Abstract 12586: Infant Pre-Listing Probability of Survival Beyond the Early Post-Heart Transplantation Period
Background: Infants listed for heart transplantation (HTx) experience the highest waitlist mortality and important early post-HTx mortality. Donor organ shortages, especially for infants, require optimal allocation of this scarce resource. We sought to create a risk stratification model predicting, at the time of listing, which infants are at decreased likelihood of surviving at least 3 mos after HTx.
Methods: All infants (<15 mos old) listed for HTx between January 2005 and October 2010 were reviewed. Multivariate logistic regression models were used to determine factors at listing associated with survival to 3 mos after HTx. Based on this model the probability at listing of survival to 3 mos post-HTx was predicted for individual patients.
Results: A total of 48 infants were included; 13 died while waiting, 6 died within 3 mos of HTx and 29 survived >3 mos post-HTx. In multivariable logistic regression models, factors at listing associated with increased odds of mortality were cardiomyopathy (vs. all congenital heart disease) (OR:6.0, p=0.02), listing on ECMO support (OR:7.8, p=0.007), on renal replacement therapy (OR:71.4, p<0.001), lower blood pH (OR:1.16/0.01 unit, p=0.02) and higher alanine aminotransferase (OR:1.02/unit, p=0.02). No patients listed on renal replacement therapy (n=10) survived and only 3/12 (25%) patients listed on ECMO support survived >3 mos post-HTx. The resulting multivariable model for probability at listing of survival beyond 3 mos post-HTx correctly assigned patients to the actual outcome in 90% of the cases, suggesting excellent discrimination. Patients with a pre-listing probability <50% experienced 93% mortality (100% mortality in <35%) in comparison to 18% mortality in patients with probability >50% (97% specificity, 68% sensitivity). Patients who had a predicted probability of mortality <50% from the time of listing and who subsequently died had longer wait time on the list (55 days vs. 27 days).
Conclusions: Based on 5 pre-listing factors we were able to assign 90% of infants to their 3 month post-HTx outcome. The concept of pre-listing survival probability should be further validated and verified, and could eventually be incorporated into decision-making around listing and, importantly, delisting for infant HTx.
- © 2011 by American Heart Association, Inc.