O2063: Comparing the new and old American Academy of Pediatrics critical congenital heart disease screening algorithms: A cost-effectiveness analysis

Background: The US Department of Health and Human Services began recommending screening for critical congenital heart disease (CCHD) using pulse oximetry in 2011, and it was implemented by all states by 2018. Most states use a screening algorithm proposed by the American Academy of Pediatrics (AAP) in 2011, which has previously been shown to be cost effective. A new screening algorithm was recently approved by the AAP in December 2024 with two key changes: (1) elimination of the second retest and (2) saturation must be ≥95% for pre- AND post-ductal saturations instead of pre- OR post-ductal saturations. The relative cost effectiveness of these two algorithms has not been assessed. We therefore aimed to quantify the relative cost-effectiveness of the two algorithms.

Methods: We developed a decision tree and Markov cohort model using all children in the United States born in 2022 without a prenatal diagnosis of CCHD (N=3,664,739). The model consists of a decision tree accounting for the first year of life followed by a Markov model for the rest of the lifetime (Figure 1). The Markov cohort model had a one-year cycle length and a lifetime time horizon. Model parameters were taken from the published literature. All costs were inflation-adjusted to 2024 dollars. Costs and quality-adjusted life years (QALYs) were discounted at 3% per year. The primary health outcome was the incremental cost-effectiveness ratio (ICER) expressed in $/QALY. A willingness to pay threshold of $100,000 per QALY was used.

Results: In the base case, the new screening algorithm led to an increase in cost of $3.385 million and an increase of 57.92 QALYs yielding an overall ICER of $58,441.31. The new algorithm also identified an additional 146 children with CCHD for a cost per additional case of $23,136.84. There were 2.5 infant deaths averted under the new algorithm with a cost of $1.35m per death averted. In one-way sensitivity analyses, the new algorithm became dominant at high specificities (Figure 2). The old algorithm was dominant when the two algorithms had the same sensitivity. The new strategy is the preferred strategy at the willingness to pay threshold if the sensitivity of the new algorithm is 2.63 percentage points higher than the old algorithm. Probabilistic sensitivity analysis showed that the new algorithm had a 51% likelihood of being cost-effective at the willingness to pay threshold.

Conclusion: The new AAP screening algorithm is cost-effective when compared to the old algorithm. Although it is slightly more expensive, it also leads to more QALYs due to fewer hospital days in the first year of life and greater survival. There is substantial decisional uncertainty due to uncertainty about the relative sensitivities of the two algorithms.