UC Irvine

Rationale: There are no validated bedside tools that can predict mortality risk of a neonate with CDH prior to initiation of ECMO and during the course of ECMO.

Objectives: To develop ECMO specific mortality risk prediction models for CDH.

Methods: The Extracorporeal Life Support Organization (ELSO) registry (2000-2015) was used to develop mortality prediction scores for CDH relative to timing of ECMO. Prediction models were developed using multivariable logistic regression models. Observed mortalities for the pre- and on-ECMO were further examined by five clinical risk groups defined by percentiles of the risk score.

Results: We identified 4,374 neonates with CDH with an overall mortality of 52%. Predictive discrimination (C-statistic) for pre-ECMO mortality model was C = 0.65 (95% CI: 0.62-0.68). Within the highest risk group, based on the pre-ECMO risk score, mortality was with 87% (144 neonates) and 75% (92 neonates), in the training and validation datasets, respectively. The pre-ECMO risk score included pre-ECMO ventilator settings, pH, prior DH repair, critical congenital heart disease, perinatal infection, and demographics. For the on-ECMO model, mortality prediction improved substantially: C = 0.73 (95% CI: 0.71-0.76) with the addition of on-ECMO associated complications and comorbidities. Within the highest risk group, defined by the on-ECMO risk score, mortality was 90% (147 neonates) and 86% (77 neonates) in the training and validation datasets, respectively. The post-ECMO mortality prediction model which accounted for the timing of CDH repair post-ECMO had the best predictive discrimination with C = 0.80 (95% CI: 0.78-0.82).

Conclusion: Mortality among neonates with CDH needing ECMO can be reliably predicted with validated clinical variables identified in this study relative to timing of ECMO.