Do MELD 3.0 have better discrimination of mortality in post-liver transplant?

Introduction: The Model for End-Stage Liver Disease (MELD) 3.0 has better discrimination of waitlist mortality in end-stage liver disease than MELDNa in the United States. This study aimed to assess MELD3.0 in post-liver transplant (LT).
Methods: We analyzed data from United Network for Organ Sharing (UNOS) between 2016-2021. We included adult(>=18 years) LT recipients and excluded patients with multi-organ transplant, re-LT, and MELD exceptions.
Results: 31,016 liver transplants were analyzed. Median age (IQR) was 56 (47-63), male was 62%, BMI was 28.3 (24.6-32.7), MELD 27 (19-34). The number of living-donor was 2,006 (6.5%) and the majority was deceased-donor (N=29010, 93.5%) Out of 29,010, brain-death-donor (DBD) was 91% (N=26427), and donation-after-cardiac (DCD) was 8.9% (N=2583). The most common indications for LT were alcohol-related liver disease (39.5%) and non-alcoholic steatohepatitis (22.4%). MELD3.0 could discriminate 30-days mortality (C-statistics 0.564; 95%CI 0.541-0.587, P<0.001), but not 90-days mortality (C-statistics 0.566; 95% CI 0.505-0.627, P=0.054). MELD3.0 had worse discrimination than MELDNa (C-statistics 0.564 vs 0.576). Based on primary indications, MELD3.0 could discriminate 30-day mortality except for viral hepatitis/cirrhosis, but did not have better discrimination than MELDNa. MELD3.0 had better discrimination in DCD than MELDNa (C-statistics 0.597 vs 0.580) and could not discriminate in living-donor. Conclusions: MELD3.0 has similar discrimination of postoperative 30-days mortality, but low predictive values.