF) is a patient-completed screen that has potential to be a substitute for more time and resource intensive nutrition screens. The aim of this cross-sectional study was to compare the PG-SGA SF and three other patient-completed screens against the nutrition assessment reference method in cirrhosis, the Royal Free Hospital subjective global assessment (RFH-SGA). We also explored whether being classified "at-risk" on a nutritional screening tool was associated with clinical outcomes of unplanned hospitalization or death. METHODS: Patients completed four nutrition screening tools with or without support from a caregiver. The RFH-SGA was carried out by a blinded registered dietitian. The four screening tools were compared against the RFH-SGA to calculate sensitivity, specificity, and positive and negative predictive value. RESULTS: A total of 121 patients were included. The PG-SGA SF screened the highest number of patients positive for malnutrition risk (52%), was the most accurate, and had the highest sensitivity. Being at risk for malnutrition on the PG-SGA SF was associated with a higher risk of unplanned hospitalization (unadjusted sHR 2.78 (95% CI 1.3-5.9), p = 0.009). CONCLUSIONS: The PG-SGA SF identifies malnutrition risk at similar or superior rates to other patient-generated screening tools in patients with cirrhosis. Our findings support its potential as a patient completed solution for identifying malnutrition risk in cirrhosis.