There are numerous tables and criteria scales which are available.
O'Grady criteria, otherwise known as the King's College criteria, can be used to predict mortality among the paracetamol and non-paracetamol groups of patients. The other well-examined set of criteria are MELD (Model for End-stage Liver Disease) These have been around since the late eighties-early nineties. A good article compares the two, and discusses their relative merits at some length.
King's College (O'Grady) Criteria - for Paracetamol Overdose
- pH of 7.3 on ABG, following fluid resuscitation, more than 24 hours post ingestion
- PT over 100 seconds (INR 6.5)
- Creatinine over 300mmol/L
- Grade 3-4 encephalopathy
- All of these must be present within a 24 hour timeframe
King's College (O'Grady) Criteria - for non-paracetamol acute liver failure
- pH of 7.3 on ABG, following fluid resuscitation
- PT over 100 seconds (INR 6.5)
- Encephalopathy AND
- Age over 10, or over 40
- Bilirubin over 300 mmol/L
- More than 7 days separate onset of jaundice from onset of encephalopathy
- Aetiology is seronegative hepatitis, or a drug-induced hepatitis
- Prothrombin time over 50 seconds
The MELD criteria is a slightly younger system, and some authors feel strongly enough about its superiority to title their articles to that effect. One can find a nice article in UpToDate about the various uses of this system, and the modifications made to it over the years.
The MELD score is the sum of logarithms:
- loge(bilirubin) x 3.8
- loge(INR) x 11.2
- loge(Creatinine) x 9.6
The lower limit is 6, and the upper limit is 40, even though the scale can go from negative digits to infinity. The scores try to predict 90 day survival; thus if you score 9 or less, you have a roughly 1.9% chance of dying, whereas if you score 40, your 90 day mortality is 71.3%.
What is the point of all this, you might ask?
Well. Prognostic stratification in acute liver failure is very important, as it helps identify patients who will benefit from a transplant. The decision to transplant a liver is far from trivial, and carries with it a serious commitment to a course of hideous complications. Indeed it would seem that the acute liver patients with a low MELD score might actually be harmed by the process, and would be better served by a conservative medical management strategy. This idea is well developed: increasing MELD score tends to parallel increasing survival benefit from transplantation.
In short, having a high MELD score or satisfying the King's College Criteria tends to put you closer to the front of the line for liver transplantation.
Using the King's College Criteria to triage liver transplant recipients
LITFL has an excellent synopsis of the problems with using the King's college criteria to guide transplantation decisions, based on some recent reviews of the issue. As a brief summary, the criteria fail to accurately identify people who might die without a transplant, and there are major differences between the outcomes mentioned by the King's College liver transplant unit and the experience of other centres.
In spite of these flaws, the American Association for the Study of Liver Diseases still supports the use of King's College criteria, according to their most recent (2011) position paper. Their decision to continue with their support of this system rests on its relatively strong predictive performance. Sensitivity is in order of 68%, and specificity 88%. People who satisfy the King's College criteria tend to have a mortality in excess of 85% without transplantation; however because of low sensitivity up to 30% of patients who will die without a transplant may not satisfy the criteria, and thus never get listed.