TO THE EDITORS:In a recent retrospective analysis, Sundaram et al. 1 used a prospectively collected database of 1798 recipients of deceased donor livers from the eras before and after the implementation of the Model for End-Stage Liver Disease (MELD) to evaluate the effects of the MELD score on the development of biliary complications. We believe that these data are important and highlight some of the limitations of the MELD score with respect to specific outcomes after liver transplantation.Two issues perhaps could be clarified. After they evaluated more than 20 potential donor, recipient, and technical risk factors, Sundaram et al.1 found that the donor age, bile leaks, duct-to-duct anastomoses, and the MELD era were independent factors for the development of anastomotic biliary strictures. Although some intraoperative and perioperative factors were evaluated, information about the total surgical time, the intraoperative transfusion requirements, and the need for reoperation in the early postoperative period was lacking in the analysis. The aforementioned risk factors play important roles in the development of anastomotic biliary strictures and may also be influenced by the MELD score. 2,3 In addition, the use of T-tubes may influence bile leaks and the development of biliary strictures. 4 The inclusion of this variable in the final multivariate model may have affected the association of bile leaks and the development of anastomotic biliary strictures.The other issue is the fact that virtually all bile leaks (97%) were found before the diagnosis of an anastomotic stricture (82 patients in the control group and 22 patients in the stricture group). These findings are puzzling because a significant number of the bile leaks (nearly half) were diagnosed 3 months after liver transplantation. It would be interesting to know more about the nature of these leaks because the majority of bile leaks appear within the first 3 months after liver transplantation in clinical practice.