Aim: Malnutrition is common in cirrhotic patients; however, there is no gold standard for nutritional assessment for this population. The aim of this study was to develop a novel nutritional screening tool based on objective indicators for cirrhotic patients chronically infected with hepatitis B virus (HBV).
Methods: This was a cross-sectional study. Patients with hepatitis B-related cirrhosis were recruited. Malnutrition was diagnosed by the presence of any of the following conditions: Nutrition Risk Screening 2002 score greater than 3 points, Subjective Global Assessment grade B or C, and body mass index (BMI) < 18.5 kg/m². Nomogram model and decision tree model were developed, and the area under the receiver operating characteristic curve (AUROC) was compared.
Results: Among the 231 studied cases, 92 (39%) were malnourished. Malnourished patients had significantly lower serum albumin, BMI and hand grip strength levels, but higher serum creatinine level and Child–Pugh grade. Two models were developed based on these variables. The nomogram model had a sensitivity of 0.696, a specificity of 0.820 and an accuracy of 0.813. The AUROC of nomogram model was 0.813 (95% CI: 0.758– 0.869, p < 0.001). For the decision tree model, the sensitivity, specificity and accuracy are 0.761, 0.885 and 0.886, respectively, with an AUROC of 0.886 (95% CI: 0842– 0.930, p < 0.001). The difference in AUROC between these two models was not statistically significant (p < 0.001).
Conclusion: The nomogram model and decision tree model developed in this study may aid assessing nutritional status for cirrhotic patients with HBV.
Keywords: cirrhosis, nutrition, screen, nomogram, decision tree