Provision of contraception immediately after an induced abortion is safe, acceptable to patients, and effective at reducing unintended pregnancy.1–3 Intrauterine devices (IUDs), subdermal implants, and depot medroxyprogesterone acetate (DMPA) can be received before a woman leaves the procedure room after a uterine aspiration or dilation and evacuation.1,2 These same methods can be initiated by the time medical abortion is complete.4 Repeat unintended pregnancy rates and repeat abortion rates are both reduced by half when women receive contraception at the time of an abortion (immediate postabortion contraception) compared with waiting for a return visit.4–12
Despite the benefits of postabortion contraception, some states prohibit abortion providers from receiving state Title X family planning funds and have refused to expand Medicaid. Furthermore, many clinics have difficulty receiving adequate reimbursement from insurance carriers.13–16 In an abortion setting, 75% of women are poor or low income and are often incurring out-of-pocket costs for abortion services.17,18 Financial barriers to immediate postabortion contraception are particularly detrimental for women who desire the more expensive methods: DMPA and long-acting reversible contraception (LARC), which include IUDs and the implant.13,15,16 Women who report having insurance start immediate postabortion LARC at twice the rate of those without.17
Although insurance coverage for contraception is related to uptake of LARC immediately postabortion, several factors related to insurance coverage and immediate postabortion contraception are unknown.13,17 Documented rather than self-reported insurance coverage has not been studied nor has this association been examined after implementation of the Affordable Care Act.
We hypothesized that women who had documented insurance coverage for contraception on the day of their abortion would be more likely to receive contraception compared with women who did not have documented coverage. Our objective was to measure the relationship between having insurance coverage for contraception and receiving IUDs, an implant, or DMPA before discharge after abortion care at one health center in North Carolina.
MATERIALS AND METHODS
This was a prospective cohort study of women obtaining an abortion at a single freestanding health center in North Carolina. The University of North Carolina institutional review board approved the study. The health center offered medical abortions and first- and second-trimester surgical abortions 3 days each week. Women traveled from South Carolina, Virginia, West Virginia, and all regions of North Carolina to receive care. Women with public, private, or no insurance were seen in the health center.
All contraception counseling and provision continued as standard clinical practice throughout the course of the study. Women received contraceptive counseling and condoms from health center clinical staff before their abortion. Methods available on-site were DMPA, levonorgestrel-releasing intrauterine systems, copper IUDs, subdermal implants, oral contraception, and vaginal rings. Women who wanted pills or rings from the health center had to pay out of pocket because the health center did not accept insurance payment for those two methods. Women undergoing medical abortion could receive levonorgestrel-releasing intrauterine systems, copper IUD, implant, or DMPA at their follow-up visit only.
There were a number of days when LARC, including implants and both types of IUDs, could be provided for free to women who qualified for a grant program. These days occurred when trainees worked with the abortion providers. A woman qualified for the grant program's free LARC if she did not have insurance coverage for contraception and had a household income at or below 300% of the 2015 Federal Poverty Level. There was never free LARC available through the grant program on medical abortion follow-up days as a result of physician scheduling. Free LARC was available to qualifying women on 18 of the 85 study days (22%).
Women were recruited for the study when they registered at the clinic on the day of their visit to initiate an abortion. Services obtained by women on the day of enrollment were uterine aspiration, laminaria placement, or mifepristone administration (Fig. 1). Women were eligible if they were interested in participating in the study, at least 18 years old, able to consent in English or Spanish, and had consented through the standard clinic protocol for an abortion.
Data were collected immediately after consent to the study. Characteristics of women in the cohort were collected from study-related, research staff-administered questionnaires and review of participants' electronic medical records (Fig. 1). Questionnaire data, including demographics and reproductive health history, were collected in private before the abortion. Insurance status, clinical data from the visit, and method of postabortion contraception received were collected from the electronic medical record. Just before discharge from the clinic, women were asked whether they would have chosen a different method had it been free and, if so, what method. Research staff entered data into the REDCap database and validated 10% of all data entry and 100% of exposure and primary outcome data entry as a group.
Women who underwent surgical abortion received LARC in the procedure room or DMPA in the recovery room immediately after their abortion and exited the study at the time of discharge from the clinic after the abortion. Women who underwent medical abortion exited after the follow-up visit when completed abortion was confirmed and contraception was received. We considered women who did not return to the health center by a month after receiving mifepristone lost to follow-up.
Our two comparison groups were defined by status of insurance coverage for contraception as documented in the electronic medical record according to standard clinic procedures. Call center staff asked women for insurance information at the time an appointment was booked. They then called the insurance company and documented current abortion and contraception coverage benefits in the electronic medical record. The women who had current insurance coverage for contraception were in the “coverage group.” Women without current insurance, insurance that did not cover contraception, or unknown contraceptive coverage were in the “no coverage group.”
Our primary outcome was defined by whether women received immediate postabortion contraception. This outcome was defined by timing and method received. We defined immediately postabortion as the same visit the surgical abortion was completed or at the end of a follow-up visit for medical abortion. We initially defined all contraceptive methods available on-site that were at least 90% effective as our primary outcome. We changed our primary outcome to IUDs, implant, and DMPA after data collection because it was at that point we learned those were the only available methods on-site that could be paid for with insurance. These primary outcome data were obtained from the electronic medical record.
The sample size was determined a priori based on our expected difference in receipt of immediate postabortion contraception between the coverage and no coverage groups. We expected the proportion of the coverage group that would receive immediate postabortion contraception to be 60% compared with 45% in the no coverage group. With a two-sided α of 0.05 and a power of 90%, we needed at least 600 participants to detect a 15% difference in receipt of immediate postabortion contraception between the two groups.
The analysis to address our main study hypothesis focused on insurance status and receipt of IUDs, an implant, or DMPA. We used Pearson χ2 testing to compare insurance status with each participant characteristic to look for any differential distribution by insurance status. We compared unadjusted participant characteristics by receipt of contraception to identify any risk factors that might confound the relationship of insurance status and receipt of contraception in the subsequent multivariable analysis. Our initial logistic regression model included insurance coverage status and all potential confounding variables. Our final reduced model adjusted for educational level and gestational age because these were the only two characteristics that substantially changed the effect size. Using β estimates from our final reduced model, we estimated the adjusted percentage of women who left the health center with contraception by insurance coverage group. We used a log binomial regression model to estimate adjusted risk ratios and 95% confidence intervals (CIs) because this was a cohort study with a common outcome.
We performed a number of subgroup analyses. Women undergoing medical abortion have different access to immediate postabortion contraception compared with women obtaining surgical abortion, so we analyzed women who underwent only surgical abortion, only medical abortion, and women who either underwent surgical abortion or had a medical abortion and returned for follow-up. We also analyzed women who received their surgical abortion on one of the days when free grant LARCs were available compared with the rest of the women. We used Stata 13 for statistical analysis.
Women enrolled into the cohort from September 24, 2015, through April 22, 2016. During this time, 1,402 women sought abortion care at the health center (Fig. 2). The demographic makeup of the 600 women enrolled in the study was similar to the 1,402 women seen in the health center over the same time period in regard to age, race and ethnicity, language, gestational age, abortion type, and insurance coverage for abortion.
Five hundred seventy-five women were included in our final analysis (Table 1). One hundred twenty-eight women (22% [128/575]) had contraceptive coverage and 447 (78% [447/575]) did not have coverage. Age, gravidity, and history of prior abortion were similar between women who did and did not have contraceptive coverage. A similar percentage of women had surgical and medical abortions in both groups. Less than one third of women in both groups were seen on days when free LARC was available. More women in the coverage group reported desiring an IUD, implant, or DMPA if it were free (59% [62/105] compared with 47% [168/357]). The most common reasons women with insurance coverage did not receive desired contraception were going to their regular doctor or a pharmacy for contraception, concerns about side effects, having pills at home, and patient confusion about insurance coverage (data not shown).
Table 2 is a descriptive table of unadjusted characteristics by receipt of immediate postabortion IUD, implant, or DMPA. More women (38% [49/128]) received an IUD, implant, or DMPA in the contraceptive coverage group than in the no coverage group (7% [33/447], P<.01). Of the 33 women in the no coverage group who received contraception, 70% (23/33) received a free grant program LARC (data not shown). Demographic and reproductive history was similar between the two groups. As expected, abortion type, desire for contraception, and availability of free LARC were statistically significant, but they were not confounders because they either only affected the receipt of contraception and not insurance coverage or were on the causal pathway to receiving contraception. They were therefore not included in the multivariable modeling.
Women with contraceptive insurance coverage were 5.6 times as likely (adjusted risk ratio 5.6, 95% CI 3.8–8.3) to leave the health center with contraception as women without contraceptive coverage (Table 3) after adjusting for educational level and gestational age. In the subgroup of women who presented for abortion on days when no free LARC was available, the adjusted risk ratio was a striking 20.8 (95% CI 9.0–48.1) for receiving contraception if a woman had insurance compared with no coverage. The final multivariable analysis was adjusted for educational level and gestational age because they were the only characteristics that changed the effect size in our model.
Women with insurance coverage for contraception on the day of their abortion in our cohort in North Carolina were 5–20 times as likely to receive immediate IUD, an implant, or DMPA than women with no coverage. Our study of women from one health center in North Carolina demonstrates that insurance coverage is strongly associated with contraceptive uptake immediately after abortion. Public insurance coverage for abortion was associated with twice the uptake of postabortion LARC compared with lack of public insurance for abortion in a number of settings across the United States.17 Women who live in states without Medicaid expansion were less likely to have insurance.19,20 North Carolina neither provides Medicaid coverage of abortion nor expanded Medicaid.19 We infer from our findings that women who have abortions in states with similar restrictions likely do not receive their desired contraceptive method immediately after an abortion. To achieve a decrease in unintended pregnancies, we believe our findings support equitable access to insurance in all settings where women receive contraceptive services, including abortion clinics.
Insurance coverage and decreased cost of all contraceptives are associated with greater contraception uptake. Among privately insured women in a nonabortion setting across the United States, lower copayments and out-of-pocket costs resulted in increased IUD and DMPA use.21–23 When free contraception was provided to women in St Louis, 75% of women in a cohort received LARC compared with 9% LARC use nationwide.24,25 Our findings also demonstrated that when LARC is available for free in an abortion setting, uptake of LARC was 10-fold higher than days when no free LARC was available.
Self-reported contraceptive desire, and actual uptake of contraception, differed in our cohort. We propose several reasons women did not leave the clinic with their self-reported desire for method of contraception. First, women with coverage cited confusion about what insurance covered. Second, some women with contraceptive coverage likely did not want to have their insurance carrier contacted by an abortion clinic. When women across the United States did not use their self-reported private insurance for their abortion, they were just as likely to leave the clinic without LARC as uninsured women.17 Third, some women may overreport wanting contraception as a result of social desirability bias, perhaps believing research staff view wanting contraception favorably. Women in four California clinics did not always want to discuss or start contraception at an abortion visit and the same could have been true among our study participants.26
We conducted our study at a single health center and this may limit generalizability. The particular health center where we conducted our study had a wide range of contraceptive methods in stock as well as access to the free LARC devices because of the relationship between the health center and physician trainees. Many abortion clinics cannot stock contraception as a result of financial, administrative, or policy barriers. These barriers include seeking preauthorization for LARC, clinic flow challenges, lack of clinician training, and the administrative burden of operationalizing Title X statutory and regulatory requirements.13,14,16 Women seen in our health center during the course of the study did not receive implants or DMPA on the day of their mifepristone visit. This proactive approach may be taken at other clinics.27,28
Women should have affordable and equitable access to all methods of contraception after an abortion. Not all women want the most effective or even any contraception after an abortion, and they should not be coerced into accepting undesired contraception.29 However, women who do want contraception should be able to access it, regardless of which state they live in. Making effective contraceptive easy for women to obtain and clinics to administer immediately after an abortion will decrease unintended pregnancy and repeat abortions.4–12 Making it straightforward for women to use insurance at abortion clinics and increasing contraceptive insurance coverage with Medicaid expansion are two initiatives that can accomplish this goal.14,16,17,20
1. Curtis KM, Tepper NK, Jatlaoui TC, Berry-Bibee E, Horton LG, Zapata LB, et al. U.S. medical eligibility criteria for contraceptive use, 2016. MMWR Recomm Rep 2016;65:1–103.
2. Okusanya BO, Oduwole O, Effa EE. Immediate postabortal insertion of intrauterine devices. The Cochrane Database of Systematic Reviews 2014, Issue 7. Art. No.: CD001777.DOI: 10.1002/14651858.CD001777.pub4.
3. Kavanaugh ML, Carlin EE, Jones RK. Patients' attitudes and experiences related to receiving contraception during abortion care. Contraception 2011;84:585–93.
4. Shimoni N, Davis A, Ramos ME, Rosario L, Westhoff C. Timing of copper intrauterine device insertion after medical abortion: a randomized controlled trial. Obstet Gynecol 2011;118:623–8.
5. Goodman S, Hendlish SK, Reeves MF, Foster-Rosales A. Impact of immediate postabortal insertion of intrauterine contraception on repeat abortion. Contraception 2008;78:143–8.
6. Heikinheimo O, Gissler M, Suhonen S. Age, parity, history of abortion and contraceptive choices affect the risk of repeat abortion. Contraception 2008;78:149–54.
7. Cremer M, Bullard KA, Mosley RM, Weiselberg C, Molaei M, Lerner V, et al. Immediate vs. delayed post-abortal copper T 380A IUD insertion in cases over 12 weeks of gestation. Contraception 2011;83:522–7.
8. Cameron ST, Glasier A, Chen ZE, Johnstone A, Dunlop C, Heller R. Effect of contraception provided at termination of pregnancy and incidence of subsequent termination of pregnancy. BJOG 2012;119:1074–80.
9. Roberts H, Silva M, Xu S. Post abortion contraception and its effect on repeat abortions in Auckland, New Zealand. Contraception 2010;82:260–5.
10. Rose SB, Lawton BA. Impact of long-acting reversible contraception on return for repeat abortion. Am J Obstet Gynecol 2012;206:37.e1–6.
11. Langston AM, Joslin-Roher SL, Westhoff CL. Immediate postabortion access to IUDs, implants and DMPA reduces repeat pregnancy within 1 year in a New York City practice. Contraception 2014;89:103–8.
12. Pohjoranta E, Mentula M, Gissler M, Suhonen S, Heikinheimo O. Provision of intrauterine contraception in association with first trimester induced abortion reduces the need of repeat abortion: first-year results of a randomized controlled trial. Hum Reprod 2015;30:2539–46.
13. Kavanaugh ML, Jones RK, Finer LB. Perceived and insurance-related barriers to the provision of contraceptive services in U.S. abortion care settings. Womens Health Issues 2011;21(suppl):S26–31.
14. Thompson KM, Speidel JJ, Saporta V, Waxman NJ, Harper CC. Contraceptive policies affect post-abortion provision of long-acting reversible contraception. Contraception 2011;83:41–7.
15. Morse J, Freedman L, Speidel JJ, Thompson KM, Stratton L, Harper CC. Postabortion contraception: qualitative interviews on counseling and provision of long-acting reversible contraceptive methods. Perspect Sex Reprod Health 2012;44:100–6.
16. Sonfield A, Gold RB. Beyond consent to sterilization: facing up to the full range of barriers to post-abortion contraception. Am J Obstet Gynecol 2014;211:3–4.
17. Rocca CH, Thompson KM, Goodman S, Westhoff CL, Harper CC. Funding policies and postabortion long-acting reversible contraception: results from a cluster randomized trial. Am J Obstet Gynecol 2016;214:716.e1–8.
20. Jones RK, Sonfield A. Health insurance coverage among women of reproductive age before and after implementation of the affordable care act. Contraception 2016;93:386–91.
21. Pace LE, Dusetzina SB, Fendrick AM, Keating NL, Dalton VK. The impact of out-of-pocket costs on the use of intrauterine contraception among women with employer-sponsored insurance. Med Care 2013;51:959–63.
22. Gariepy AM, Simon EJ, Patel DA, Creinin MD, Schwarz EB. The impact of out-of-pocket expense on IUD utilization among women with private insurance. Contraception 2011;84:e39–42.
23. Postlethwaite D, Trussell J, Zoolakis A, Shabear R, Petitti D. A comparison of contraceptive procurement pre- and post-benefit change. Contraception 2007;76:360–5.
24. Peipert JF, Madden T, Allsworth JE, Secura GM. Preventing unintended pregnancies by providing no-cost contraception. Obstet Gynecol 2012;120:1291–7.
25. Kavanaugh ML, Jerman J, Finer LB. Changes in use of long-acting reversible contraceptive methods among U.S. women, 2009–2012. Obstet Gynecol 2015;126:917–27.
26. Matulich M, Cansino C, Culwell KR, Creinin MD. Understanding women's desires for contraceptive counseling at the time of first-trimester surgical abortion. Contraception 2014;89:36–41.
27. Raymond EG, Weaver MA, Tan YL, Louie KS, Bousiéguez M, Lugo-Hernández EM, et al. Effect of immediate compared with delayed insertion of etonogestrel implants on medical abortion efficacy and repeat pregnancy: a randomized controlled trial. Obstet Gynecol 2016;127:306–12.
28. Raymond EG, Weaver MA, Louie KS, Tan YL, Bousiéguez M, Aranguré-Peraza AG, et al. Effects of depot medroxyprogesterone acetate injection timing on medical abortion efficacy and repeat pregnancy: a randomized controlled trial. Obstet Gynecol 2016;128:739–45.
29. Gomez AM, Fuentes L, Allina A. Women or LARC first? Reproductive autonomy and the promotion of long-acting reversible contraceptive methods. Perspect Sex Reprod Health 2014;46:171–5.