Objective: Two recent randomized controlled trials (RCTs) reported that inhaled nitric oxide (iNO) decreased the incidence of extracorporeal membrane oxygenation (ECMO) or death in term and near-term newborns with hypoxic respiratory failure. Our objective was to estimate the cost-effectiveness ratio of iNO in this population.
Methods: We studied 1000 simulation cohorts (n = 483 for each cohort) of term/near-term newborns with hypoxemic respiratory failure. We conducted our study following US Public Health Service Panel on Cost-Effectiveness in Health and Medicine guidelines, adopting the US societal perspective. We constructed a decision tree reflecting iNO use, subsequent ECMO use, death, and long-term neurologic and respiratory morbidity in survivors, as determined from the combined outcomes of the 2 RCTs (n = 483). We estimated costs on the basis of length-of-stay data for the initial episode of care from 1 of the RCTs, unit costs from administrative data sets, and current pricing for iNO. We ran a Monte Carlo simulation to generate estimates of differences in costs and effects at 1 year, along with the stochastic uncertainty around these estimates. We expressed effects as quality-adjusted survival, assuming quality of life = 1 with no comorbidity, 0.7 with 1 comorbidity, and 0.49 (0.7 x 0.7) with 2 comorbidities. We constructed a base case, in which iNO was initiated at tertiary care ECMO centers (mimicking the RCTs) and a Public Health Service Panel on Cost-effectiveness in Health and Medicine reference case, in which iNO was initiated at the local hospital before transfer (mimicking real-world practice). We exposed our assumptions to a sensitivity analysis.
Results: Direct application of the trial results (base case) suggested that iNO was both more effective and cheaper (cost savings of 1880 dollars per case despite acquisition costs of 5150 dollars, predominantly as a result of decreased need for ECMO), with 84.6% probability that the cost-effectiveness ratio was better than 100,000 dollars per quality-adjusted life-year. Under the reference case, iNO was also more effective (though slightly less so) and was even cheaper (cost savings of 4400 dollars per case), with 71.6% probability that iNO was cheaper and more effective and 91.6% probability that the cost effectiveness ratio was better than 100,000 dollars per quality-adjusted life-year. Sensitivity analyses showed these estimates to be sensitive to patient selection and the price of iNO but insensitive to assumptions regarding quality of life.
Conclusions: From a US societal perspective, iNO has a favorable cost-effectiveness profile when initiated either at ECMO centers or at local hospitals in term/near-term neonates with hypoxemic respiratory failure.