to the editor: Dr. Canavan's article on the appropriate use of the intrauterine device (IUD)1 may have misled physicians and, indirectly, their female patients. In the article, Dr. Canavan stated, “It is important for the patient to understand that the IUD does not appear to be an abortifacient but, rather, prevents conception.”1 Of interest, he cited studies that have been interpreted as not supporting his premise.2
In addition, Dr. Canavan said, “women who use IUDs have fewer ectopic pregnancies than noncontraceptive-using control subjects.”1 If, as he asserts, the exclusive action of the IUD is to prevent fertilization (and then there would be no possible postfertilization effect), then any reduction in the percentage of intrauterine pregnancies should be matched by an equal reduction in the percentage of ectopic pregnancies. However, several studies that were not mentioned by Dr. Canavan have shown an excess risk of ectopic pregnancy versus intrauterine pregnancy among patients who use IUDs.3–6
Dr. Canavan's failure to discuss these studies is critically important. Why? Because the results of these studies, which clearly show an increased risk of ectopic pregnancy, can only be explained by greater success in preventing implantation than in blocking fertilization and/or by excess ectopic pregnancies caused by the IUD. Both possible explanations represent postfertilization effects; the mechanism that Dr. Canavan feels is “important for the patient to understand” does not occur.1
In a systematic review2 of the mechanisms of action of the IUD, published in 1997 in the American Journal of Obstetrics and Gynecology, Dr. Spinnato wrote, “Several important conclusions can be drawn from the available literature regarding the mechanism of action in IUDs: (1) The available evidence supports a continued significant role for a postfertilization mechanism of action of IUDs. (2) There is little compelling evidence to suggest that . . . IUDs reliably eliminate the likelihood of fertilization.” Dr. Spinnato concluded,“the analysis of the evidence strongly suggests that the contraceptive effectiveness of [IUDs] is achieved by both a prefertilization spermicidal action and a postfertilization inhibition of uterine implantation.”2
It seems obvious that, in order to accurately reflect the medical literature, complete and accurate informed consent about the use of IUDs should include explanation of the possible postfertilization action of the IUD. This information would be especially important for patients with a moral objection to this effect. Because it would be difficult to predict which patients might object, the possible effect should be disclosed to all patients. Therefore, any patient education handout that lacks information on this potential effect represents, at the very least, a failure to provide the patient with complete informed consent. If the potential effect violates the moral requirements of a woman and she is not informed of the possibility, then this failure of disclosure seriously jeopardizes her autonomy. Further, I believe that conscious withholding of such information constitutes unethical deception.
Unfortunately, the patient education handout that was published with Dr. Canavan's article is incomplete, inaccurate and potentially misleading when it says, “The IUD prevents sperm from joining with an egg.”1 Even if Dr. Canavan rejects Dr. Spinnato's conclusions, I feel that Dr. Canavan has failed the readers of American Family Physician. Furthermore, AFP was potentially remiss in not letting its readers and their patients (via the patient education handout) know of the potential postfertilization effect of the IUD. Hopefully, the handout will be revised to sensitively and accurately address the potential postfertilization effect of the IUD, which is recognized in the medical literature but not discussed in Dr. Canavan's article.
in reply: I am pleased to respond to Dr. Larimore's letter, as I am very sensitive to his concerns about the mechanisms of action of the intrauterine device (IUD). A large volume of data exists on the mechanisms of action of the IUD. However, after weighing the evidence, I felt that the preponderance of data supports the conclusion that the IUD works by preventing fertilization. Dr. Larimore's evaluation of similar data with an alternative conclusion clearly shows that the data are open to different interpretations. A study with absolutely irrefutable evidence and unquestionable statistical power is presently unavailable and is unlikely to become available.
Dr. Larimore asserts that, “any reduction in the percentage of intrauterine pregnancies should be matched by an equal reduction in the percentage of ectopic pregnancies,” and discusses several studies that have found an increase in the ratio of ectopic pregnancies to intrauterine pregnancies in IUD users. These studies are unreliable. They present data on IUDs that are not available in the United States. The IUD that is used in the United States and discussed in my article is the TCu 380 (ParaGard), which may act differently than other IUDs because of its high copper content.
Moreover, no attempt was made in these studies to compensate for the age-related increase in ectopic pregnancies. The study findings may be affected by sample bias because other risk factors for ectopic pregnancy were not known, such as vaginal douching, cigarette smoking and number of sexual partners.1 Also, the number of pregnancies in patients using an IUD is so low that statistical power is poor. Although the Marchbanks study1 reported an 11.9 percent relative risk for ectopic pregnancy in IUD users, the 95 percent confidence interval was 2.3 to 62.0 percent.
In the Marchbanks study,1 an increased risk of ectopic pregnancy with current use of an IUD was noted, but the researchers admitted that “any condition that prevents or retards migration of the fertilized ovum to the uterus could predispose a woman to ectopic gestation.” A study by Cramer and associates2 found that patients who have used an IUD in the past have an increased rate of tubal infertility, which suggests that tubal function is affected by the IUD through an inflammatory or infectious etiology. Because these factors are both associated with ectopic pregnancy, the increased ratio of ectopic pregnancies reported in some studies may reflect these effects.
Sivin3 found that the rate of pregnancy (and the rate of ectopic pregnancy) varied inversely with the copper surface area of the IUD. These findings would support the dose-related spermicidal effect of copper. If we assume that an implantation effect is related to the presence of a foreign body, it would be highly unlikely that increasing the copper content of the IUD would decrease the number of pregnancy failures and ectopic pregnancies.
Dr. Spinnato's article,4 quoted by Dr. Larimore, was published after my article was written; time did not permit me the opportunity to present his opinions. Dr. Spinnato repeatedly points out the limited value of available clinical trials but leaves out numerous studies that contradict an implantation effect of IUDs. Sivin reviewed this same information and their data, which showed strong support for an exclusive contraceptive effect.5
It is often difficult and, sometimes, harmful to present to a patient all of the extremely rare risks of any device or procedure. The possibility that the IUD may cause an implantation failure seems remote enough to make it unreasonable and inconsistent with present standards to place it on patient information material. Many medical procedures carry severe risks that are not disclosed to the patient because the risks are so remote. Patients who seek preconception counseling are rarely informed of the risk of death from childbirth, just as mothers who are considering cesarean section are not informed of the risk of death from such surgery. These discussions should come directly from the physician to those patients who wish to know about these extremely rare risks.
My opinion remains that the TCu 380 IUD represents a very safe and effective form of contraception for selected patients.
editor's note: Since this issue is so emotionally charged, upon review of both sides of the argument, I believe it is prudent to acknowledge the controversy regarding whether IUDs exert an implantation/abortifacient effect. For this reason, we have modified the electronic/Web-based version of the patient education handout on IUDs that accompanied this article to acknowledge both viewpoints.