Current issues of ACP Journal Club are published in Annals of Internal Medicine


Review: The levonorgestrel-releasing intrauterine system prevents more pregnancies than do intrauterine devices (IUDs) ≤ 250 mm2 but not > 250 mm2


ACP J Club. 2001 Sep-Oct;135:64. doi:10.7326/ACPJC-2001-135-2-064

Source Citation

French RS, Cowan FM, Mansour D, et al. Levonorgestrel-releasing (20 µg/day) intrauterine systems (Mirena) compared with other methods of reversible contraceptives. BJOG. 2000 Oct;107:1218-25. [PubMed ID: 11028571] (All 2001 articles were reviewed for relevancy, and abstracts were last revised in 2007.)



In women of reproductive age, is the levonorgestrel-releasing (20 µg/d) intrauterine system (LNG-20) more effective and tolerable than other reversible contraceptive methods?

Data sources

Studies were identified using MEDLINE, EMBASE/Excerpta Medica, PsycLIT, POPLINE, Cochrane Controlled Trials Register, and specialist databases between January 1974 and July 1998. Bibliographies of relevant articles were also reviewed. Experts in the field and organizations were contacted for unpublished data.

Study selection

Randomized controlled trials were selected if participants were women of reproductive age, they compared LNG-20 with another reversible contraceptive, and they had predetermined outcomes.

Data extraction

Data were extracted on predetermined outcomes of pregnancy as a result of method or user failure and continuation of contraceptive method. Data on these outcomes were extracted when either number of events per women months or single-decrement life-table probabilities were reported.

Main results

5 studies met the inclusion criteria. LNG-20 was compared with 2 groups of intrauterine devices (IUDs), categorized by the surface area of the copper wire: ≤ 250 mm2 (Nova-T, CuT 200, and CuT 220 IUDs) and > 250 mm2 (CuT 380 Ag IUD). At 5 years of follow-up, LNG-20 was more effective than IUDs ≤ 250 mm2 for preventing unplanned pregnancy (rate ratio [RR] 0.08, 95% CI 0.04 to 0.18) but not more effective than IUDs > 250 mm2 (RR 0.66, CI 0.25 to 1.75) (Table). Overall, continued use was more likely with LNG-20 than with IUDs ≤ 250 mm2 (RR 1.16, CI 1.02 to 1.31) and equally as likely with IUDs > 250 mm2 (RR 0.97, CI 0.83 to 1.14) (Table). LNG-20 users were more likely to discontinue use because of hormonal side effects than were users of IUDs ≤ 250 mm2(RR 5.2, CI 1.3 to 20.3) and IUDs > 250 mm2 (RR 4.2, CI 2.0 to 9.0). LNG-20 users were also more likely to discontinue use because of amenorrhea than were users of IUDs ≤ 250 mm2(RR 29.2, CI 1.7 to 488) and IUDs > 250 mm2 (RR 48.9, CI 16.9 to 141).


In women of reproductive age, the levonorgestrel-releasing (20 µg/d) intrauterine system (LNG-20) is more effective than intrauterine devices (IUDs) ≤ 250 mm2 but as effective as IUDs > 250 mm2 for preventing pregnancy. LNG-20 is more likely to lead to discontinued use because of hormonal side effects and amenorrhea.

Source of funding: NHS R&D Health Technology Assessment Programme.

For correspondence: Dr. R. French, Department of Sexually Transmitted Diseases, Royal Free and University College Medical School, Mortimer Market Centre, off Capper Street, London WC1E 6AU, England, UK. FAX 44-207-679-9669.

Table. Levonorgestrel-releasing (20 µg/d) intrauterine system (LNG) vs intrauterine devices ≤ 250 mm2 (IUD ≤ 250) or > 250 mm2 (IUD > 250) in women of reproductive age at 5 years of use*

Outcomes Comparison Event rates RRR (95% CI) NNT (CI)
Pregnancy LNG vs IUD ≤ 250† 0.009% vs 0.11% 92% (81 to 97) 17 (13 to 26)
LNG vs IUD > 250 0.017% vs 0.026% 34% (-74 to 75) Not significant
Continued use LNG vs IUD ≤ 250† 1.11% vs 0.96% 16% (2 to 31) 12 (7 to 61)
LNG vs IUD > 250 0.85% vs 0.88% 3% (-14 to 17) Not significant

*RBX = relative benefit reduction. Other abbreviations defined in Glossary; RRR, RBI, RBR, NNT, NNH, and CI calculated from data in article.
†2 studies were meta-analyzed using a fixed-effects model and reported weighted event rates; other comparisons were based on 1 study.


The systematic review by French and colleagues addresses an important clinical question. The study methods reported were appropriate, and the search for relevant studies was comprehensive. The willingness by volunteers to be randomly allocated, however, implies that either contraceptive method was acceptable. The copper IUD comparative studies would therefore be unlikely to include women with heavy or painful menses, who are most likely to benefit from the “side effects” of LNG-20 (i.e., reduction of menstrual flow and pain). Some studies also noted the importance of preinsertion counseling to reduce concern about amenorrhea or the occurrence of irregular bleeding in the early months of use.

The category of IUDs ≤ 250 mm2 could be misleading because the only studies that were analyzed related to the CuT 380 Ag IUD, which has copper on the side arms and in the center. No available comparisons existed with other high copper-load devices positioned only centrally (e.g., Multiload 375). The distinction in the positioning of the copper may be one reason why the CuT 380 Ag IUD has such low pregnancy rates.

This review has a broader title than that of the abstract because a search was made for comparison with all reversible methods. Only 1 study not using a copper IUD was identified as eligible. This was a 3-year comparison with Norplant 2 that involved only 200 women and found no significant difference in measured outcomes other than fewer removals of devices because of bleeding with LNG-20.

The LNG-20 and the CuT 380 Ag IUD are equally effective and acceptable for women with normal menses. Questions remaining include which women are likely to benefit from a specific choice of 1 device or the other and the issue of cost-effectiveness.

Rosemary Kirkman, MB, ChB
Palatine Centre
Manchester, England, UK