Continuation of copper and levonorgestrel intrauterine devices: a retrospective cohort study

Abstract

Studies conflict on whether the duration of use of the copper intrauterine device is longer than that of the levonorgestrel intrauterine device, and whether women who continue using intrauterine devices differ from those who discontinue. We sought to assess continuation rates and performance of levonorgestrel intrauterine devices compared with copper intrauterine devices over a 5-year period. We performed a retrospective cohort study of 1164 individuals who underwent intrauterine device placement at an urban academic medical center. The analysis focused on a comparison of continuation rates between those using levonorgestrel intrauterine device and copper intrauterine device, factors associated with discontinuation, and intrauterine device performance. We assessed the differences in continuation at discrete time points, pregnancy, and expulsion rates using χ2 tests and calculated hazard ratios using a multivariable Cox model. Of 1164 women who underwent contraceptive intrauterine device insertion, 956 had follow-up data available. At 2 years, 64.9% of levonorgestrel intrauterine device users continued their device, compared with 57.7% of copper intrauterine device users (P= .11). At 4 years, continuation rates were 45.1% for levonorgestrel intrauterine device and 32.6% for copper intrauterine device (P < .01), and at 5 years continuation rates were 28.1% for levonorgestrel intrauterine device and 23.8% for copper intrauterine device (P= .33). Black race, primiparity, and age were positively associated with discontinuation; education was not. The hazard ratio for discontinuation of levonorgestrel intrauterine device compared with copper intrauterine device >4 years was 0.71 (95% confidence interval, 0.55-0.93) and >5 years was 0.82 (95% confidence interval, 0.64-1.05) after adjusting for race, age, parity, and education. Copper intrauterine device users were more likely to experience expulsion (10.2% copper intrauterine device vs 4.9% levonorgestrel intrauterine device, P < .01) over the study period and to become pregnant in the first year of use (1.6% copper intrauterine device vs 0.1% levonorgestrel intrauterine device, P < .01). We found a difference in continuation rates between levonorgestrel and copper intrauterine device users at 4 years but not at 5 years. Copper intrauterine device users were more likely to experience expulsion and pregnancy.