BACKGROUND: Insertion of an intrauterine device (IUD) immediately after an abortion has several potential advantages. The woman is known not to be pregnant, a major concern for clinicians. For example, many clinicians refuse to insert an IUD in a woman who is not menstruating [Stanback 1997]. After induced abortion, a woman's motivation to use contraception may be high. Among women who have limited access to a clinician, abortion care may provide a unique opportunity to address a woman's need for contraception [Mahomed 1997; McLaurin 1993; Wolf 1994]. A copper IUD confers nearly the same contraceptive efficacy as does tubal sterilization [Peterson 1996], yet it is simpler, less expensive, and promptly reversible. In addition, insertion of an IUD immediately after abortion may avoid discomfort related to insertion, and any bleeding from the insertion will be disguised by the expected bleeding after abortion. However, insertion of an IUD immediately after a pregnancy ends carries potential risks as well. For example, the risk of perforation may be increased due to softening of the myometrium. One IUD manufacturer in the United States warns that "The PROGESTASERT(R) system is not intended for immediate postabortion or postpartum insertion It should not be inserted until involution of the uterus is complete. The incidence of perforation and expulsion is greater if involution is not completed." [ALZA 1998] The package labeling for the Copper 7 stated that it should not be inserted after an abortion until involution was complete "in order to avoid a higher incidence of perforation or expulsion."[Gillett 1980] Another potential concern is infection. Insertion of an IUD after a clandestine or unsafe abortion may increase the risk of upper genital tract infection compared with interval insertion (remote from pregnancy). OBJECTIVES: To assess the safety and efficacy of IUD insertion immediately after spontaneous or induced abortion. SEARCH STRATEGY: We used Medline, Popline, and EMBASE computer searches, supplemented by review articles and contacts with investigators. SELECTION CRITERIA: We sought all randomized controlled trials that had at least one treatment arm that involved IUD insertion immediately after an induced abortion or after curettage for spontaneous abortion. We identified 12 trials which described random assignment but excluded three from this review. Two of these revealed unethical research conduct, and one used alternate assignment to treatments. DATA COLLECTION AND ANALYSIS: We evaluated the methodological quality of each report and abstracted information onto a data collection form. We focused on gross discontinuation rates (single-decrement life table analysis) for accidental pregnancy, perforation, expulsion, and pelvic inflammatory disease. We entered the data into RevMan 3.1 for analysis of Peto odds ratios. MAIN RESULTS: In large multicenter trials, the TCu 220C device proved superior to either the Lippes Loop D or the Copper 7 IUDs for immediate post-abortal insertion. The Peto OR for discontinuation for pregnancy was 0.4 (95% CI 0.2-0.7) compared with the Lippes Loop D and 0.5 (95% CI 0.3-0.8) compared with the Copper 7. Expulsions were also significantly less common with the TCu 220C than with the other two IUDs. In single-center trials, the Nova T IUD had a signicantly higher discontinuation rate for pregnancy than did the Multiload 250 (OR 4.5; 95% CI 1.0-19.8), while the Nova T had a significantly lower discontinuation rate for pregnancy than did the TCu 200 (OR 0.3; 95% CI 0.1-0.9). Only one trial compared immediate vs. delayed insertion. In this trial, the performance of the Copper 7 IUD inserted immediately after abortion was inferior to that after interval insertion (remote from pregnancy), although the differences were not statistically significant. Addition of copper sleeves to a Lippes Loop D improved its performance, but addition of a topical hydrogel to a Spring Coil did not offer benefit. Overall, rates of perforation and pelvic inflammatory disease were low. REVIEWER'S CONCLUSIONS: Insertion of an IUD immediately after abortion is both safe and practical. This was true for both induced and reported "spontaneous" abortions, many of which may have been induced under clandestine circumstances. IUD expulsion rates were higher after second-trimester abortions than after earlier abortions, so delaying insertion may be advisable after later abortions. Although Copper 7 devices inserted unrelated to pregnancy may perform better than those inserted after abortion, many women who express interest in intrauterine contraception do not return for the scheduled insertion. These women may remain unprotected against unintended pregnancy. However, evidence is inadequate to compare the safety and efficacy of IUD insertion immediately after abortion vs. insertion some weeks later.