D-MELD does not predict post-liver transplantation survival: a single-center experience from Brazil

Abstract

Background. The D-MELD score was designed to prevent donor-recipient matches with a high risk of unfavorable outcome. The main objective of the present study was to assess the predictive value of the D-MELD score for 1-month and 3-month post-transplant mortality in a cohort of patients who underwent deceased-donor liver transplantation in Southern Brazil.Material and methods. A cohort study was conducted. Receiver operating characteristic c-statistics were used to determine the ability of the D-MELD score to predict mortality. The Kaplan-Meier method was used to analyze survival as a function of time regarding D-MELD scores, and the Cox model was employed to assess the association between D-MELD and mortality.Results. Most recipients were male, with a mean age of 54.3 ± 9.6 years (n = 233 transplants). Mean donor age was 44.9 ± 16.8 years (19.3% of donors were aged ≥ 60 years). Mean MELD and D-MELD scores were 16.3 ± 7.1 and 733.1 ± 437.8 respectively. Overall survival at 1 and 3 months was 83.6%. The c-statistic value for 1- and 3-month mortality was < 0.5 for the D-MELD. Analysis of Kaplan-Meier curves for groups with D-MELD scores < 1,600 and ≥ 1,600 did not show statistically significant differences in survival (p = 0.722).Conclusion. D-mElD scores were unable to predict survival in this cohort of Brazilian liver transplant recipients.

Keywords

• Prognostic models
• Liver transplant
• Outcome