As noted by Canavan in this issue of American Family Physician,1 the intrauterine device (IUD) remains a seldom-used method of contraception. Surveys of U.S. women have shown that IUDs have a poor reputation. However, the same surveys showed that women using IUDs have the highest satisfaction with this method of contraception. Among physicians, fear of litigation is the most common reason for not offering IUDs to their patients.2 However, in reality, neither manufacturers nor physicians are being sued over contemporary devices.3 These paradoxes—among both women and clinicians—have hurt women's health in recent decades.
Upper genital tract infection and subsequent infertility are the principal concerns about IUD safety. These fears are largely unfounded. Pelvic inflammatory disease occurs much less frequently with IUD use than most physicians and patients think; moreover, the occurrence of pelvic inflammatory disease is associated with the insertion of the IUD and not with the device itself. Both epidemiologic and bacteriologic evidence points to contamination of the endometrial cavity at the time of insertion as the mechanism for IUD-related infection. Despite this transient contamination, infection rarely occurs. Because prophylactic antibiotics significantly reduce the risk of infection after induced abortion, several trials have examined the potential usefulness of antibiotics for IUD insertion. In a recent trial in Los Angeles County,4 azithromycin (Zithromax), given in a dose of 500 mg orally one hour before insertion, had no significant benefit when compared with placebo. The most important finding of this trial was the near absence of upper genital tract infection in women during the first three months after IUD insertion.
Canavan raises but does not settle the question of the IUD string and its relation to pelvic inflammatory disease. First, the risk of IUD-related pelvic inflammatory disease is inversely associated with the duration of use. If the string drew bacteria into the uterus, then the risk of infection would increase with time. It does not. Second, cohort studies from China comparing the use of IUDs with and without tail strings have found no important differences in the low rates of pelvic inflammatory disease. Third, a randomized controlled trial5 comparing the same device with and without a tail string reached the same conclusion. Thus, the hypothesis that the string poses a risk is untenable, and the issue of “biofilm” becomes clinically irrelevant.
The risk of tubal infertility related to IUD use is low. Although findings of some case-control studies have raised this concern, findings of cohort studies, which are less subject to bias than are case-control studies, have provided reassurance. Women in New Zealand and Norway who discontinued using their IUDs because of problems did not have a significant increase in infertility compared with women who stopped using an IUD to become pregnant.6,7
Women infected with human immunodeficiency virus (HIV) need highly effective contraception. Based on theoretic concerns about increased susceptibility to upper genital tract infection, the World Health Organization and International Planned Parenthood have recommended that women with HIV infection not use an IUD.8,9 However, a recent cohort study10 in Nairobi, Kenya, showed that HIV-infected women, compared with uninfected women, do not have a significant increase in the risk of pelvic inflammatory disease in the early months after insertion of an IUD. Long-term follow-up of these women is continuing.
Some distinctions between IUDs are important. Table 6 in Canavan's article notes heavier bleeding in association with IUDs. While this is true for the copper-bearing device (Paragard T 380A), the opposite is true with the T-shaped progesterone system (Progestasert): it reduces average blood loss. Indeed, progestin-releasing devices can treat heavy menses or dysmenorrhea. Table 6 in Canavan's article also indicates a 26-fold increase in the risk of septic abortion in women who use IUDs. This figure for septic second-trimester fetal loss comes from one case-control study conducted 20 years ago and relates to IUDs that are no longer marketed in the United States. However, in that study, if the IUD were removed in the first trimester, the relative risk of septic second-trimester fetal loss was 1.2 (95 percent CI, 0.2 to 6.9). Other investigators have not found an increased risk of febrile spontaneous abortion among IUD users.11
Women and their clinicians need to take a fresh look at IUDs. Contemporary IUDs are very safe and are also highly effective.The cumulative 10-year probability of an accidental pregnancy with the copper-bearing T 380A is similar to that with tubal sterilization. When compared with tubal sterilization, an IUD is less expensive, safer, more convenient and immediately reversible. Indeed, the copper T 380A can be considered a “reversible method of sterilization.”
As noted a decade ago by a World Health Organization scientific group,12 “the currently available copper and hormone-releasing IUDs, when properly used, are probably the most effective and reliable reversible method of fertility regulation.” This upbeat assessment still holds true today. Nevertheless, use of the IUD has languished in the United States. As implied in Canavan's review, family physicians have a key role in the renaissance of the IUD.