Preventing unintended pregnancy among young women in Kenya: prospective cohort study to offer contraceptive implants.

BACKGROUND Subdermal contraceptive implants have low discontinuation rates but are underused among young women in Africa. This study aimed to isolate the role initial contraceptive method has on preventing unintended pregnancy. STUDY DESIGN We recruited 399 Kenyan women aged 18-24 years into a prospective cohort study if they wanted short-acting hormonal methods (injectable or oral contraceptives). We offered an implant and formed two study groups: implant and short-acting. For contraceptive discontinuation/pregnancy, we used log-rank tests and proportional hazards models. We applied intent-to-treat principles to evaluate the role of initial method choice on future pregnancy. RESULTS Twenty-four percent opted for an implant (n=97), and the remainder opted for a short-acting method (n=299). The 18-month discontinuation probability was 21 per 100 for implant users and 43 per 100 for the short-acting method group (p=.001). Twenty-two unintended pregnancies occurred; all were among the short-acting group. The adjusted relative risk of pregnancy among the short-acting group vs. implant group was 7.4 (95% confidence interval: 1.6-34.5). CONCLUSIONS Many young Kenyan women found implants to be a reasonable alternative to short-acting methods. Having choice is essential, and starting on implants provides substantial and clear protection from unintended pregnancy relative to short-acting methods.