A Recipient Risk Prediction Tool for Short-term Mortality After Pediatric Heart Transplantation

Swati Choudhry, Yunfei Wang, Susan W. Denfield, Antonio G. Cabrera, Jack F. Price, Hari P. Tunuguntla, Vikas R. Dharnidharka, Charles E. Canter, William J. Dreyer

Research output: Contribution to journalArticlepeer-review

1 Scopus citations

Abstract

BACKGROUND: The first year after heart transplantation (HT) has the highest risk of mortality. We aim to derive and validate a recipient risk prediction tool for early mortality after pediatric HT. METHODS: The International Society for Heart and Lung Transplantation (ISHLT) registry was used to identify patients (≤18 y) who underwent primary HT during January 2000-December 2014. Independent predictors of 1-year mortality were identified based on recipient characteristics at HT. Risk scores were assigned based on the magnitude of relative odds of 1-year mortality. The predictive capability of the ISHLT registry derived recipient risk score was externally validated using the Scientific Registry of Transplant Recipients registry data from 2015 to 2017 to ensure a cohort of patients completely exclusive from the derivation cohort. RESULTS: A total of 5045 eligible patients were included in the analysis. The 20-point risk scoring system incorporated 8 recipient variables, including age at HT, diagnosis, pre-HT ventilator use, extracorporeal membrane oxygenation, inhaled nitric oxide use, infection, estimated glomerular filtration rate, and serum bilirubin. Compared with low-risk score group, high-risk group had 7-fold increased risk of 1-year mortality (hazard ratio 7.4; 95% CI [5.2-9.1]; P < 0.001). The C-statistics (0.77) and Hosmer-Lemeshow goodness of fit (0.9) for recipient risk score using derivation cohort from ISHLT registry performed well and was similar to the internal and external validation cohort (C-statistics 0.75, 0.78 and Hosmer-Lemeshow goodness of fit P = 0.4, 0.3, respectively). CONCLUSIONS: This study derived and externally validated a simple risk predictive model based on recipient characteristics at HT that has good prediction characteristics for 1-year post-HT mortality. This model may help clinicians identify candidates who are at a higher risk for post-HT mortality and may optimize organ sharing.

Original languageEnglish
Pages (from-to)2434-2439
Number of pages6
JournalTransplantation
Volume103
Issue number11
DOIs
StatePublished - Nov 1 2019

Fingerprint Dive into the research topics of 'A Recipient Risk Prediction Tool for Short-term Mortality After Pediatric Heart Transplantation'. Together they form a unique fingerprint.

Cite this