How Operation Rescue and National Right to Life Spin Abortion Research
A new study was released last week that indicated that medical abortions with RU-486 (mifepristone/Mifeprex in the United States) is as safe as surgical abortion. The authors conclude, “We found no evidence that a previous medical abortion, as compared with a previous surgical abortion, increases the risk of spontaneous abortion, ectopic pregnancy, preterm birth, or low birth weight.” Operation Rescue wasted no time putting their spin on the results – this statement is from Operation Rescue Senior Policy Advisor Cheryl Sullenger:
“Reporters are drawing the erroneous conclusion that this study means RU486 is completely safe. That’s not what the study says, and nothing could be further from the truth. Women are dying at an alarming rate from RU486 abortions and its widespread misuse in the abortion industry. That has not changed. Women who have had abortions have greater risks of miscarriage and infertility than women who have not had abortions. It is no accident that the study refused to compare these two groups of women, because we know they would have found that abortion hurts women, and that is obviously a conclusion that they did not want to reach.”
The piece goes on to insist that, “This misinformation campaign is really a horrific thing.” So, let’s check some information. Bear with me, this is a lengthy one, but I want to try to do it justice.
Operation Rescue is correct that media headlines may be inaccurate, as they so often are. CNN’s headline was “Abortion pills don’t boost later miscarriage risks,” although the very first sentence of the article explicitly states that this is in comparison to surgical abortion. Several other media outlets repeated this headline gaffe. I do think headlines are important, which I alluded to in this previous post. However, if one actually reads any of these news pieces, do they actually “draw the erroneous conclusion that this study means RU486 is completely safe?” No, they do not. They clearly state that the study addressed medical vs. surgical abortion. I’m all for holding media outlets accountable for their reporting, but doing so in an honest way matters. If you’re going to call out media bias, it’s important to recognize your own bias. For example, I’m only looking at anti-choice representation of evidence in this post, and am making no claims as to how pro-choice groups represent their evidence.
As for the point that “Women are dying at an alarming rate from RU486 abortions and its widespread misuse in the abortion industry?” – I suppose that depends on what you call an “alarming rate.” How many U.S. deaths has the FDA reported since the drug was approved in 2000? Six. This FDA Q&A reported that at the time of the 1st four deaths, the drug had been used approximately 460,000 times, for a rate of about .87 (<1) per 100,000. CDC statistics for deaths from abortion in general show a rate of <1 per 100,000 cases for almost every year since abortion was made legal. On the other hand, the CDC estimates the maternal mortality rate in 2003 at 12.1 per 100,000 live births. In other words, pregnancy and birth kill many more women each year than abortion, be it medical or surgical. It is misleading to claim that “women are dying at an alarming rate” without putting that rate into a context of overall maternal health and safety.
One thing O.R. gets right is that four of the 6 cases of death involved off-label dosing of the drug – this seems to be an emerging trend as we look at this piece – the statements are just close enough to the truth, without presenting the whole picture, as to be convincing on the face of it. While this off-label use was undoubtedly more common prior to the deaths, when the FDA suggested that vaginal dosing of the drug *might* have been associated with the deaths, Planned Parenthood immediately changed its protocol to administer the drug orally alone. Given that Planned Parenthood is the first name in abortion services in the United States, it is unlikely, if they follow their own guidelines, that this off-label use continues to be “widespread.”
So what of the claim that “Women who have had abortions have greater risks of miscarriage and infertility than women who have not had abortions?” The Operation Rescue release provides no references to back up those claims. National Right to Life makes the same claims, stating, “Overall, women who have abortions face an increased risk of ectopic (tubal) pregnancy and a more than doubled risk of future sterility. Perhaps most important of all, the risk of these sorts of complications, along with risks of future miscarriage, increase with each subsequent abortion,” and does provide some references. The first red flag I noticed is that NRLC’s “medical facts” about abortion still includes the statement, “There is strong evidence that abortion increases the risk of breast cancer,” which has been debunked by the National Cancer Institute.
The four references, provided by NRLC about fertility effects of abortion are:
93 . David N. Danforth, Ph.D., M.D., ed., et al, Obstetrics and Gynecology, 5th ed. (Philadelphia: J.B. Lipincott, 1986), pp. 217, 257, 382-383. See also Jack Pritchard, et al, Williams Obstetrics, 17th ed. (Norwalk, CT: Appleton-Century-Crofts, 1985), p. 484.
94. Danforth, cited above, p. 887, and David H. Nichols, M.D., Gynecologic and Obstetric Surgery (St. Louis: Mosby-Year Book Inc., 1993), p. 260, and Leon Speroff, Robert H. Glass, Nathan G. Kase, Clinical Gynecological Endochrinology & Infertility (Baltimore: Williams & Wilkins, 1983), pp. 156-157.
These are textbooks. The newer (2003) version of Danforth’s Obstetrics and Gynecology reviews possible abortion complications and asserts, “The lay literature erroneously states that abortion causes cancer and reproductive dysfunction. Current medical evidence gives no indication that abortion causes either. There has also been no proven link between abortion and breast cancer. Future reproductive health is not affected by having a medically supervised legal abortion.” Also on the topic of future fertility, the more recent version (2005) of Williams Obstetrics states, “Fertility does not appear to be diminished by an elective abortion, except infrequently as a consequence of infection. Vacuum aspiration does not increase the subsequent incidence of second-trimester spontaneous abortion or preterm delivery. Similarly, subsequent ectopic pregnancies are not increased, except possibly in women with preexisting chlamydial infection or in those who develop postabortion infections.” I don’t have the Speroff work on hand, but can check a 1999 version at the library. While textbooks are never completely up to date, it’s worth noting that the more recent versions of at least 2 of the 3 cited texts contradict the assertion that future fertility is definitely and widely impacted. These texts do not make a distinction within these quotes with regards to surgical vs. medical abortion, but neither does the Operation Rescue passage quoted above.
95. A. Levin, et al, “Ectopic Pregnancy and Prior Induced Abortion,” American Journal of Public Health, Vol. 72, No. 3 (March 1982), pp. 253-256.
Believe it or not, I actually managed to find this rather old article electronically. The authors looked at women with ectopic pregnancies between July 1976 and May 1978 and compared them with women with normal pregnancies, and tried to examine the risk factors for an ectopic pregnancy. The women were interviewed about past pregnancies, prior pelvic infections or surgery, smoking, past induced abortion, and other items. After adjusting for factors such as pelvic infection (not related to abortion) that could also complicate pregnancy and confound the results, the researchers found a relative risk of ectopic pregnancy of 1.3 for women with one prior induced abortion (meaning their risk of ectopic pregnancy would be 1.3x greater than those with no prior induced abortion) and 2.6 for those with 2 or more prior induced abortions. However, the confidence interval for the first figure is 0.6-2.7, and the CI for the second is (0.9-7.4). This is important – because the confidence interval “crosses” 1, the findings have no statistical significance. The CI ranges are a bit like the margin of error you see when political polls are reported – in the first case, the risk could have been up to 2.7 times higher, or it could have been .6, lower than the risk for those with no previous abortion. It’s also important to note that about 1/3 of the women in the study with prior abortions had illegal (and presumably less safe) abortions.
96. Anastasia Tzonou, et al, “Induced abortions, miscarriages, and tobacco smoking as risk factors for secondary infertility,” Journal of Epidemiology and Community Health, Vol. 47 (1993), p. 36.
I can’t get access to the full-text of this at the moment, but according to the abstract, “The logistic regression adjusted relative risks (and 95% confidence intervals) for secondary infertility were 2.1 (1.1-4.0) when there was one previous induced abortion and 2.3 (1.0-5.3) when there were two previous induced abortions.” These figures would suggest that there was some risk associated with previous abortion. It was conducted in Greece in 1987-1988, but abortion was not made legal in Greece until 1986 – it is likely that many of the women in the study with prior abortions had them illegally, although I do not know whether illegal abortions were any more safe than in the United States. A more recent review article (Thorp JM et al, 2003) on the topic cites this as one of two studies that suggest an association with abortion and later subfertility among 7 studies that were examined. The second study was also Greek.
97. A. Levin, et al, “Association of induced abortion with subsequent pregnancy loss,” Journal of the American Medical Association, Vol. 243, No. 24 (June 27, 1980), pp. 2495-2496, 2498-2499.
This is also not available electronically – the abstract states, “No increase in risk of pregnancy loss was detected among women with a single prior induced abortion. We conclude that multiple induced abortions do increase the risk of subsequent pregnancy losses up to 28 weeks’ gestation.“
Look, I obviously think women’s health is important. But I also respect women enough to try to be honest about it. The data on induced abortion and fertility is mixed, with some studies finding increased risk, others finding none, and some finding risk but acknowledging biases and limitations, or that the findings represent a slight increase in a rare outcome. Is this enough evidence to say for certain how or if abortion will affect most women’s fertility? No, it really is not. I can do a PubMed search and selectively choose articles to support either view, which suggests that there is no true consensus that risk is always raised by abortion. Cherry-picking older studies does little to clear up this confusion. The Thrope review article sums it up better than I could:
The long-term health effects of elective abortion are difficult to study and thus poorly understood. This lack of knowledge stems from a variety of causes. First and foremost, exposure to abortion cannot be assigned on an experimental basis, restricting researchers to rely on observational studies and precluding randomized trials. Thus, all research in this realm is prone to an array of different sources of bias that complicate the process of drawing conclusions. Second, it is not clear what group of women constitutes an appropriate comparison group for these observational studies. Third, the decision to terminate a pregnancy is emotionally difficult for many women. Hence, regret, remorse, or shame may cause them to not disclose having made such a decision when queried about their reproductive histories. Fourth, the long-term health consequences of elective abortion have been highly politicized. Those who would grant a moral status to an embryo or fetus and thus limit elective abortion, often use adverse health consequence claims as a tool to further their moral agenda, while those who support no restrictions on abortion access are at times unwilling to consider that pregnancy interruption could affect future mental and physical health. Finally, the effect sizes are small with risk ratios when present falling in the range of a doubling or less of risk for comparatively rare outcomes. The potential for modest influence on events that are unlikely and distant for an individual woman hinders the ability of clinicians or patients to use their experience and judgment to use such information in decision making.
Finally, the authors of the new study don’t deliberately mislead with regard to what they are comparing. They specifically state that first trimester surgical abortion is assumed to be and has generally been demonstrated to be safe, and thus they wanted to see how medical abortion stacked up in comparison. They also note that newer data on abortion and complications generally finds little increased risk to women.
“We chose not to compare women who had medical abortions directly with women who had no prior abortions, since these groups differ with respect to factors that affect pregnancy outcomes, such as socioeconomic status, smoking status, and other health-related conditions and behaviors. The implications of our data for the long-term safety of medical abortion therefore rely on the premise that surgical abortion in the first trimester is safe, which is supported by the majority of studies in the literature. A review of available data published in 1990 concluded that early surgical abortion by vacuum aspiration, currently the most commonly used method, was not associated with ectopic pregnancy, spontaneous abortion, low birth weight, or preterm birth in a subsequent pregnancy. Most studies published since then have supported this conclusion. Among studies reporting increased risks of adverse outcomes in subsequent pregnancies, the findings have been inconsistent; this inconsistency may reflect the performance of multiple comparisons or recall bias in case–control studies of induced abortion and ectopic pregnancies and the use of older methods for abortion in some cases.”
Why does it matter at all? Because anti-choice folks have been trying over the past several years to get the FDA to withdraw approval on the drug. Based on the safety data.