Specific Nutrition Evaluation for Advanced Chronic Liver Disease: Validation through psychometrics analysis

Validated instruments specifically designed to assess the nutritional status of individuals with chronic liver disease (CLD) are scarce. The purpose of this study was to evaluate the psychometric properties of the Specific Nutrition Evaluation for Advanced Chronic Liver Disease (SNE-ACLD) instrument and establish cut-off points for the nutritional assessment of individuals with CLD.

Hospitalized individuals with CLD were evaluated by SNE-ACLD and different nutritional assessment tools. Initial evidence of structural construct validity was assessed using principal component analysis; reliability using Cronbach's alpha; cut-off points were determined using cluster analysis. Criterion validity, in which the appendicular muscle mass index was adopted as a reference for concurrent validity using ROC curve; and Cox regression to assess the one-year mortality risk for predictive validity.

Two hundred participants were evaluated, of which 75.0% were male, 50.5% were elderly, and 80.8% were classified as Child-Pugh B or C. Principal component analysis demonstrated that SNE-ACLD has four components that explain 69.9% of the total variance of the construct. Cronbach's alpha of 0.70 indicated a good interrelationship between the items. The established cut-off points classified the participants into four groups: without malnutrition, mild, moderate, or severe malnutrition. The SNE-ACLD showed concurrent validity with an AUC of 0.714 and predictive validity, with severely malnourished participants presenting one-year mortality risk almost five times higher (HR = 4.98; 95% CI: 2.42-10.2) than those without severe malnutrition.

SNE-ACLD effectively addresses the domains of the malnutrition construct in this public, presenting evidence of both internal and external validity.