Risk Index Predicts Pediatric Heart Allograft Non‐Utilization

Abstract

AbstractBackgroundChildren listed for heart transplantation face the highest waitlist mortality among all solid organ transplant patients (14%). Attempts at decreasing donor allograft non‐utilization (41.5%) could potentially decrease waitlist mortality for pediatric heart transplant patients. Our aim was to quantify the non‐utilization risk of pediatric donor heart allografts at the time of initial offering.MethodsUsing the United Network of Organ Sharing (UNOS) database, we retrospectively analyzed 8823 deceased donors (≤18 years old) data through univariable and multivariable analysis and logistic regression models. These factors were divided into a training (n = 5882) and validation set (n = 2941). Donor clinical characteristics and laboratory values were used to predict non‐utilization of donor hearts. The multivariable analysis used factors that were significant from the univariable analysis (p‐value < .05), and the pediatric non‐utilization risk index (pDRSI) included significant factors from the multivariable analysis, producing an overall risk score for non‐utilization. With these data, we created a non‐utilization risk index to predict likelihood of donor allograft non‐utilization.ResultsFrom the 24 potential factors that were identified from univariable analysis, 17 were significant predictors (p < .05) of pediatric heart non‐utilization in the multivariable analysis. Low left ventricular ejection fraction (odds ratio (OR)‐35.3), hepatitis C positive donor (OR‐23.3), high left ventricular ejection fraction (OR‐3.29), and hepatitis B positive donor (OR‐3.27) were the most significant risk factors. The phDSRI has a C‐statistic of 0.80 for the training set and 0.80 for the validation set.ConclusionUsing over 8000 donors, the phDSRI uses 17 significant risk factors to predict risk of pediatric heart donor allograft non‐utilization.