Contraception after medication abortion in the United States: results from a cluster randomized trial.

ABSTRACT

BACKGROUND: Understanding how contraceptive choices and access differ for women having medication abortions compared to aspiration procedures can help to identify priorities for improved patient-centered postabortion contraceptive care. OBJECTIVE: The objective of this study was to investigate the differences in contraceptive counseling, method choices, and use between medication and aspiration abortion patients. STUDY DESIGN: This subanalysis examines data from 643 abortion patients from 17 reproductive health centers in a cluster, randomized trial across the United States. We recruited participants aged 18-25 years who did not desire pregnancy and followed them for 1 year. We measured the effect of a full-staff contraceptive training and abortion type on contraceptive counseling, choice, and use with multivariable regression models, using generalized estimating equations for clustering. We used survival analysis with shared frailty to model actual intrauterine device and subdermal implant initiation over 1 year. RESULTS: Overall, 26% of participants (n = 166) had a medication abortion and 74% (n = 477) had an aspiration abortion at the enrollment visit. Women obtaining medication abortions were as likely as those having aspiration abortions to receive counseling on intrauterine devices or the implant (55%) and on a short-acting hormonal method (79%). The proportions of women choosing to use these methods (29% intrauterine device or implant, 58% short-acting hormonal) were also similar by abortion type. The proportions of women who actually used short-acting hormonal methods (71% medication vs 57% aspiration) and condoms or no method (20% vs 22%) within 3 months were not significantly different by abortion type. However, intrauterine device initiation over a year was significantly lower after the medication than the aspiration abortion (11 per 100 person-years vs 20 per 100 person-years, adjusted hazard ratio, 0.50; 95% confidence interval, 0.28-0.89). Implant initiation rates were low and similar by abortion type (5 per 100 person-years vs 4 per 100 person-years, adjusted hazard ratio, 2.41; 95% confidence interval, 0.88-6.59). In contrast to women choosing short-acting methods, relatively few of those choosing a long-acting method at enrollment, 34% of medication abortion patients and 53% of aspiration abortion patients, had one placed within 3 months. Neither differences in health insurance nor pelvic examination preferences by abortion type accounted for lower intrauterine device use among medication abortion patients. CONCLUSION: Despite similar contraceptive choices, fewer patients receiving medication abortion than aspiration abortion initiated intrauterine devices over 1 year of follow-up. Interventions to help patients receiving medication abortion to successfully return for intrauterine device placement are warranted. New protocols for same-day implant placement may also help patients receiving medication abortion and desiring a long-acting method to receive one.