Summary: Intrauterine devices (IUDs) are among the safest and most effective forms of contraception available. They are also the most popular reversible contraceptive in the world. Yet lingering apprehension about their association with pelvic inflammatory disease (PID) limits their acceptance in some areas. This FHI brief summarizes six recent studies and review articles confirming the safety of the IUD in general and especially with respect to PID and infertility.
Key Points
-
Copper IUDs are among the safest, most effective, and least expensive reversible contraceptives available.
-
Nulligravid women can be appropriate candidates for the copper IUD.
-
In a population with STI prevalence of 10 percent, the calculated risk of full clinical PID attributable to IUD insertion is 0.15 percent (less than one in 600 women).
-
IUDs may be an appropriate contraceptive method for HIV-1-infected women with ongoing access to medical services.
-
With their high effectiveness, long duration of use and favorable safety profile, modern IUDs offer advantages over surgical sterilization and Norplant, especially for women who are at low risk of STIs.
-
Contemporary IUD use is extremely safe, with or without the use of prophylactic antibiotics.
Use of copper intrauterine devices and the risk of tubal infertility among nulligravid women. Hubacher D, Lara-Ricalde R, Taylor DJ, et al. N Engl J Med 2001;345:561–67.
Mindful that concern over increased risk of infertility has limited the promotion and acceptance of IUDs in some parts of the world (particularly the United States), the authors undertook a case-control study of the copper-containing TCu380A and subsequent infertility. Between 1997 and 1999, 1,895 women were recruited to supply information about their sexual histories, genitourinary health, and past use of contraceptives, including IUDs.
The study population included 358 women with primary infertility and radiographically confirmed tubal occlusion; 953 controls with primary infertility but no tubal occlusion; and 584 primagravid controls.
Stratified analyses and logistic regression revealed no association between prior TCu380A use and tubal occlusion, either when using infertile controls or primigravid controls.
The authors also found tubal infertility unassociated with duration of IUD use, reason for IUD removal, or the presence or absence of gynecologic problems related to IUD use. However, they did find an association between infertility and the presence of chlamydia antibodies.
Risk of clinical pelvic inflammatory disease attributable to an intrauterine device. Shelton JD. Lancet 2001;357:443.
Using disease-prevalence and STI risk-assessment data from other studies, the author calculated the risk of clinical PID attributable to IUDs, presuming that PID results from IUD insertion in the presence of cervical gonorrhea (GC) or chlamydia (CT) and occurs in the first few months post-insertion. Absolute risk of PID was calculated as the product of the prevalence of GC/CT multiplied by the probability of contracting PID when an IUD is inserted in the presence of GC/CT. The values used by the author in his calculations of PID risk are summarized in the following table:
|
Risk |
Percent |
|
Prevalence of GC/CT in general population |
10.00 |
|
Reduction in GC/CT prevalence by screening |
50.00 |
|
Risk of PID with IUD if GC/CT are present |
5.00 |
|
Absolute risk of PID in IUD user |
0.25 |
|
Relative risk of PID in non-IUD user |
40.00 |
|
Absolute risk of PID in non-IUD user |
0.10 |
|
Risk of PID attributable to IUD |
0.15 |
Assuming that GC/CT prevalence among IUD recipients can be reduced by 50 percent with simple screening and questions about STI risk, in a population with GC/CT prevalence of 10 percent, the calculated risk of full clinical PID attributable to IUD insertion is 0.15 percent (less than one in 600 women); with GC/CT prevalence at 20 percent, the PID risk rises to 0.3 percent; and with a prevalence of five percent, the risk falls to 0.075 percent.
Is the intrauterine device appropriate contraception for HIV-1-infected women? Morrison CS, Sekadde-Kigondu C, Sinei SK, Weiner DH, Kwok C, Kokonya D. Br J Obstet Gynaecol 2001;198(8):784-790.
This study examined whether HIV-1-infected women are more likely to experience complications following IUD insertion than non-infected women. The authors collected data on PID, infection, pain, bleeding, expulsions, and pregnancies from a sample of 156 HIV-1 infected and 493 non-infected women in Nairobi, Kenya. Patients received 1-, 4-, and 24-month follow-up visits. The authors identified complications in 14.7% of infected women and 14.8% of non-infected women. Subsequent multivariate analyses suggested there was no relationship between HIV-infection status and risk of overall complications or infection-related complications. However, inadequate statistical power, due to low PID rates in both infected (2.0%) and non-infected groups (.4%), prevented the authors from ruling out a slight increase in risk of PID among HIV-1-infected women.
Intrauterine devices: the optimal long-term contraceptive method? Fortney JA, Feldblum P, Raymond E. J Reprod Med 1999;44:269–74.
Fortney et al. reviewed selected data on the efficacy, safety, cost, and "technical ease" of the IUD, levonorgestrel implant (Norplant), and surgical sterilization; they reported the following:
Efficacy: The contraceptive efficacy of the IUD is comparable to that of sterilization and may be superior to that of Norplant. The approved effective service life of the copper T 380A (TCu380A) IUD is 10 years (additional research has shown effectiveness for up to 12 years); the levonorgestrel intrauterine system (LNg IUS), at least 7 years; Norplant, a maximum of five years.
Safety: The chief safety concern with the IUD is increased risk of PID, which can be substantially reduced with careful patient screening and proper insertion technique. The LNg IUS may not be associated with increased risk of PID. Menstrual disturbances are the most common adverse events associated with the IUD and IUS. The TCu380A is associated with increased bleeding. The LNg IUS, like Norplant, is associated with reduced or absent bleeding.
Cost: The TCu380A is the most cost-effective reversible contraceptive in the United States, and data from Thailand show it to be more cost-effective than implants or injectables in that country. No comparative data were available on the LNg IUS.
Technical ease: The IUD/IUS is easier to insert and remove than Norplant.
Prophylactic antibiotics for intrauterine device insertion: a meta-analysis of the randomized controlled trials. Grimes DA, Schulz KF. Contraception 1999;60:57–63.
The authors reviewed and analyzed data from all the conventionally indexed (MEDLINE, EMBASE) randomized controlled trials of antibiotic prophylaxis to prevent early infection following IUD insertion. They found four trials conducted from 1966 to 1998. All the trials compared either oral doxycycline or azithromycin versus placebo or no treatment.
They found no significant protective effect against pelvic inflammatory disease (PID) or premature IUD discontinuation. The only significant effect noted was a decrease in the frequency of unscheduled return visits, possibly due to decreased rates of subclinical endometritis, which can cause pain or bleeding that might prompt a woman to return to the clinic.
This work was supported by the U.S. Agency for International Development (USAID). The contents do not necessarily reflect USAID views and policy.
For more information, contact FHI's Research to Practice initiative at rtop@fhi.org.
This document is also available as a PDF in English (380 KB), a PDF (115 KB) in Spanish, and a PDF in French (370 KB).
