What scoring system is used to assess acute liver failure?

Scoring Systems for Acute Liver Failure

The MELD score is the recommended prognostic scoring system for acute liver failure, with a cutoff of 30.5 or higher predicting need for liver transplantation, and should be used preferentially over King's College Criteria. 1

Primary Scoring Systems

MELD Score (Preferred)

• The American Gastroenterological Association recommends using MELD score rather than King's College Criteria as the prognostic scoring system for acute liver failure. 1
• MELD demonstrates superior sensitivity (77%, range 70-92%) compared to King's College Criteria (61%, range 47-76%) for predicting mortality, though King's College Criteria has higher specificity (86% vs 72%). 1
• A MELD score ≥30.5 is the fixed cutoff value that predicts need for liver transplantation in acute liver failure. 1
• The MELD score was evaluated in 6 studies involving 526 ALF patients with 58% mortality, showing diagnostic odds ratio of 8.79 (95% CI, 5.19-14.89). 1
• MELD offers the advantage of optimizing specificity without losing significant sensitivity, making it the optimal choice. 1

King's College Criteria (Alternative)

• King's College Criteria has been evaluated in 8 studies involving 962 patients with ALF (47% mortality), showing pooled sensitivity of 61% and specificity of 86%. 1
• While more specific than MELD, the lower sensitivity means it may miss patients who will deteriorate and require transplantation. 1
• The 2005 AASLD position paper noted that King's College Criteria showed positive predictive values ranging from just below 70% to nearly 100%, but negative predictive values ranging from only 25% to 94%. 1

Important Clinical Context

Limitations of Current Scoring Systems

• Currently available prognostic scoring systems do not adequately predict outcome and determine candidacy for liver transplantation in all cases. 1
• The 2005 AASLD guidelines explicitly state that reliance entirely upon these scoring guidelines is not recommended. 1
• MELD score fails to predict mortality in approximately 15-20% of patients with end-stage liver disease. 2, 3

Etiology-Specific Considerations

• The most significant predictor of outcome in the largest U.S. multicenter ALF study was etiology: acetaminophen, hepatitis A, shock liver, or pregnancy-related disease showed 50% transplant-free survival, while all other etiologies showed only 25% transplant-free survival. 1
• Patients presenting in grade III or IV encephalopathy were less likely to survive without liver transplantation compared to those presenting in grade I or II encephalopathy. 1

Critical Distinction: Acute Liver Failure vs Acute-on-Chronic Liver Failure

Note that acute liver failure (ALF) and acute-on-chronic liver failure (ACLF) are distinct entities requiring different scoring approaches. 1

• For ACLF specifically, CLIF-C ACLF score, NACSELD ACLF score, and AARC score are more appropriate than standard MELD, as they incorporate extrahepatic organ failures. 1
• MELD and MELD-Na may underestimate mortality in ACLF patients because they capture intrinsic liver disease but not the impact of extrahepatic organ failures. 1

Practical Application Algorithm

1. Confirm diagnosis of acute liver failure (not acute-on-chronic liver failure)
2. Calculate MELD score using serum bilirubin, creatinine, and INR 1
3. Use cutoff of MELD ≥30.5 to identify patients likely requiring liver transplantation 1
4. Consider etiology as acetaminophen-related ALF has better prognosis than other causes 1
5. Assess encephalopathy grade as grade III-IV predicts worse outcome 1
6. Do not rely solely on scoring systems for transplant decisions given their limitations 1