Absence of chorionic gonadotropin in sera of women who use intrauterine devices.

A controlled study was undertaken to determine whether unnoticed pregnancies routinely occur in users of the intrauterine contraceptive device (IUD). Starting on day 10 of the menstrual cycle and continuing through the onset of menstruation or until the diagnosis of pregnancy, we collected daily blood samples from three groups of normally menstruating young women. The study groups were (A) IUD users (n = 30), (B) women with tubal ligation (n = 30), and (C) women trying to become pregnant (n = 15). The sequential serum samples were analyzed by radioimmunoassay for progesterone (P), human luteinizing hormone (hLH), and human chorionic gonadotropin (hCG). No positive hCG assays in luteal phase blood sera of IUD users were observed. The only positive hCG determinations of IUD users coincided with the preovulatory surge of hLH. Two subjects who became pregnant, as judged by progressive increases in hCG and P levels in the luteal phase, belonged to the group planning pregnancy. The finding of two pregnancies in 15 months of exposure is consistent with the assumption of natural fertility. The probability of no pregnancies in 30 months at risk, as observed among the IUD users, is between 1 in 200 and 1 in 100,000, depending on the assumption made for natural fertility. The study demonstrates that IUD users do not retain their natural fertility, and that IUDs do not exert their antifertility effect as abortifacient agents. If a confirmed pregnancy is detected in an IUD user, it may be assumed to represent an isolated case of contraceptive failure.