What Helps Us Decide to Adopt an Intervention: Efficacy, Costs, or Both?

In this issue of Pediatrics, Huang et al¹  conduct an economic evaluation analyzing data from the High-Flow Use in Non-Tertiary Centers for Early Respiratory Distress (HUNTER) trial²  comparing the cost-effectiveness of “bubble” continuous positive airway pressure (CPAP) with nasal high-flow therapy (nHF) in infants ≥31 weeks’ gestation and ≥1200 g receiving care in Australian nontertiary special care nurseries (comparable to level II nurseries in the United States). Using a health care sector perspective, which only includes costs of hospitalization and transport to another facility (as opposed to a societal perspective, which would include costs to all stakeholders, including family), they assessed the outcomes of need for endotracheal intubation and transfer to a tertiary-level NICU when using CPAP versus nHF under 2 scenarios: one with rescue CPAP available if treatment failure occurred with nHF and the other without the option of rescue CPAP. In the study, Huang et al¹  were able to manage infants from birth to 12 months of age and assess inpatient medical costs as well as interhospital transfers. The authors concluded that CPAP was more effective and cheaper when there was no back-up CPAP available for infants on nHF. In scenarios in which rescue CPAP was available, however, there was no difference in costs or effectiveness outcomes between the primary use of CPAP versus nHF.¹ 

In a recent systematic review, researchers found that <2% of neonatal randomized controlled trials (RCTs) published in the last 3 decades have included economic evaluations.³  Therefore, we applaud the authors for their economic study, with which they provide a guide for resource allocation decisions in nontertiary NICUs. There are several advantages to performing economic evaluations alongside RCTs, such as decreased susceptibility to bias, given the benefits of randomization and real-time cost estimates that are internally valid. The external validity of an economic evaluation may vary depending on the specific patient population and baseline risk of disease as well as institutional practices and available resources.

The authors undertake a transparent and comprehensive approach for this cost-effectiveness analysis, following the recommended guidelines for evaluating such a trial.⁴  Remarkably for a neonatal study, they were able to track hospital costs over 12 months of life for 95% of infants in the study, perhaps a reflection of the health care infrastructure within Australia. One feature of the economic analysis that is particularly important is the family perspective. As Huang et al¹  pointed out, the effectiveness outcomes were chosen to reflect not only resource use but also relevant outcomes to families and clinicians. The authors were explicit in using a health care sector perspective, although they were able to include some nonmedical costs, such as travel, hospital parking, and accommodations. This viewpoint, however, excludes other out-of-pocket costs that may be burdensome to caregivers, such as time away from work or child care. Quantifying these costs may be challenging in a clinical trial, and there is a paucity of data published on this topic, particularly for NICU families. With a societal perspective, researchers would have the additional benefit of including all relevant costs regardless of who is paying and would broadly take into account the interests of a population at a national level. The estimated societal costs of prematurity in the United States, for example, is ∼$25.2 billion per year, considering medical care, delivery services, early intervention, special education, and lost productivity.⁵ 

Furthermore, in the subgroup analysis in which the researchers evaluated the geographic distance from tertiary-level NICUs, they found that CPAP was the dominant strategy when the nearest NICU was ≥70 km away, highlighting, per the authors, the role of the perception of risk for nontertiary units. When considering remote areas in low- and middle-income countries with limited resources, CPAP as the primary respiratory support would seem more appealing, but the authors note that the results should be interpreted with caution, given that CPAP expertise may differ from the participating hospitals in the RCT. The idea of preventing transfers of neonates to higher levels of care is particularly appealing, especially when it can be done safely. Respiratory distress requiring mechanical ventilation is frequently the reason for transfer to a higher level NICU. Thus, it is important to evaluate interventions that reduce the need for transfer, such as optimal noninvasive respiratory support (as studied in the HUNTER trial) and perhaps other interventions, like the recently studied aerosolized surfactant.⁶  In addition, preventing the transfer of moderate and late preterm infants may result in significant cost savings, given that this subgroup constitutes the majority of preterm births.⁷  Although morbidity and mortality is relatively low among infants born at 32 to 36 weeks’ gestation, they are more likely to require respiratory and nutritional support, compared with term infants.⁸  Moreover, because they account for the majority of preterm births, moderate to late preterm infants have estimated societal costs of $10.6 billion through 18 years of age, compared with $5.5 billion for infants of an extremely low gestational age.⁹ 

The ultimate question with evaluating the results of the HUNTER trial and the economic evaluation is how do we (as clinicians, health care leaders, and families) decide on adoption of an intervention? The HUNTER trial has some interesting efficacy data with primary and secondary outcomes. There are clear financial implications for the hospital, payers, and family that are associated with transfer to a higher level of neonatal care. The transparent approach of the HUNTER trial economic evaluation does allow for the said stakeholders to tailor the decision to their own health care setting. To that end, we must point out 1 aspect of the cost-effectiveness ratio interpretation. The authors did not explicitly state the incremental cost-effectiveness ratio, which is the difference in costs divided by the difference in outcomes, when comparing CPAP with nHF in the trial scenario. The incremental cost-effectiveness ratio for each NICU transfer avoided is negative because nHF costs more than CPAP and is less effective, but this requires context for interpretation. However, the probabilistic sensitivity analysis for rescue CPAP, which varies each model input across a distribution, is intriguing as shown in the cost-effectiveness plane, whereby the x-axis represents incremental effectiveness and the y-axis represents incremental costs. The nearly even distribution among all 4 quadrants of the cost-effectiveness plane suggests that 4 different scenarios are equally likely: rescue CPAP may cost more but also be more effective, cost more with worse outcomes (“dominated”), cost less with worse outcomes, or cost less but result in better outcomes (“dominant”).¹⁰  In contrast, the probabilistic sensitivity analysis for CPAP as the sole primary support clearly reveals that the majority of dots are in the right lower quadrant, indicating that CPAP is less costly and more effective.

Overall, the authors should be commended for recognizing the realities of constrained resources and limited budgets in health care, as well as incorporating the family perspective. With their study, they allow decision-makers to concurrently evaluate costs and effects of 2 commonly used treatment strategies in the NICU and help us understand the value in health care.¹¹  Given increasing budgetary constraints in health care, the economic results of this analysis, coupled with the efficacy data, suggest to us that nontertiary special care nurseries should consider using solely CPAP as the primary mode of respiratory support because it is more effective and cheaper than nHF. Future RCTs should include funding for economic evaluations to identify high-value interventions that optimize neonatal health, reduce costs, and improve the experience of families.¹² 

CPAP

continuous positive airway pressure

nHF

nasal high-flow therapy

RCT

randomized controlled trial