Effectiveness of IUDs: a review.

Large, multicenter, randomized trials of IUDs conducted between 1970 and 1986 encompass more than 50,000 woman-years of experience in the two-year interval following device insertion. Taken singly and collectively, these trials demonstrate that IUD failure rates are strongly affected by the age of participants, notably by the proportion of women under age 25 admitted to the studies. Individual trials and overall comparisons indicate further that: 1. Non-medicated devices such as the Mahua steel ring and Lippes loop D have pregnancy rates above 2 per 100 woman-years. 2. The first approved copper IUDs, the Copper 7 and the TCu 200, do not markedly differ in pregnancy rates from standard plastic devices with rates significantly above 2 per 100 woman-years. 3. Five more recent, widely used copper IUDs have had failure rates significantly below 2 per 100 woman-years in multicenter trials. The point estimate of the failure rate was less than 1 pregnancy per 100 woman-years for three of these IUDs. One copper-bearing device had a failure rate significantly below 1 per 100 woman-years. 4. Devices releasing either 25 mcg/day of progesterone or 2 mcg/day of levonorgestrel have had significantly higher failure rates than the more effective copper-releasing IUDs; however, devices releasing 20 mcg/day of levonorgestrel appear at least as effective as the most effective copper IUDs. Long-term IUD failure rates derived from straight assignment as well as randomized trials indicate 4-year gross cumulative failure rates do not exceed 10 per 100 for any of the standard copper devices or the Lippes loop D. Point estimates of the four-year gross cumulative pregnancy rates have been below 5 per 100 users in studies of the TCu 220C, the Nova T, the TCu 380A, and the MLCu 375.