The dishonest claim that abortion pill reversal is dangerous

Photo Credit 愚木混株 cdd20 on Unsplash

Have you ever heard people say that abortion pill reversal (APR) is unsafe, because they abandoned the only randomised controlled trial done on it over safety concerns? In this post, I illustrate just how dishonest this claim is. In short, the researchers abandoned the study because current treatment was (possibly) harming women.

The study enrolled 10 patients. Five had current treatment (i.e. expectant management, not taking misoprostol), while five had APR (i.e. taking progesterone instead of misoprostol). The study was cancelled after 10 patients because three patients had “haemorrhages”.

1. No one has raised safety concerns about current treatment.

The question is whether progesterone is unsafe. But progesterone – at the same dose used for APR – is widely used and recommended by current guidance on threatened miscarriage in women with previous miscarriage. So straight off the bat, there is a general medical consensus that (a) not taking misoprostol (i.e. current treatment) is safe; and (b) taking high-dose progesterone in threatened miscarriage is safe. APR is just the combination of these. So does this study show anything to break this consensus? If so, what?

2. Haemorrhage is a standard complication of medical abortion.

If you give women abortion pills, you should expect some of them to have haemorrhages and attend the emergency department (ED). Indeed, the under-reporting of women attending ED with haemorrhages from medical abortion has been a significant cause of concern over the last few years. But one relevant question is: how should we define “haemorrhage”?

Studies trying to prove the safety of medical abortion have perhaps the highest threshold possible: the need for a blood transfusion (so that the study can ignore any haemorrhage that falls short of this). But in this study, researchers lowered the bar for what constitutes a haemorrhage to one of the lowest possible thresholds. Why? So that they could frame the narrative around how “dangerous” APR is. If they used the transfusion threshold, only one of the participants (in the current treatment group, not in the progesterone group) would have counted as having a haemorrhage. That is not a basis for cancelling a study and refusing to do any further research, even if the haemorrhage were related to APR.

3. Who actually had the haemorrhages in this study?

Of the three haemorrhages, two of the patients had current treatment – and these were by far the more serious events. One of those women needed surgical intervention (as is common after medical abortion). This patient had two previous deliveries and one previous abortion. The other had a body mass index (BMI) of 34.3, was a smoker and a former alcohol and drug user, and had six previous deliveries and five previous abortions. This woman, hardly a typical abortion patient, was at far higher risk of significant complications. She needed surgery (again, common after medical abortion) as well as a blood transfusion.

So two of the five women needed surgery and one also needed a blood transfusion. Does this show that current treatment is unsafe? It’s hard to answer, because of the small sample size. But note that no one is claiming that we need to revise, ban, and/or demonise current treatment as a result. If these haemorrhages were cause for genuine, sincere concern, there would be outcry calling for a new approach to this clinical scenario. But no one has batted an eyelid.

4. How are we defining “haemorrhage”?

What about the women who took progesterone? Of those five, one had a “haemorrhage” and presented to the ED. Again, she was hardly a typical, healthy abortion patient: she had a BMI of 36.4 and previously had seven abortions and a miscarriage. No information about comorbidities was given. She had perhaps the mildest anaemia possible (haemoglobin [Hb] 10.4) and required no surgery or transfusion at all. If the researchers had used the threshold normally used to study the safety of medical abortion (i.e. the need for transfusion), they would not have counted her as having a haemorrhage at all. (Additionally, no baseline Hb was taken, so we have no idea if there was any drop at all.) This woman did not even need surgery, despite surgery being commonly needed after an ordinary medical abortion. Finally, the haemorrhage occurred just two days after taking mifepristone, so even if she had gone through with a normal abortion (taking misoprostol two days later, at home), in all likelihood, she would have had the exact same symptoms.

In short…

In short, two patients receiving current treatment (which no one raises concerns over) had a haemorrhage, both needing surgery and one needing a transfusion. One patient receiving progesterone had a “haemorrhage” but appears to have had the mildest “haemorrhage” imaginable, not requiring any treatment at all.

So, what did the authors conclude? That APR is super dangerous and that the study had to be cancelled. Nothing about the safety of current treatment.

Of course, the media ran with this narrative, failing to raise even the most basic questions about the study and the real reasons for cancelling it. Note thus far the study suggested APR was working (80% baby survival in the progesterone group vs 40% in the current treatment/control group), which would upend pro-abortion narratives about APR. If women bleeding and needing surgery really showed that an intervention was too dangerous to even study, there is no way anyone would ever have authorised medical abortion to begin with. Bleeding and the need for surgery are common – even very common – consequences of medical abortion. But again, the researchers raised no concerns about that.

So, in summary, what is the entire evidence base for claiming that APR is too dangerous to even be studied? That one single woman had a medical abortion, also had some progesterone, and had some bleeding. That’s it. It is a most risible, dishonest evidence base. But because APR offers a woman an actual choice to keep her baby and challenges the narrative that abortion is always what a woman definitely wants, it cannot possibly be allowed.

This post was adapted from a Twitter thread.

If you appreciate our work and would like to help, one of the most effective ways to do so is to become a monthly donor. You can also give a one time donation here or volunteer with us here.