Postplacental intrauterine device expulsion by 12 weeks: a prospective cohort study.

BACKGROUND: An intrauterine device placed immediately following a delivery can serve as an effective and safe contraceptive strategy in the postpartum period. There is limited evidence that the levonorgestrel intrauterine system may have a higher rate of expulsion compared to the copper intrauterine device; however, rates of expulsion for these 2 intrauterine device types have not been compared directly.
OBJECTIVE: We sought to compare expulsion rates by 12 weeks' postpartum for the levonorgestrel intrauterine system and copper intrauterine device. STUDY
DESIGN: We enrolled women who received postplacental intrauterine devices at 2 urban hospitals. Eligible women were ≥18 years old, English- or Spanish-speaking, with singleton vaginal delivery at ≥35 weeks' gestation. Intrauterine devices were inserted within 10 minutes of placental delivery by trained providers using ring forceps or the operator's hand. Intrauterine device location was evaluated via abdominal ultrasound at 24-48 hours' postpartum, and via transvaginal ultrasound 6 and 12 weeks later, categorizing position of the intrauterine device at the fundus, below the fundus but above the internal os, any part of the intrauterine device below the internal os (partial expulsion), or no intrauterine device visualized. Outcomes included intrauterine device expulsion and method continuation. We used multivariable logistic regression to identify factors associated with expulsion.
RESULTS: We enrolled 123 women ages 18-40 years. Of these, 68 (55%) initiated levonorgestrel intrauterine system and 55 (45%) initiated copper intrauterine device. Groups were similar except more copper intrauterine device users were Hispanic (66% vs 38%) and fewer were primiparous (16% vs 31%). Among the 96 (78%) with 12-week follow-up, expulsion was higher for levonorgestrel intrauterine system users (21/55 or 38%) than for copper intrauterine device users (8/41 or 20%) (odds ratio, 2.55; 95% confidence interval, 0.99-6.55; P = .05). At 24 hours' postpartum, there was no significant difference in median distance from the intrauterine device to the fundus between intrauterine device types or between those who did or did not experience expulsion. Of expulsions, 86% occurred ≤6 weeks' postpartum. All complete expulsions were clinically identified, but of the partial expulsions, only 4/10 (40%) were clinically suspected prior to ultrasound. The only independent predictor of expulsion was intrauterine device type. Including reinsertions, intrauterine device use at 12 weeks was not significantly different for levonorgestrel intrauterine system and copper intrauterine device users (80% vs 93%; P = .14).
CONCLUSION: Women initiating postplacental levonorgestrel intrauterine system are more likely to experience complete expulsion than those initiating copper intrauterine device. Using sonographic criteria results in higher expulsion rates than previously reported. It is unclear if such high expulsion rates would be identified following standard clinical practice.