MELD score: utility and comparison with King's College criteria in non-acetaminophen acute liver failure

Abstract

Objective: To compare Model for End-stage Liver Disease Score (MELD Score, MS) and King's College Hospital (KCH) criteria for finding correlation of mortality in non-acetaminophen induced acute liver failure (NAI-ALF).

Study design: An analytical cross-sectional study.

Place and duration of study: The Aga Khan University Hospital, Karachi, from 2005 to 2007.

Methodology: The study included patients with NAI-ALF. KCH criteria were labelled as good and bad prognosis groups. MELD score were calculated by using the MELD calculator. ROC was plotted and sensitivity analysis was done. ETA was used to see correlation between MELD and KCH.

Results: Ninety-one patients with mean age of 32.5 + 16.3 years were studied; 49 were males (54%). Out of these, 57 patients died (63%); two leading causes of non-acetaminophen induced acute liver failure (NAI-ALF) were hepatitis hepatitis B virus (HBV) (n = 30, 33%) followed by hepatitis E virus in (n = 23, 25.3%). According to King's College Hospital (KCH) criteria, 50 patients (88%) who died had bad prognosis and 24 patients (70.6%) who survived had good prognosis. The ROC determined MELD score of 32 was the best predictor of mortality with sensitivity and specificity of 79% and 71%, respectively and positive predictive value (PPV) and negative predictive values (NPV) of 82% and 67% respectively. There was significant association between mortality and bad prognosis according to KCH criteria (p < 0.001). Overall mean MELD score (MMS) was 35.35 + 8.64. MMS on admission was 38 + 7.32 in patients who died and 30.7 + 8.77 in those who survived (p = < 0.001). MMS correlated equally with KCH criteria (ETA = 0.52).

Conclusion: The admission MELD score has an excellent utility and correlates equally with KCH criteria for mortality in NAI- ALF.