Immediate postabortion initiation of levonorgestrel implants reduces the incidence of births and abortions at 2 years and beyond

Abstract

ObjectivesThe aims of this study were to compare immediate postabortion uptake of recently subsidized (no-cost) levonorgestrel-releasing implants with already available intrauterine and shorter-acting methods and to compare the incidence of subsequent pregnancies (ending in birth or abortion) within 2 years. Study designRetrospective chart review of 4698 women attending a New Zealand public hospital abortion clinic over 2 years (2010–2012) to describe postabortion contraceptive choice, with follow-up via clinic and national births records to assess subsequent pregnancies at 12, 24, 36, and 48 months. ResultsTwenty percent of the cohort (934/4698) received an implant, 26% an intrauterine method (927 copper intrauterine device, 301 levonorgestrel-releasing intrauterine system [LNG-IUS]), and 54% chose other shorter-acting methods (2536/4698). Method choice was significantly associated with age, ethnicity, and pregnancy history (p < .001). At 24 months, the unadjusted incidence of subsequent abortion for implant users was 3.8% (95% confidence interval [CI] = 2.5–5.0) and 11.6% (95% CI = 10.3–12.8) for those choosing other short-acting methods. By 48 months, 6.6% of implant users had a subsequent abortion (compared with 18.3% for short-acting methods). The incidence of continued pregnancy at 24 months was 6.3% (95% CI = 4.4–8.1) for implant users and 15.7% (95% CI = 14–17.2) for those choosing other short-acting methods. Adjusted hazard ratios (HRs) for subsequent abortion were lowest for women initiating an implant (HR = 0.26, 95% CI = 0.20–0.35) or LNG-IUS (HR = 0.26, 0.16–0.44, reference group: short-acting methods). ConclusionsImmediate postabortion insertion of an implant significantly reduced rates of subsequent pregnancy for at least 2 years. Abortion service providers should ensure women have barrier-free access to all long-acting reversible contraceptions to delay or prevent pregnancy. ImplicationsInitiation of an levonorgestrel implant immediately postabortion was associated with a 74% reduction in subsequent abortion over the next 4 years compared with use of short-acting methods. Implants were popular among adolescents—a group at high-risk of subsequent pregnancy, and who have not historically been considered appropriate candidates for intrauterine contraceptive methods.