Background The Model for End Stage Liver Disease (MELD) predicts mortality in end stage liver disease. Incorporation of serum sodium into the MELD may improve diagnostic accuracy in decompensated patients with ascites. However, other complications of cirrhosis are not reflected. This study investigates whether quantitative liver function tests predict survival and increase prognostic accuracy of the MELD.

Methods 604 patients with suspected cirrhosis were staged clinically and haemodynamically. Galactose-elimination-capacity, sorbitol clearance, lidocaine metabolism and indocyanin green (ICG) half life were determined. Survival was the primary end point of the study. Prognostic effects of individual parameters were calculated using Cox regression models and ROC curves.

Results 321 patients on standard pharmacological and endoscopic treatment (PET) and 74 patients undergoing transjugular portosystemic shunting (TIPS) were studied. Of all quantitative liver function tests, ICG half life was the most accurate in predicting survival. Upon incorporation into the MELD, it modified the score in patients with PET up to 35 points. Clinically relevant changes to the score, however, occurred in patients with a MELD score between 10 and 30, allowing an objective prognostic discrimination of individual survival based on laboratory liver function and blood flow. The MELD-ICG was validated in the second cohort of patients undergoing TIPS implantation.

Conclusion ICG had the highest predictive value of the examined tests. Its incorporation into the MELD adds an estimation of liver blood flow and renders the new score MELD-ICG more accurate in predicting survival in intermediate to advanced cirrhosis than the MELD and MELD-Na.