Pediatrics and Neonatology (Mar 2022)
Cost-effectiveness analysis of extended extracorporeal membrane oxygenation duration in newborns with congenital diaphragmatic hernia in the United States
Abstract
Background: The duration of extracorporeal membrane oxygenation (ECMO) has been historically confined in many centers to two weeks. We evaluated the cost-effectiveness of additional weeks on ECMO beyond two weeks for newborns with congenital diaphragmatic hernia (CDH) who may require longer stays to maximize survival potential. Methods: We modeled lifetime outcomes using a decision tree from the US societal perspective. Survival at discharge, probability of long-term sequelae, direct medical costs, indirect costs, and quality-adjusted life years (QALY) for long-term disability were considered. Considering the nature of severity of CDH, we used $200,000 per QALY as the willingness-to-pay threshold in the base case. Results: The lifetime costs per CDH infant generated from staying on ECMO for ≤2 weeks, 2–3 weeks, and >3 weeks are $473,334, $654,771, $1,007,476, respectively (2018 USD), and the total QALYs gained from each treatment arm are 1.83, 3.6, and 5.05. In the base case, the net monetary benefits are −$108,034 for ECMO ≤2 weeks, $64,258 for 2–3 weeks, and $2955 for >3 weeks. In probabilistic simulations, a duration of ≤2 weeks is dominated by a duration of 2–3 weeks in 65.3% of cases and dominated by > 3 weeks in 60.2% of cases. A duration of 2–3 weeks is more cost-effective than >3 weeks in 68.6% of simulations. Conclusion: Our findings suggest that 2–3 weeks of ECMO may be the most cost-effective for CDH infants that are unable to wean off at 2 weeks from the US societal perspective. Regardless of ECMO duration, ECMO use generates positive incremental NMB at WTP of $200,000 if the survival probability is greater than 0.3. Future research must be conducted to evaluate the long-term outcomes and sequelae of CDH patients post-discharge to better inform the clinical decision-making in neonatal intensive care unit.
