At ISPOR 2025 in Montreal, PurpleLab presented a research poster titled “Intersectional Multilevel Analysis of Individual Heterogeneity and Discriminatory Accuracy (MAIHDA) for Social Determinants of Health and Pregnancy Outcomes.”
The study, conducted by Anna Wang, PhD, Francis Concannon, BS, Daria Eremina, PhD, and Douglas Londono, PhD, applied MAIHDA to analyze how overlapping social factors influence pregnancy outcomes using a large U.S. healthcare claims dataset.
Study overview:
- Analyzed pregnancy outcomes (full-term delivery, miscarriage, termination) among women aged 18–49 from 2014 to 2024.
- Six social variables were used to create intersectional strata: age, race/ethnicity, education, marital status, Medicaid enrollment, and U.S. region.
- MAIHDA was used to examine how these factors, in combination, correlate with different pregnancy outcomes.
Key findings:
- Certain combinations of social factors are linked to a higher likelihood of miscarriage or termination than would be expected from looking at single factors alone.
- MAIHDA helps identify disparities that conventional models often overlook.
- The study reinforces that health outcomes are shaped by the intersections of social positions, not isolated traits.
The analysis reflects PurpleLab’s ongoing work to apply advanced methodologies to better understand structural influences on health outcomes.
See the full research poster below for more detail.
