System-Level Predictors of Long-Acting Reversible Contraception Provision in the Veterans Health Administration

To examine the provision of long-acting reversible contraceptive (LARC) methods across the Veterans Health Administration's (VA) 140 regional healthcare systems and investigate system-level correlates of low provision as an indicator of potential access barriers.

We conducted a cross-sectional analysis of national VA electronic health record (EHR) data. For each regional healthcare system, we calculated the percentage of pregnancy-capable Veterans who received a LARC method (intrauterine device or contraceptive implant). We categorized healthcare systems in the bottom quartile as low-provision. We examined associations between low-provision and system-level factors, including gynecologist staffing per pregnancy-capable Veteran, Women's Health Medical Director protected time, percent of pregnancy-capable Veterans visiting a women's health clinic, and LARC provision at ≥ 1 community-based outpatient clinic (CBOC).

We performed a secondary analysis of EHR data for female pregnancy-capable Veterans ages 18-44 who visited VA primary care or gynecology in 2019. We evaluated associations with chi-squared tests and multivariable logistic regression adjusting for Veteran-level factors.

The median percentage of Veterans receiving LARC methods across healthcare systems was 4.9%, varying from 0% to 12.0%. In multivariable modeling, each 5% increase in gynecologist half-days per 100 pregnancy-capable Veterans was associated with an average two-percentage point decrease in the probability of being a low-provision system (average marginal effect [AME] = -0.02, 95% CI: -0.02, -0.01). LARC provision at ≥ 1 CBOCs was associated with an average 17-percentage point decrease in the probability of being a low-provision system (AME = -0.17, 95% CI: -0.29, -0.05).

We found significant variation in LARC provision across the VA's 140 regional healthcare systems. Importantly, this EHR analysis is limited as it does not incorporate patient demand for methods. Our findings, however, indicate potential access barriers. Interventions, such as increasing gynecologist staffing and investing in LARC provision in CBOCs, could help ensure access to these methods.