Decision and economic evaluation of abortion availability in the United States military

Background: Active-duty service women rely on the civilian sector for most abortion care due to limits on federal funding for abortion. Abortion is now banned in many states with large military presences. The Department of Defense has implemented policies to assist active-duty service women in accessing abortion, but there is debate to reverse this support.

Objective: Our goal was to compare the cost-effectiveness and incidence of adverse maternal and neonatal outcomes of a hypothetical cohort of active-duty service women living in abortion-restricted states comparing restricted abortion access (abortion not available cohort) to abortion available with Department of Defense travel support (abortion available cohort).

Study design: We developed a decision tree model to compare abortion not available and abortion available cohorts for active-duty service women living in abortion-restricted states. Our cohorts were subdivided into normal pregnancies and those with a major fetal anomaly. Cost estimates, probabilities, and disability weights of various health conditions associated with abortion and pregnancy were obtained and derived from the literature. Effectiveness was expressed in disability-adjusted life years and the willingness-to-pay threshold was set to $100,000 per disability-adjusted life year gained or averted. We completed probabilistic sensitivity analyses with 10,000 simulations to test the robustness of our results. Secondary outcomes included numbers of stillbirths, neonatal deaths, neonatal intensive care unit admissions, maternal deaths, severe maternal morbidities, and first and second trimester abortions.

Results: The abortion not available cohort had a higher annual cost to the military ($299.1 million, 95% confidence interval 239.2-386.6, vs $226.0 million, 95% confidence interval 181.9-288.5) and was associated with 203 more disability-adjusted life years compared to the abortion available cohort. The incremental cost-effectiveness ratio was dominant for abortion available. Abortion not available resulted in an annual additional 7 stillbirths, 1 neonatal death, 112 neonatal intensive care unit admissions, 0.016 maternal deaths, 24 severe maternal morbidities, 27 less second trimester abortions, and 602 less first trimester abortions. Probabilistic sensitivity analysis revealed that the chance of the abortion available cohort being the more cost-effective strategy was greater than 95%.

Conclusion: Limiting active-duty service women's access to abortion care increases costs to the military, even with costs of travel support, and increases adverse maternal and neonatal outcomes. This analysis provides important information for policymakers about economic and health burdens associated with barriers to abortion care in the military.