Background: Paracetamol (acetaminophen) toxicity remains the leading cause of acute liver failure (ALF) in the developed world. In the UK, the recently modified King's College Criteria are used to list patients for emergency liver transplantation, but these criteria have been criticized for their low sensitivity and for spectrum bias in their application.
Aim: To evaluate existing prognostic criteria critically for predicting death without transplantation in paracetamol-induced ALF.
Methods: MEDLINE, EMBASE and CINAHL were searched to identify studies containing adult patients with paracetamol-induced ALF. Selected studies were evaluated and data were pooled if appropriate, to calculate sensitivity, specificity and diagnostic odds ratios (DORs) of applied prognostic tests.
Results: Of 6507 studies identified, 14 were eligible for inclusion, evaluating 1960 patients. The original King's College Criteria had a pooled sensitivity of 58.2% and specificity of 94.6%, with a DOR of 27.7. Addition of arterial lactate to the King's College Criteria reduced the DOR to 26.1. Several other clinical and laboratory variables had higher DORs than the King's College Criteria, but were only evaluated in single studies of limited quality.
Conclusions: The original King's College Criteria remain well-validated criteria with high prognostic accuracy. Other potential prognostic variables should be prospectively assessed in multicentre studies to refine the criteria further.