Mifepristone-Misoprostol Medical Abortion Mortality

Mitchell Creinin, MD; Paul Blumenthal, MD, MPH; Lee Shulman, MD


Medical abortion is a non-surgical option for women who elect to terminate a pregnancy. In September 2000, the United States Food and Drug Administration (FDA) approved a medical abortion regimen for women up to 49 days gestation consisting of 600 mg mifepristone (a progesterone antagonist) followed 36 to 48 hours later by 400 mg oral misoprostol (a prostaglandin analogue). Women return to the office 2 days after taking the mifepristone to be evaluated and to receive the misoprostol tablets; follow-up occurs in approximately 2 weeks. However, a number of alternative evidence-based regimens have become common in clinical practice in the United States and around the world. Such "off-label" use of medications frequently becomes the local "standard of care" when there is sufficient evidence that the off-label use is effective and has benefit for the patient. With regard to medical abortion, rigorous research and evaluation in clinical trials of these alternative regimens indicate they offer clinical benefit to patients.[1] In the United States, the most common deviations from the FDA-approved regimen are the following:

  • Use of mifepristone 200 mg;

  • Use of misoprostol 800 mcg vaginally;

  • Decreasing the time interval between the drugs to 6 to 24 hours;

  • Use of the mifepristone-vaginal misoprostol regimen through 63 days gestation; and

  • Supplying the misoprostol at the same time as the mifepristone for the patient to administer herself at home, and follow-up in 1 week with vaginal ultrasound confirmation of pregnancy expulsion.

The evidence indicates that these changes increase efficacy (decrease surgical interventions), decrease side effects, and improve the acceptability of medical abortion to women primarily by shortening the time interval to successful abortion.[1]

In fact, the FDA-approved regimen has been less stringently evaluated than the evidence-based regimens. More women have participated in published trials who have received a regimen of mifepristone and vaginal misoprostol as compared with those who have used mifepristone and oral misoprostol. On the other hand, the FDA-approved regimen has a long history of safe clinical use in Europe and China without any reported deaths due to infection. However, the meticulousness of monitoring and the reliability of reporting, especially from China, are not well delineated.

In the United States, reporting of surgical abortion procedures and complications is a passive process that often meets resistance both at the provider level and at the state level; indeed, 3 states (Alaska, California, and New Hampshire) do not provide abortion data to the Centers for Disease Control and Prevention. Reporting in the United States is further complicated by the social and political environment, which often creates situations in which women may be unlikely to report having had an abortion.[2] Accordingly, we cannot be certain of the true incidence of serious complications and death with early abortion, whether the procedure was surgical or medical. Nevertheless, such problems appear to be rare.

As with any new treatment, especially a politically contentious one, serious complications of medical abortion using mifepristone are more likely to be reported. Seven deaths have been reported to the company responsible for the US distribution of mifepristone and to the FDA since 2001, 4 of which are related to infection with a rare organism, Clostridium sordellii.[3] Although 2 more deaths were recently reported,[4] 1 of these deaths was unrelated to the medical abortion.[5] This organism is also implicated in the deaths of 2 women who had a second trimester miscarriage and received no mifepristone or misoprostol.[6] Notably, deaths through miscarriage, which are not usually reported with considerable detail, appear, like medical abortion-related deaths, to be primarily associated with infection.[7]

These issues have brought us to a crossroads of science and politics. Because serious complications of medical abortion, surgical abortion, and miscarriage are all rare, it is virtually impossible to comparatively assess with any statistical certainty whether one is truly more complicated or risky than another. It is similarly difficult to differentiate the putative causes of increased risk of different treatment options.

What can be said with certainty is that death from abortion of any kind is still rare. Even after accounting for the recent deaths among mifepristone users, the best estimates for the United States confer the following risks:

  • 1/1,000,000 with surgical abortion through 63 days gestation[8]

  • 1/100,000 with medical abortion through 63 days gestation[9]

  • 1/100,000 with miscarriage[7]

  • 1/10,000 with a term delivery[10]

However, we must also keep in mind that these estimates, including those for early surgical abortion, are highly unstable statistically. In addition, we must consider that access becomes important when considering these rates. In many areas of the United States, a medical abortion is actually more accessible than a surgical one. Thus, from an access perspective, a medical abortion that would have occurred at 8 weeks gestation may be replaced by a surgical abortion at 10 weeks. Accordingly, when assessing risk, we may not always be comparing medical and surgical options within the same gestational age range.

The best strategy to decrease the death rate associated with medical abortion in the United States is unclear. It is important for providers, patients, and government officials to maintain perspective -- the risk of death from a motor vehicle accident in 1 year is still significantly greater than the risk of death from a medical abortion. Banning such an option for women may only increase the rate of deaths from surgical abortion performed at a later gestation. This seems irrational. When it became apparent that toxic shock syndrome, a rare but deadly infection, could be caused by prolonged use of tampons, the FDA did not remove all tampons from the market; rather, relevant studies were conducted to understand why infections occurred.

Healthcare providers and policymakers also should not be so quick to blame mifepristone, misoprostol, or both as the "causative" agent. On the basis of all the information we have to date, serious infection and death from medical abortion seem most likely related to the physiologic process of medical abortion, not the medicines themselves.

Some may infer that vaginal administration of misoprostol is the cause of infectious deaths in the United States. However, many medical abortion providers throughout the United Kingdom have switched to using vaginal misoprostol over the past few years, with unsupported estimates that up to half of all medical abortions in the United Kingdom use provider-administered vaginal misoprostol. Given that the absolute numbers of medical abortion performed in Europe far exceed those in the United States, we would expect to be seeing death rates in Europe comparable to those in the United States if vaginal misoprostol was the cause. Moreover, vaginal misoprostol is used routinely for labor induction and second-trimester labor-induction abortion without any reports of serious infection with similar organisms. Typically, with these indications, the provider inserts the misoprostol.

We believe that maintaining use and availability of the most effective options for medical abortion is important for women who are choosing early abortion. Currently, there is no evidence that banning use of vaginal misoprostol for abortion, miscarriage treatment, or labor induction would be beneficial. Without more details about the cases of the women who died and more research into the prevalence, diagnosis, prevention, and treatment of C sordellii, we cannot begin to provide evidence-based suggestions for altering care. Thus, we consider the following to be the most important relevant policy positions:

  • Ensure that all women interested in medical abortion receive appropriate counseling regarding the risk of serious infection and death. Inform them that the risk of death appears to be 10 times greater with medical abortion than with surgical abortion, but that the death rate is very low (approximately 1 per 100,000) -- about the same as the risk of death with a miscarriage, and much lower than the risk of death from a delivery.

  • Consider provider insertion of misoprostol vaginally, via a speculum, as opposed to self-administration. Individual patients would have to consider the inconvenience vs the theoretical risks if vaginal self-administration is the underlying link to these rare infections.

  • Consider adding antibiotics for 1 week with medical abortion. Serious infection from standard sexually transmitted infections appears rare, and metronidazole (an inexpensive and commonly prescribed medication) may be the best option for such treatment.

  • Increase government funding for surveillance and investigation of all serious infections and deaths in pregnant women in order to better understand the relative risks.

  • Increase government funding for surveillance of the prevalence of C sordellii in pregnant and non-pregnant women.

The goal of any medical abortion program is to provide safe and effective services for women who choose to terminate a pregnancy in this way. Protecting the health of our patients is our first priority. However, it is important for both the professional and lay communities to recognize that the evidence-based protocols for provision of mifepristone and misoprostol were developed through rigorous scientific studies and, despite tragedy, we need an equally thorough understanding of the causes of abortion-related death before a single drug, regimen, or protocol is held responsible and deemed inappropriate in this setting.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.