ABSTRACT
Introduction: Malnutrition is a common condition in liver cirrhosis (LC), which is associated with poor survival. Despite the wide range of tools, there is no agreement on a standard nutritional assessment method applicable to LC. Aim: To determine the validity and prognostic value of the Patient-Generated Subjective Global Assessment (PG-SGA) as a nutritional assessment tool in LC patients. Material and methods: In 2019-2021, 161 patients with LC (aged 55.2 ±11.6 years) were involved, of whom 23, 57, and 81 patients were classified as Class A, B, and C Child-Turcotte-Pugh (CTP), accordingly. Fifty patients died during follow-up (489 (293-639) days). The PG-SGA, Controlling Nutritional Status (CONUT), handgrip strength, and skeletal muscle index (SMI) were used to assess nutritional status. Results: According to the PG-SGA 29.8% of patients were moderately malnourished and 29.8% were severely malnourished. 50.6% of CTP C patients were severely malnourished. Numerical PG-SGA correlated with CTP, Model for End-Stage Liver Disease, CONUT, SMI, and handgrip strength. Low SMI and handgrip strength were present in 87.5% and 66.7% of severely malnourished patients, respectively. PG-SGA predicted mortality (AUC = 0.775, p < 0.001). Severely malnourished patients had significantly lower survival than moderately malnourished and well-nourished patients in the Kaplan-Meier analysis. Hepatic encephalopathy (HR = 2.29, p = 0.046), hypoalbuminemia (HR = 2.27, p = 0.022), and severe malnutrition according to PG-SGA (HR = 2.39, p = 0.016) were independent predictors of mortality in Cox proportional hazards regression analysis. Conclusions: The PG-SGA is a reliable nutritional assessment tool and can predict mortality in LC patients.
