Levonorgestrel capsule implants in the United States: a 5-year study

Abstract

Objective: To measure effectiveness, adverse event experience, and acceptability of the Food and Drug Administration-approved variant of levonorgestrel capsule implants in the United States through 5 years and to examine determinants of these outcomes.

Methods: In a prospective, multicenter study, 511 sexually active women selecting contraceptive implants were monitored four times in the 1st year, then semiannually through 5 years. Adverse events were elicited by query and physical examination, and their incidence was measured. Lifetable analyses computed pregnancy and other discontinuation rates. Cox regression models examined effects of age, parity, and preadmission desire for more children on continuation. Removal times were analyzed by analysis of variance.

Results: Three pregnancies occurred, yielding a 5-year cumulative rate of 1.3+/-0.8 per 100 users, an average annual rate of three per 1000 women, and an ectopic pregnancy rate of 0.6 per 1000 woman years. No pregnancies occurred to women weighing less than 79 kg. Prolonged or irregular menstrual bleeding, followed distantly by headache, weight gain, and mood changes, was the most frequent medical conditions leading to removal. Weight gain averaged 1 kg per year. Each annual continuation rate was above 80 per 100, for a cumulative 5-year rate of 39 per 100. Continuation was age-dependent, with younger women (younger than 25 years at entry) having lower 5-year continuation rates than older subjects (P < .01). Tissue trauma from deeply placed or poorly aligned implants or severe reactions to local anesthetic affected subjects in 3.1% of removals (nine cases).

Conclusion: As measured by annual continuation rates of 80 per 100 or higher and annual pregnancy rates below one per 100, implant contraception in the United States was found to be highly acceptable and effective, year after year, regardless of the woman's age or family formation status. The cumulative 5-year pregnancy rate, 1.3 per 100, is comparable to that of tubal ligation.