In this issue of Midwifery Today Jennifer Enoch presents an excellent, thorough review of the use of misoprostol (Cytotec) for induction (1). A careful reading of this paper, however, raises a number of urgent questions: Misoprostol is on the market as a prescription drug because the Food and Drug Administration (FDA) has approved misoprostol for stomach problems, but not for induction of labor. Why not? What does the FDA say about this “off-label“ use? What does the company manufacturing the drug say about this use? What do the scientific data show, and what do scientists say about this ongoing off-label use? A brief review of the evolution of the use of misoprostol for induction clearly illustrates several problems related to obstetric and midwifery practice in the United States.
In South Dakota three months ago, an obstetrician bragged to me over lunch that he had introduced into his community the use of Cytotec for induction. When questioned, he admitted knowing the FDA does not approve such use of this drug but that nevertheless he does not inform women it is not approved for induction, nor does he ask for their informed consent. He scoffed at my suggestion that he is experimenting on women without their knowledge, much less consent. His excuse: “We will wait forever for the bureaucrats in Washington, D.C., at the FDA to approve drugs, so we must try them out ourselves if we want progress.”
One month later, in Oregon, a local doctor told me (and repeated it on her local weekly TV Health program) that obstetricians in Medford told her they are thrilled with Cytotec for induction because they can bring women in first thing in the morning, give them Cytotec and have the babies out before 5 p.m.—a welcome return to daylight obstetrics. None of the hospitals in the Medford area required informed consent when using Cytotec for induction. The Oregon State Health Department told me that while collecting their statewide data on induction, they have observed that Cytotec has now become the most common method of induction.
While the United States has a system in place to ensure that all drugs must be evaluated by the FDA before being allowed on the market and that certain drugs are to be dispensed only through physician prescription, there is a hole in this system. Once a drug has been approved by the FDA for one use and put on the market, there is nothing to prevent a physician from using that drug for whatever use he wants and at any dosage. Trials for new uses of drugs are important as long as the trials are done as research and everyone understands that this use is experimental, with informed consent from the patient. Misoprostol induction shows potential for certain benefits, but these benefits must be documented by careful research that at the same time looks carefully for the risks.
We can’t just throw drugs at people in an uncontrolled way. If a practitioner hears about a new use and simply starts using the drug this new way, this is experimenting on patients without the usual safeguards in place for research subjects. And while practitioners should report to the FDA on such off-label trials and should always report to the FDA any side effects and risks found, in reality only a very small number of practitioners ever report anything to the FDA. As a result, a large information vacuum exists in the United States with regard to what prescription drugs are being used for which purposes and what side effects and serious risks have occurred.
So when practitioners simply begin to use a drug for a new purpose, there follows a phenomenon I have experienced for years as a practicing clinician but rarely see described in print—the informal spread of clinical experience. In hospital corridors, lunchrooms and staff lounges, doctors, midwives and nurses share their ideas and experience.
A recent technology makes it possible to listen in on such clinical chat—online Web pages and chat rooms. By accessing the World Wide Web and then using key words such as “misoprostol“ and “pregnancy,” many Web pages and chat lines appear. Clinicians, scientists, policy-makers and patients should read these Internet pages from time to time. While clinicians writing on the Web are not necessarily representative of all clinicians, it is possible to discover how at least some of today’s clinicians think and act. It seems like eavesdropping because of their candor and their blunt way of expressing ideas and opinions and revealing their attitudes.
It is the informal communication of uncontrolled clinical experience that has driven the spread of misoprostol induction, as is apparent from the following actual statements taken from the Internet in 1998 (2):
“Cytotec is extremely effective at very low doses, is very cheap, and has been used on many, many women without their being aware that it really is still an experimental use.“
“I must say that I have heard some great things about Cytotec myself. I know some people who have used it and say that they have pretty good luck with it. It sounds like your ladies are pretty happy with its effects—two-hour labors and such. Just be careful. I would have to say that the biggest danger is leaving the woman alone. The stuff turns the cervix to complete MUSHIE (Web message emphasis, not mine) and opens it with a couple of contractions. So whatever you do, remember that you must not stay gone too long.”
“At my suggestion our high risk OB referral hospital tried Cytotec—one-half tab per vagina—and after two cases of hyperstimulation stopped its use.“
“We’ve seen no cases of hyperstimulation after Cytotec that did not respond to a two-gram bolus of MgSo4. You can almost count on a delivery 12 hours after inserting the Cytotec tablet.”
“We are using it at Yale and although there is a format for how to give it, there is still controversy on to whom to give it. Pharmacy uses one of their nifty little pill cutters and sends us one-fourth of a 100 microgram tablet (remember this stuff was made for treatment of ulcers!).“
“I’ve personally used it twice and had excellent results in women wanting homebirths, but going postdates. I’m attaching my own protocol for anyone interested. Again I warn that I am no expert and I consider this protocol to be a ”work in progress“—it will certainly change as I gather experience and information about this drug.“
“We are using misoprostol regularly for induction—my department loves it. We use one of the protocols published on OBGYN.Net Web page.”
“Our biggest fear is that the company will pull Cytotec from the market, since our internist/GI buddies tell us that it isn’t worth a darn for its labeled indication.“
What is apparent from this Internet medical practice is the lack of appreciation of any borderline between experimenting on patients and practicing medicine on patients and the absence of concern for patients’ rights to informed consent.
Also apparent from reading the Internet is the inability of many clinicians to critically review published papers. The general assumption is that since there are, as stated in one Web message, “gobs of references” (2), the scientific work has been done, and it is OK to use this drug for this purpose. The tendency for clinicians to misinterpret scientific papers is in part because of a difference in approach, since scientists must believe they don’t know, while clinicians, in order to do what they do, must believe they do know. A common attitude among clinicians, revealed by Internet messages, is that pregnancy and birth are dangerous until proven safe, while technology and drugs are safe until proven dangerous.
In the review of published randomized controlled trials (RCT) of misoprostol induction, Enoch mentions several weaknesses of these trials (1). A number of additional weaknesses of the trials Enoch reviewed must be added to produce the following, more complete, list:
- All trials compare misoprostol with another inducing drug, but not with nonuse of a drug. Studies of the new use of a drug need to begin with comparing its use with nonuse, or no baseline data on effect exists.
- None of the RCTs is blind.
- No RCTs control for the risk status of the woman.
- No RCTs control for whether the woman is in active labor and if so, what stage of labor.
- No RCTs have adequate standard dosage regimes. Defining dose according to “adequate contraction pattern“ is most inadequate.
- Since both the dosages used and the drugs compared with vary from study to study, the studies cannot be compared with each other, nor can they be used for meta-analysis.
- All RCTs have been done at university hospitals. While this may provide preliminary data on efficacy, it gives no information on effectiveness—that is, how effective is this use of this drug in the real world of community practice.
- The total sample size (both arms of the trial) in all studies is far too small—between 126 and 220. This sample size is completely inadequate for measuring risks. Since the studies cannot be combined because they lack any standardization of methodology, adding up the samples to increase the sample size is not an option.
- In no trials is woman or baby followed for any significant period of time to identify side effects or risks.
- No trials report on certain, possibly rarer, risks, such as cervical laceration and severe perineal tears.
- Of the six RCTs, five show significantly more fetal tachycardia in the misoprostol arm, and the sixth RCT shows more fetal tachycardia, which does not reach statistical significance. This sixth study also reports more abnormal fetal heart rate patterns and more meconium in the misoprostol arm. These results are preliminary given the small sample size, but they are most worrisome. When these results on risks in the RCTs are combined with the case reports Enoch reviews concerning uterine rupture after misoprostol induction, these very preliminary findings regarding risks of misoprostol induction cry out for further research.
In summary, the studies to date might be a bit helpful in fine tuning dose and dose interval and in suggesting possible efficacy, but they leave wide open serious concerns about risks. Turning again to the Internet and the gold standard of perinatal science, the Cochrane Library, a review of misoprostol induction in the second online issue in 1998 concludes:
“In dosages of 25 micrograms three hourly or more, misoprostol is more effective than conventional methods of cervical ripening and labour induction. The increase in uterine hyperstimulation with fetal heart rate changes found in this review is a matter of concern. Although no differences in perinatal outcome were shown, the studies were not sufficiently large to exclude the possibility of uncommon serious adverse effects. The increase in meconium-stained liquor in one study also requires further investigation.
”Thus, though misoprostol shows promise as a highly effective, inexpensive and convenient agent for labour induction, it cannot be recommended for routine use at this stage. It is also not registered for such use in many countries.
“Because of the enormous economic and possible clinical advantages of misoprostol, there is the need for trials to establish its safety.“ (3)
In other words, the opinion of the best perinatal scientists is that misoprostol induction is still experimental and should be done only in a controlled research setting with the usual protection of research subjects, including fully informed consent. This is because to date our scientific data are inadequate to tell us whether or not misoprostol induction is safe.
The alacrity with which a technology or drug is adopted and used is related to its advantages for the patient and, equally or more importantly, for the practitioner. We have struggled for years with little success to keep a lid on the medically unnecessary use of that most convenient obstetrical procedure—cesarean section. During this same time we have struggled with little success to promote the adoption of the evidence-based but inconvenient VBAC (vaginal birth after cesarean).
How do we hold back the rapid spread of misoprostol induction, which heralds the return of all the conveniences of daylight obstetrics? That the drug is not approved by the FDA for this purpose, not approved for this use by the drug manufacturer, not endorsed for this use by the American College of Obstetricians and Gynecologists or midwifery organizations, and not recommended for routine use by scientists (who tell us we do not know if it is safe) has had no apparent effect on the enthusiasm with which clinicians, both doctors and midwives, are starting to use it.
It is particularly disconcerting to learn from the Internet and from chatting with practitioners that some midwives are jumping on the misoprostol induction bandwagon. A homebirth midwife on the Internet talks of her “work in progress” protocol—an oxymoron, because protocols should never be based on brief experience but always on a thorough review of all the best current scientific data. Midwives need to make every effort to achieve evidence-based practice, particularly when using drugs and invasive technologies, and the clear lack of data on serious risks of misoprostol induction should be sufficient to deter all midwives from this procedure, whether in hospital or out of hospital.
The issue here is consumer protection and quality assurance in maternity care. We need a system of rational pharmaceutical management that guarantees adequate evaluation of every use of a drug prior to its use for that purpose as well as drug protocols developed by an officially recognized group of scientists, clinicians (including midwives), policy-makers, and consumers and based on the best scientific evidence. Present consumer protection systems in some countries, for example in Scandinavia, include mandatory prior evaluation and officially endorsed consensus protocols, and there is no evidence that progress in maternity care is held back.
However, in some countries such as the United States and the United Kingdom, the only way consumers of health care have found to protect themselves is through the courts, because this is the only place doctors and hospitals cannot successfully stonewall information and opinion. Too many times clinicians are sued for the wrong reasons because there is not a system to guarantee prior adequate evaluation and evidence-based protocols with some weight to them.
The case reports on uterine rupture after misoprostol induction recommend not using this drug if the woman has a scarred uterus, and Enoch, in her review, wisely echoes this recommendation (1). But after more than a decade of cesarean section rates in the United States above 20 percent, a significant proportion of American women of childbearing age have a scarred uterus, and such a policy would sharply reduce the opportunities for daylight obstetrics. How many uteri will be ruptured before a court case finally applies a needed brake in this practice? I would welcome learning of cases where misoprostol induction was used without fully informed consent and there was subsequent uterine rupture, cervical laceration or other serious complications.
- Enoch, Jennifer. (1999). Misoprostol (Cytotec): a new method of inducing labor. Midwifery Today, No. 49.
- Copies of 17 pages of messages on misoprostol induction printed off the World Wide Web on May 26, 1998, available from this author.
- Hofmeyr, G. J. (1998). Misoprostol administered vaginally for cervical ripening and labour induction with a viable fetus. The Cochrane Library, Issue 2.
- Midwifery Today’s gateway page about cytotec.