Critical Analysis of the Pediatric End‐Stage Liver Disease Scoring System: A Single Center Experience

Critical Analysis of the Pediatric End-Stage Liver Disease Scoring System: A Single Center Experience. Shneider BL, Neimark E, Frankenburg T, Arnott L, Suchy FJ, Emre S. Liver Transplant 2005:11:788-795. Summary: In 2002, a new system for donor liver allocation was introduced by the United Network for Organ Sharing (UNOS) to determine medical urgency for children and adults awaiting liver transplantation. The scoring system was developed by the liver transplant community to comply with a directive from the Department of Health and Human Service (DHHS). The directive indicated that organs should be first offered to the sickest patients but not futile cases and that the system for organ allocation should be based on objective clinical criteria, not waiting time. Consequently, the Pediatric End-Stage Liver Disease (PELD) scoring system was developed and implemented in February, 2002 to assess the risk of death to allocate organs according to the principles outlined by the DHHS. The PELD scoring algorithm is based on total bilirubin, International Normalized Ratio (INR), albumin, evidence of growth failure, and age less than 1 year. Centers may request addition of exception PELD points from their regional review board if the providers judge that the calculated PELD score does not reflect the severity of the patient's disease. In the present study, Shneider et al. (1) evaluated the impact and effectiveness of the PELD scoring system. To do so, the authors retrospectively analyzed 48 children who underwent liver transplantation or died awaiting transplantation between February 27, 2002 and February 26, 2004 at a single liver transplant center. In addition to the elements of the PELD scoring system, they also recorded serum sodium and the presence of gastrointestinal bleeding, ascites, spontaneous bacterial peritonitis, cholangitis, hepatopulmonary syndrome, severe pruritus, and pathologic bone fractures. The rationale for requesting addition of exception points to the calculated PELD score and for refusal of an organ offer were recorded. The disease distribution was typical for a large pediatric liver transplant center. Biliary atresia accounted for 31% of listed children, acute liver failure for 25%, and chronic cholestasis for 13%. Of the 48 children who underwent liver transplantation, 25 were excluded from PELD analyses, 4 because they were older than 17 years, 10 because the patient had acute liver failure, and the remaining 11 for a variety of reasons. Of the remaining 23 patients, 4 children were transferred to the intensive care unit and were made status 1 based on the severity of their disease. The remaining 19 patients underwent liver transplantation based on their PELD score; 17 of 19 had PELD score based on exception points. During the time period reviewed, four listed children died without transplant. Two of the children, one with mitochondrial disease and one with severe portopulmonary syndrome, were withdrawn from the list. One child died of brain stem herniation as a complication of acute liver failure, and the other child died of overwhelming sepsis. Shneider et al. conclude that the PELD scoring system underestimates short-term risk of mortality and that without the potential for exception points, the death rate for children awaiting liver transplantation would exceed that for adults. Comments: In 1998, the U.S. DHHS released a policy that included the final rule intended to improve the effectiveness and equity of the nation's organ transplantation system. The focus of the proposed final rule was to allocate organs to those patients in most urgent medical need within a geographic radius of potential delivery on the basis of the cold ischemic times for liver viability. The proposed final rule attempted to eliminate geographic disparity in organ allocation “so that neither place of residence nor place of listing shall be a major determinant of access to a transplant.” Congress asked the Institute of Medicine (IOM) to examine the potential consequences of implementation of the final rule (2). In its 1999 report, the IOM stated that despite the best efforts of all of the liver transplant community, demand exceeded the supply of available organs and that for patients and providers, the system was confusing and difficult to understand. Both the IOM and the DHHS were in agreement that donor organs should be allocated on the basis of medical need and acknowledged the need for a measure of medical urgency for liver transplantation candidates. From this series of events, the medical end-stage liver disease (MELD) and PELD scoring systems were born. Because the UNOS database did not capture sufficient detail for pediatric patients from which to develop an objective scoring system, the data for development of the PELD scoring algorithm were retrieved from the Studies of Pediatric Liver Transplantation registry (3). From this registry, which includes 39 centers from the United States and Canada, 884 children with chronic liver disease listed for their first liver transplant as of June 15, 2000 were included in the analyses. Forty-five percent were less than 1 year of age at the time of listing, and 46% were diagnosed with biliary atresia. Seventeen factors that might impact on poor outcome before transplantation were identified, and six factors were found to be objective, verifiable, and not event based. The outcomes assessed were death pretransplant and transfer to the intensive care unit pretransplant. In univariate analysis, age, total bilirubin, INR, and albumin were significant predictors (P < 0.01) of both outcome endpoints. Glomerular filtration rate estimated by the Schwartz formula was not a significant predictor of either outcome, and growth failure was significant for death/moved to intensive care unit outcome but not for death alone. In multivariate analyses, age less than 1 year, serum albumin concentration, total serum bilirubin concentration, INR, and growth failure were identified as predictors of outcome. With this background in mind, what can we conclude from the results presented and the questions posed by Shneider et al. (1)? In the IOM report, it was suggested that the success of the scoring system be measured by outcomes. With the implementation of PELD, the death rate for children awaiting liver transplantation decreased. However, the proportion of children who were transplanted as status 1 by exception increased from 23% to 29%, and there was significant variance among regions (4). So, by just looking at risk of death on the waiting list, we might see improvement over the pre-PELD period but not when assessing other measures of success including geographic variation. What is not addressed by these measures is the proportion of patients for whom the PELD scoring system was used to allocate organs. Shneider et al. report that their team has aggressively sought exception points. Only 2 of 22 patients who received organs were allocated from a decreased donor according to calculated PELD scores. The authors raise concerns that without exception points, children may die awaiting liver transplantation. In summary, the report by Shneider et al. gives us insight into the impact of the newly devised PELD scoring algorithm. It also raises several questions. Does the failure of the the PELD scoring algorithm to predict risk suggested by Shneider et al. reflect study bias because transfer to the intensive care unit was used as an outcome measure? Might the model be improved by inclusion of additional measures such as serum sodium? What is a realistic prediction for the proportion of children who will be allocated organs as a result of their calculated PELD score given that children are listed with disorders (tumors, metabolic liver disease, hepatopulmonary syndrome) who typically have low calculated PELD scores? Because adults and children compete for organs, we must ask whether MELD and PELD scores represent equivalent risks for death before transplant. Furthermore, we must recognize that the algorithm was based on past patient populations and not on current or future populations. As a result, as the patient population changes, the model will need to be reassessed. In the end, we must recognize that allocation of organs is a dynamic process and requires ongoing review and a collaborative effort by members of the transplant community. PELD is an excellent first step, and reports like the one of Shneider et al. should help us to improve this system. John C. Bucuvalas Cincinnati Children's Hospital Medical Center Cincinnati, OH [email protected]

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.