Introduction: The relationships between prenatal diagnosis of congenital heart defects (CHDs) and outcomes are inconsistent. Some studies have shown that compared to postnatal diagnosis, prenatal diagnosis is associated with later surgical timing and worse outcomes. These findings could be due to confounding because more severe CHDs are likelier to be diagnosed on prenatal ultrasonography. To bypass this confounding, we evaluated prenatal diagnosis as a mediator. Aim: To evaluate whether prenatal diagnosis of CHD mediates the association between disease severity and timing to postnatal intervention. Hypothesis: Infants with severe CHD receive intervention sooner than those with mild CHD due to a higher prenatal diagnosis prevalence. Methods: This is a retrospective cohort study of all infants who received their initial cardiac catheterization or surgery between 0-60 days (i.e., critical CHD) at Ann&Robert H. Lurie Children’s Hospital of Chicago (2015-2021). Birthing individuals with inadequate prenatal care were excluded. The primary exposure variable was a three-level categorical variable derived from existing literature, categorizing disease severity based on anatomic factors (mild, moderate, and severe CHD). The outcome was age at intervention, i.e., cardiac catheterization or surgery, whichever occurred sooner, in days. The mediator variable was whether prenatal diagnosis occurred. Control variables included birth weight, sex, surgery during the COVID-19 pandemic, race, ethnicity, birth hospital’s level of perinatal care as designated by Illinois state, presence of a syndrome or non-cardiac abnormality, and insurance status. Results: In total, 429 parent-infant pairs met inclusion criteria; 68.3% had prenatal diagnosis and 28.4% had severe CHD. In adjusted analyses, compared to those with the mild CHD, those with moderate CHD had 40.0% higher probability of prenatal diagnosis (p<0.001) and on average, received intervention 14 days sooner (p=0.001). Those with severe CHD had a 59.5% higher probability of prenatal diagnosis (p<0.001), and on average received intervention 19 days sooner (p<0.001). In a mediation analysis, prenatal diagnosis accounted for 27.2% of the difference in surgical timing between moderate and mild disease, and 29.5% of the difference in surgical timing between severe and mild disease ( Figure 1 ). Conclusion: The relationship between CHD severity and timing to postnatal intervention is partly attributable to prenatal diagnosis.
Read full abstract