The effect of contraceptive access reform on privately insured patients: Evidence from Delaware Contraceptive Access Now

Abstract

Background: Many states are implementing comprehensive programs aimed at reducing persistent barriers to contraceptive care. Evidence on the effectiveness of these programs is essential for practice improvement and policy development.

Objective: To evaluate changes in the probability of initiating a contraceptive method by women with employer sponsored insurance after implementation of Delaware Contraceptive Access Now (DelCAN), a statewide initiative that aimed to increase access to long-acting reversible contraceptives (LARCs).

Design, setting, and participants: We used a difference-in-differences design to examine contraceptive initiation rates. Data came from IBM Marketscan and covered women age 15-44 enrolled in employer sponsored insurance. The primary outcome was insertion of a LARC, both in the overall study population and in the immediate postpartum (IPP) setting. Secondary analysis examined changes to other contraceptive method types.

Results: The cohort of 4,550,459 enrollees generated a sample of 11,888,837 person-years and 615,670 childbirth hospitalizations. Difference-in-differences estimates suggested that DelCAN was associated with a 0.3 percentage point (95% CI [0.2, 0.5], p<0.001) increase in the LARC insertion rate in the overall study population and a 0.4 percentage point increase (95% CI [0.2, 0.6], p<0.001) in the percent of births adopting IPP LARC. Associations between DelCAN and LARC insertion appeared stronger for adolescents compared to older women. Results for other method types were less consistent.

Conclusions: A comprehensive statewide program was associated with increased LARC insertion rates among enrollees with employer sponsored insurance. Understanding the effect of these programs is critical for on-going policy development for states engaged in contraceptive access reform.

Fig 1. Average LARC insertion rates and IPP LARC insertion rates among women age 15–44 in delaware and control states.

Source: IBM Marketscan Commercial Claims and Encounters Database (2012–2019). Notes: The rate reported in the left-hand panel each year is calculated as the number of women having a LARC device inserted in that year, divided by the total number of women enrolled in a plan in the Marketscan database. The rate reported in the right-hand panel is calculated as the number of births in a hospital where a LARC device was provided during the dates of hospitalization, divided by the total number of births occurring in a hospital recorded in the Marketscan database for each year.

Fig 2. Difference-in-differences effect of the DelCAN program on LARC insertions, for women age 15–44 by age group and policyholder status.

Source: IBM Marketscan Commercial Claims and Encounters Database (2012–2019). Notes: Difference-in-differences estimates are from a linear probability model estimated with individual data at the person-year level for any LARC insertions and at the birth level for IPP LARC placements. Both the unadjusted and adjusted difference-in-difference estimates include state and year fixed effects. The adjusted model controls for individual level covariates and state by year covariates for demographics, health care access, and other state contraceptive policies. See the text for sample inclusion rules and full list of covariates. Standard errors are clustered at the individual level and 95% confidence intervals are shown.