|| Checking for direct PDF access through Ovid
Allocation policies for liver transplantation (LT) are worldwide based on MELD score. However, if on one side this score represents the cornerstone of priority process, on the other side poor evidence exists on its ability to predict post-LT death. As a consequence, new scores able to predict post-LT survival are required. The comprehensive complication index (CCI) calculated during the post-operative hospital stay represents an interesting predictor of long-term mortality. The aim of the present study is to firstly identify potential predictors of poor CCI, and then to compare CCI with other risk factors in terms of death prediction. A retrospective analysis based on 240 consecutive LT patients undergoing transplant at Sapienza University of Rome during the period 2008-2016 were evaluated. Median time after LT was 40 (IQR: 12-98) months. At ROC analysis, the best predictor of CCI≥50 was the MELD value at the moment of LT (AUC=0.695; p=0.001). UNOS status 1A, donor and recipient age, HCV-related cirrhosis or presence of hepatocellular cancer were not predictors of post-LT high CCI values. Testing CCI as a predictor of post-LT death, it was the unique statistically significant variable able to predict death, with an excellent 0.831 of AUC (p<0.0001). Nor MELD neither UNOS status 1A were similarly efficacious. Also after removing early (<3 months) post-LT deaths, CCI maintained its diagnostic ability (AUC=0.867; p<0.0001). As for post-LT survival, patients with CCI≥50 had a 5-year survival rate of 55.2% vs. 94.4% of cases meeting this threshold value (p<0.0001). Initial clinical course identified by CCI value strongly influences the entire post-LT story of the patient: CCI is an excellent predictor of death when compared with other risk factors. Studies aimed at externally validate this score are needed.