A multinational case-control study of ectopic pregnancy. The World Health Organization's Special Programme of Research, Development and Research Training in Human Reproduction: Task Force on Intrauterine Devices for Fertility Regulation.

We conducted a multinational case-control study of ectopic pregnancy in which 1108 ectopic cases were matched by age, parity and marital status with an equal number of pregnant and non-pregnant controls. When cases were compared to pregnant controls there was an increased relative risk (RR) of ectopic pregnancy associated with the use of an intrauterine device (IUD; RR = 6.4) and sterilisation (RR = 10.9) at time of conception, but there was no increased risk associated with oral contraception or other reversible methods. There was an excess of ovarian pregnancies among cases wearing an IUD, and more cases than controls using an IUD had evidence of pelvic inflammatory disease (PID). This suggests that the IUD increases the risk of ectopic gestation possibly by providing greater protection against intrauterine (rather than extrauterine) pregnancy, and by predisposing women to PID and tubal damage. When cases were compared to non-pregnant controls the relative risk of ectopic pregnancy was reduced with all methods, but this protective effect was less marked with the IUD (RR = 0.5) than with the pill (RR = 0.1), other interval methods (RR = 0.2) or sterilisation (RR = 0.2). IUD use prior to conception did not affect the risk of ectopic gestation (RR = 0.7). A past history of PID or sexually transmitted disease (STD) was associated with an increased risk of ectopic pregnancy compared to pregnant (RR = 2.8) and non-pregnant (RR = 2.0) controls. This risk may be higher with multiple episodes of previous PID, but the results were not consistent. Prior PID in the presence of an IUD did not increase the risk over and above that associated with a high risk of recurrence (RR = 7.0 for pregnant and RR = 9.3 for non-pregnant controls). Induced abortion did not significantly affect the risk of ectopic gestation, but spontaneous abortion was associated with an increased risk compared to non-pregnant controls. This finding may have been due to selection bias. Cigarette smoking was associated with an increased relative risk of 3.1 for pregnant and 1.8 for non-pregnant controls. This was not consistent in all comparison groups and may have resulted from confounding, but requires further investigation. We did not observe a significant risk of ectopic pregnancy following gynaecological surgery.