UC San Diego

Background

Prenatal detection (PND) has benefits for infants with hypoplastic left heart syndrome (HLHS) and transposition of the great arteries (TGA), but associations between sociodemographic and geographic factors with PND have not been sufficiently explored. This study evaluated whether socioeconomic quartile (SEQ), public insurance, race and ethnicity, rural residence, and distance of residence (distance and driving time from a cardiac surgical center) are associated with the PND or timing of PND, with a secondary aim to analyze differences between the United States and Canada.

Methods

In this retrospective cohort study, fetuses and infants <2 months of age with HLHS or TGA admitted between 2012 and 2016 to participating Fetal Heart Society Research Collaborative institutions in the United States and Canada were included. SEQ, rural residence, and distance of residence were derived using maternal census tract from the maternal address at first visit. Subjects were assigned a SEQ z score using the neighborhood summary score or Canadian Chan index and separated into quartiles. Insurance type and self-reported race and ethnicity were obtained from medical charts. We evaluated associations among SEQ, insurance type, race and ethnicity, rural residence, and distance of residence with PND of HLHS and TGA (aggregate and individually) using bivariate analysis with adjusted associations for confounding variables and cluster analysis for centers.

Results

Data on 1862 subjects (HLHS: n=1171, 92% PND; TGA: n=691, 58% PND) were submitted by 21 centers (19 in the United States). In the United States, lower SEQ was associated with lower PND in HLHS and TGA, with the strongest association in the lower SEQ of pregnancies with fetal TGA (quartile 1, 0.78 [95% CI, 0.64-0.85], quartile 2, 0.77 [95% CI, 0.64-0.93], quartile 3, 0.83 [95% CI, 0.69-1.00], quartile 4, reference). Hispanic ethnicity (relative risk, 0.85 [95% CI, 0.72-0.99]) and rural residence (relative risk, 0.78 [95% CI, 0.64-0.95]) were also associated with lower PND in TGA. Lower SEQ was associated with later PND overall; in the United States, rural residence and public insurance were also associated with later PND.

Conclusions

We demonstrate that lower SEQ, Hispanic ethnicity, and rural residence are associated with decreased PND for TGA, with lower SEQ also being associated with decreased PND for HLHS. Future work to increase PND should be considered in these specific populations.