NRL News

Study Touts Telabortion as Safe, Effective, Ignores Fate of Hundreds of Missing Patients

by | Feb 22, 2024

By Randall K. O’Bannon, Ph.D., NRL Director of Education & Research

As certain as the summer sunshine, you can always count on the abortion industry to have some new “study” lined up as soon as they’ve got a new abortion policy or method to defend. The latest is “Effectiveness and safety of telehealth medication abortion in the USA,” by Ushma Upadhyay and a team from the University of California, San Francisco (UCSF) which appeared in the journal Nature Medicine, published online February 15, 2024.

Not surprisingly, it concluded that “Telehealth medication abortion is effective, safe and comparable to published rates of in-person medication abortion care.”

UCSF researchers claim that after examining thousands of cases where women got their abortion pills by mail after an online, email, or text interview, 97.7% of these were effective and less than one percent suffered any serious complication.

Those familiar with these drugs and knowledgeable as to how they work recognize that there is some serious spin going on here. A closer examination of the data and the unwarranted assumptions of the study continue to show a process still fraught with many serious problems for women.

Safety issues long an obstacle to home use

From its earliest days, it became apparent that the abortion industry wanted chemical abortion – abortions using pills like mifepristone and misoprostol – to be something that could be done outside the clinic, “in the privacy of one’s own home.” There were a number of safety issues with the drugs, however, prompting the government to impose some safeguards. They required that the patients get screened in person, have their pregnancies dated, confirm that they did not have an ectopic pregnancy, and have a plan in place in case they had bleeding or other issues. The pills’ effectiveness declined, and the likelihood of complications rose, the farther along a woman was in her pregnancy.

It is very important that women understand that mifepristone does not work in circumstances of ectopic pregnancy.

At the industry’s insistence, the Biden administration did away with required in person visits in 2021, opening the door to the mailing of abortion pills and telehealth abortions.

There is ongoing dispute about the safety of such abortions, evidenced by the case of the Alliance for Hippocratic Medicine v. the U.S. Food & Drug Administration now heading for the Supreme Court. The Alliance challenged the safety and legality of the government’s recent moves.

Why this study appears now

Now, right on schedule, the abortion industry puts out a study claiming to provide evidence that these abortions are safe, safe, safe, just as safe and effective as the ones where the woman comes to the office, is personally screened and examined by a physician before being given the pills and is then monitored by his clinic staff.

Those who’ve watched them do this for the past three decades or so know the drill. The researchers get access to the abortion pill provider’s medical logs and records, note the number of patients, the scarcity of recorded complaints, and then pronounce the new method or modification absolutely safe and effective.

But those of us who’ve read these studies and studied them carefully, know that there’s a lot that gets minimized and left out of the press release, things that potentially paint a much different picture.

More than 1,800 patients missing from the study analysis Start with the number of patients involved (“more than 6,000″ says the New York Times 2/15/24) and the claims that a high percentage of these abortions (97.7%) were “effective.” 

But the 97.7% applies not to the original 6,974 patients whose records they sought, or even the 6,154 whose charts showed them being sent the “medications.” The number applies only to the 4,454 for whom the abortion outcome was known.

Maybe these pills were generally “successful” for women who responded to follow up from the virtual clinic and whose outcomes were “known.” Nonetheless, this represented less than three quarters of the original sample!

Why trust a remote stranger?

The authors want you to assume that their effectiveness figures held for the rest of the sample or might have even been better, but this makes a lot of unwarranted assumptions.

Why would a woman receiving pills from people she has never met in person from a virtual clinic maybe hundreds (or thousands) of miles from her home contact them when she has a problem?  What could they do other than send her to the nearest emergency room, which she can easily and more quickly do herself?

The point here is that the authors really only know for certain that these pills were “successful” for 4,352 of the 6,154 patients who received them. That would be an “effectiveness” rate of just 71%, a far cry from the advertised 97.7%, and that leaves a lot of room for “failed abortions” or other complications.

In truth, those pills are probably “effective” for some of those who do not recontact the virtual clinic. But one cannot simply assume, as the authors of this study do, that they worked as well for those who followed up as for those who, for whatever reason, dealt with the consequences in some other way.

Basis for determination of effectiveness

Before turning to safety issues, one more caution about “effectiveness” – the ability of abortion pills to end pregnancy. The authors of the study were not relying on the examination of fetal remains or the results of later physical examination. Instead, they say outcomes were “ascertained using a test or the patient’s history.”

Thus, what they know is not that the pills worked, but that the record noted a successful outcome. People familiar with these kinds of abortions know that is possible to have heavy bleeding and pass large clots without actually aborting the child. The child could survive, or the child could die and remain in the womb, leading to infection or other problems. The point is, without an actual test or physical examination, one cannot be sure.

Ambivalence and APR misinformation

Another point about effectiveness. We already know, from the study, that some of the women (120) who were sent the pills did not take them. Though these were not included in the final effectiveness analysis they do indicate some ambivalence among patients regarding their wishes.

Among the patients who did receive and take at least the first pill, authors say there were two who “requested abortion pill reversal” This alone is proof against claims of abortion pill promoters that women never change their minds after initiating this process.

So how did sellers of these pills respond? Despite thousands of cases and published studies attesting to the high rates of abortion pill reversal (APR), Upadhyay and team says patients “were advised that evidence-based reversal treatment does not exist and referred to urgent in-person care.”

Rather than being given progesterone to try and block the effects of mifepristone and potentially save their babies, it seems that they were shipped out to their local emergency rooms and told to wait and see what happened. Other studies have shown this to be ineffective at reversal and potentially dangerous.

Unbelievable claims of abortion pill safety

Claims of “safety” are at least as dubious as those regarding “effectiveness,” but the data is nonetheless revealing.

Again, one cannot make any solid claim about the safety of the 1,802 patients who received the pills but did not report back to the virtual clinic. As noted above, there is legitimate reason to believe that women dealing with hemorrhage, infection, a ruptured ectopic pregnancy, or an ongoing pregnancy did not turn to far away online strangers or faceless phone operators from a virtual clinic on the other side of the country. Instead, they were more likely to seek the help they needed from a local doctor they knew or an area emergency room that could actually treat or address their problem.

If that is likely and logical, then the complications reported by

Upadhyay would represent just the tip of the iceberg, at best maybe just some hint of the type of issues other women might be facing.

Defining away a host of safety problems

Naturally, as reliable defenders of the abortion pill and the abortion industry, Upadhyay and team claim that complications were exceedingly rare, with 99.7% reporting “no major abortion related adverse events.” If you’ve followed abortion pill studies for some years, you know a lot of the work here is done by limiting what it is that can be counted as a “major abortion-related adverse event.”

This information cannot be found in the online article, but in a supplement. The supplement specifically identifies “serious adverse events” as those including “blood transfusion, abdominal surgery (including salpingectomy, laparotomy and laparoscopy to treat ectopic pregnancy), hospital admission requiring overnight stay, or death.”

While these are certainly serious outcomes, the fact that they occurred at all is significant. But note what it is that isn’t included.  A woman could hemorrhage, be gushing blood, rushed to the emergency room, be given uterotonics to stop the bleeding, etc., but so long as she is not given a transfusion or admitted to the hospital for an overnight stay, it does not count as a major complication.

Surgery doesn’t count?

It is unclear whether further surgical treatment will merit the more serious classification. We know that surgical procedures like salpingectomy or laparotomy to remove the child in an ectopic pregnancy count. But what about surgery to stop the bleeding, or to complete the abortion?

We know from data published in the study that 63 women, 1.4% of their effectiveness sample, required a “Procedure, aspiration or surgery,” more than would fit in the 0.3% they indicate had “major adverse events.” So clearly, for the authors, it was possible to have a complication severe enough to require surgery without them counting it as a major complication.

That’s how you come up with the sort of claims you commonly read in the popular media of “serious complications” being less than one percent (e.g., Washington Post, 4/10/23), despite much higher numbers ending up in the emergency room.

Visits to the Emergency Room

Indeed, even in this study, 81 patients, nearly 2% of their sample (that is, those for whom they had information) visited the emergency room. Though obviously not a big deal for the researchers, in context, even this number is very concerning. 

The sales pitch for telemedical abortion is that women can avoid having to go to the clinic. They can order their abortion pills online, wait for their package to arrive, and have their abortions at home.

When you pick up your pills in person from the clinic down the street, you have someone to call, somewhere to go when or if you have a problem. For example, when you begin gushing blood, when you start running a fever, when the pain becomes unbearable, or when you sense that something is going wrong beyond the usual bloody, painful chemical abortion. 

You should be able to get help when you need it.

An issue in particular for rural patients

That isn’t always the case for the chemical abortion client of the virtual clinic. The prescriber isn’t typically someone they know, someone available to provide emergency surgery, and may not even live in the same state.

If they come from a suburban or rural community, like nearly one in ten (9.9%) in this study did, they may not be anywhere near the sort of qualified medical help they need. That could put their life at risk even if they aren’t presenting with the sort of complication Upadhyay and her team count as “major.”

Implications for the broader U.S. population

In general, even if you take them at face value, there is a willful, callous blindness to the implications of their study. If 56% of abortion in the U.S. are now chemical (see the U.S. Centers for Disease Control’s 2021 Abortion Surveillance), and there are somewhere between 900,000 and 1 million abortions a year, then even what looks like a small percentage can represent thousands of women.

A failure rate of 2.3% against a backdrop of 950,000 annual abortions, with 56% or 532,000 being chemical abortions (the number today is certainly higher) would mean more than 12,000 women might require some “intervention to complete [the] abortion.”  Even if they ignore the unknown outcomes of 26% of the patients in Upadhyay’s study and accept the reported 1.8% of remaining patients visiting the emergency room, we’re still talking about more than 9,500 patients.

But if there is a higher failure or complication rates for the women the study was unable to track, these numbers could mushroom.

Even researchers doubt gestational dating

Claims of safety and efficacy of telemedical abortion depend in part on the ability of virtual prescribers being able to determine gestational age and identify ectopic pregnancies without ultrasound or a direct physical examination. If they are to be believed (and one ignores the fate of those patients lost to follow up), they appear to have been unusually successful in this screening.

The records Upadhyay and her team examined from virtual clinics had no cases of women over 70 days attempting to abort, despite advertising by Planned Parenthood and others for abortions pills at 11 weeks or more and adverse event reports by the FDA indicating widespread use of mifepristone on women with gestational ages weeks past the FDA mandated 10 weeks LMP (after a woman’s last menstrual period). There were four cases where the gestational age was unknown, though.

This means that women were somehow unusually knowledgeable and forthcoming about the timing of their pregnancy or that prescribers were exceptionally good at verbally screening their patients. However, another explanation is that some women were neither accurate nor honest in their assessments and that records reflected merely reported gestational ages rather than actual, accurate ones.

Even the authors of the study were somewhat dubious. Upadhyay and her team remarked that they were surprised that there were no unexpected pregnancy durations beyond 70 days (previous studies implied there should have been at least a few) but say this may be due to underreporting by patients. The UCSF team speculated that women who may have underestimated their gestations may not have reported their errors to the clinic.

Still missing a few ectopic pregnancies

Ectopic pregnancy is said to occur in 1-2% of all pregnancies, where the embryo implants outside the womb. If left untreated, it can result in the rupture of the fallopian tube and may lead to death of mother and child. Mifepristone is not able to treat ectopic pregnancy.

Despite what appears to be a remarkable ability on the part of prescribers to screen their patients for ectopic pregnancy, the study shows a few got through.

Given the percentages above, there should have been some 60 -120 ectopic pregnancies among the initial 6,034 patient records which researchers first examined. It appears that the virtual screeners were able to catch many of these by verbal screening, with data indicating that 486 women were referred for ultrasounds for confirmation of gestational age or intrauterine pregnancy.

Still, six patients reported ectopic pregnancies, including at least one that ruptured. Again, if help is available and these are treated, at least the mother should be able to survive. If these occur at an inopportune time and place where qualified, direct medical assistance is not around, it could prove disastrous.

Again, if ectopic pregnancy affects 1-2% of mifepristone patients, as it does the rest of the population of reproductive aged women, that would potentially impact five to ten thousand women a year. But even they were somehow subjected to the most brilliant, insightful verbal screening by text, email, or webcam, as reported here, and it affected just .013 of those patients, that would still be more than 700 patients a year in the U.S., putting a considerable number of women at unnecessary risk.

Just as good as in person care?

One of the contentions of researchers is that this study demonstrates that telehealth abortion or abortion managed by a virtual prescriber is as safe as those managed directly by an in-person prescriber.

This depends on so many questionable assertions and assumptions that the claim is virtually meaningless. If both in person and telehealth chemical abortions depend on studies where a sizeable number of patients are missing from safety and efficacy calculations—and excluding those patients who are the most likely to experience problems—then the results from both types of studies will be flawed.

Studies may be fairly consistent with each other, reporting what appears to be largely matching data. But that is because each is generally limited to those patients who are both most satisfied with their abortions and most likely to report back. The outcomes that truly reveal the safety and efficacy of these abortion drugs are those of the patients for which none of these studies have data.

Data from other western countries paint a much different safety and efficacy profile

This is why studies from other countries with national healthcare systems and other ways of collecting and recording data get much different results. This, despite using the same drugs and having similar (or better) patient profiles.


In 2009, researchers from Finland found 20% of chemical abortion patients reporting adverse events (e.g., hemorrhage, incomplete abortion), nearly four times the “adverse events” reported for standard surgical abortions. Under pressure from chemical abortion defenders, the authors agreed that some of these adverse events may not have been as serious as others. That these events were still significant enough to prompt women to re-access the health care system with these concerns needs to be noted.


Research from Britain, which recently adopted its own “Pills by Post” program, mailing abortion pills to women’s homes, also shows lower efficacy and higher complication rates. While the government was claiming extremely low adverse event rates (identifying just one single complication among 23,061 chemical abortion patients), a researcher who contacted the hospital system directly found that 5.9 % reported complications connected to an incomplete abortion with “retained products of conception,” Three percent of patients required surgery to deal with incomplete abortions and 2.3% of patients were treated for hemorrhage in Britain’s National Trust hospitals (Percuity, 10/27/21).

These numbers are not as high as they are in Finland, perhaps because the United Kingdom has a higher population density, giving patients readier access to emergency help. But they are still higher than those reported by Upadhyay and her team from California.

British newspapers wrote about increased calls to ambulances once this program was in place. These calls jumped once “Pills by Post” was instituted, increasing by more than 50% in some areas, up at least 25% in others (London Daily Express, 4/25/23)


A Canadian study from 2023 in the online edition of the Annals of Internal Medicine (January 3, 2023) found even higher complication rates among patients picking up mifepristone prescriptions from pharmacies, similar to the protocol recently authorized here in the U.S. Among the 39,856 patients in that study, emergency room visits were reported by 10.3% — at least one out of every ten patients!

All these foreign studies, performed in modern western medical environments, utilizing the same drugs and the same basic protocol, reported much lower efficacy and much higher complication rates. The only significant difference appears to be that they were performed by private or more neutral state observers rather than those regularly published and cited in the U.S. by members of the medical establishment involved in and committed to the promotion and performance of chemical abortion.

Close scrutiny of other American studies reveal higher complication rates

That said, when examined more closely, even American studies reveal a much less positive safety profile. An earlier study of emergency room visits in 2015 by Upadhyay herself is often one of the ones cited as proof that the rate of serious complications with mifepristone is “less than 1%.” Indeed, in “Incidence of emergency department visits and complications after abortion,” from the January 2015 issue of Obstetrics & Gynecology, Upadhyay officially found that “The major complication rate was 0.23%,” less than a quarter of one percent.

But again, this depends on several of these now familiar questionable moves to finesse the data. First, Upadhyay specifically limits what can be counted as a “serious” or “major” complication. “Major complications were defined as serious unexpected adverse events requiring hospital admission, surgery, or blood transfusion,” the article asserts. “Minor complications were all other expected adverse events.”

While this sounds reasonable, consider the things included in Upadhyay’s “minor complications”: hemorrhage, infection, incomplete or “failed” abortion requiring “uterine aspiration” (i.e., surgical abortion). Even things like “uterine perforation” were classified as “minor.”

Second, with this knowledge, consider that when Upadhyay added in and counted both major and “minor” complications, the complication rate for chemical abortions was 5.19% – considerably higher than the “less than one percent” advertised.

Calling these complications “minor” diminishes the significance of the fact that these incidents were sufficiently serious as to prompt so many of these patients to visit their local emergency rooms.

And this was among only those who somehow revealed their chemical abortion attempt when many of those promoting and selling these pills were telling them they didn’t have to.

Summing Up

As distressing as it is that the abortion industry has put out yet another misleading study on chemical abortion, worse is that the media spin and reaction is likely to be exactly the same as that of researchers – ignoring the hundreds or even thousands of desperate women who are missing from the counts.

There will likely be no place in the news media stories for the terrified, hemorrhaging women, doubled over in pain who spurn the faceless, useless bureaucrats of the virtual clinic for the more direct, personal, accountable help of their own local doctor or emergency room.

These women and their complications won’t get counted in studies like Upadhyay’s. They may not even be recorded in hospital reports if women follow the advice of many abortion pill prescribers and don’t reveal their chemical abortions to ER staff.  Their risks and injuries will be very real, nonetheless.

Relying on reports from abortion pill suppliers and prescribers, U.S. abortion industry studies are structured in such a way as to miss this critical data. But studies from other countries better track those patients and reveal much poorer safety and efficacy rates with mifepristone.

If American women want to know the truth about mifepristone safety and efficacy, if the U.S. Food & Drug Administration wants to have an accurate assessment of mifepristone’s risks, they’re going to have to stop relying on data sifted and spun by abortion pill promoters.

Categories: Abortion Pill