A Preventive Approach to Obstetric Care in a Rural Hospital: Association Between Higher Rates of Preventive Labor Induction and Lower Rates of Cesarean Delivery

Second Response: Dr. Klein raises seven issues with our study that he believes should be addressed. I am happy to do so.

1. Could the results be primarily due to differing physician style irrespective of AMOR-IPAT exposure? It is certainly possible that differing physician styles accounted for some, or all, or the differences in outcomes between the two groups. However, it is also possible that differing levels of AMOR-IPAT exposure accounted for these differences. Or it could be both things working together. To get some insight into the possible relative importance of the two possibilities, lets look at the family physicians in the study. All seven were in private practice, worked in the same call group, believed strongly in continuity of care, and actively pursued the ideals of family medicine. All very much wanted their patients to have safe vaginal deliveries, and all came in to do deliveries on days they were not on call. However, the family physicians classified as AMOR-IPAT providers performed more PGE-2 assisted labor induction, and had lower nullip (3.7% vs. 10.9%) and multip (3.0% vs. 6.8%) cesarean delivery rates. Did differing labor management styles account for these differences, or was it the level of AMOR-IPAT exposure?

Prior to my use of AMOR-IPAT, patients of my personal practice required cesarean delivery ~15-20% of the time. I vividly remember the frustration caused by the need to deal regularly with failure to progress and acute fetal intolerance of labor. I did not have, in my early years of practice, any “global labor management approach” that regularly prevented cesarean delivery for acute fetal intolerance of labor in women who smoked, had hypertension or were older than 35 years of age. However, in recent years I have used an approach (AMOR-IPAT) that increases the likelihood that labor begins before the placenta becomes grade 3, before there is confirmed oligohydramnios, and before there is thick meconium at rupture of membranes. All of these adverse events occur with increasing frequency as a function of increasing gestational age during the term period of pregnancy , so the idea of encouraging delivery early in the term period of pregnancy so as to prevent utero-placental insufficiency has at least some face validity. Similarly, in my early years of practice I did not have any “global labor management approach” that could regularly avoid cesarean delivery for failure to progress in my patients with short stature, excessive weight gain and a borderline LGA fetus. However, in recent years I have used an approach (AMOR-IPAT) that increases the likelihood that labor begins before the fetus has grown unnecessarily large and so that it might pass more easily through the maternal pelvis. This often involves the use of preventive labor induction if spontaneous labor has not started in a timely manner (as estimated by the UL-OTD, or the upper limit of the optimal time of delivery). We know that the risk of CPD/ failure to progress increases as function of increasing gestational age during the term period of pregnancy. AMOR-IPAT theory suggests that vaginal delivery will be more likely to occur if delivery is encouraged earlier during the term period, especially in patients at increased risk for CPD/FTP. Our paper provides support for this theory, but certainly does not prove the theory.

It is important to note that a patient persistent approach to preventive labor induction is probably required in order to maximize favorable outcomes associated with AMOR-IPAT. Is this a characteristic that all maternity care providers have? Unfortunately, this is probably not the case. However, other examples of preventive approaches require significant adjunctive requirements. Surgeons must maintain sterile technique, and must have excellent anesthesia, when removing an appendix before it ruptures. Physicians giving immunizations must store their vaccines correctly, and given the injections properly. Providers performing PAP smears must send their samples to competent cytologists, and have adequate colposcopy services available? All methods of preventive care are at some level complex. However, getting back to our study, there is no conclusive evidence that AMOR-IPAT physicians used methods of care that were more likely to promote vaginal delivery than non-AMOR-IPAT providers. We believe that the most likely reason the cesarean delivery rate was lower in the AMOR-IPAT-exposed group is that more women in the exposed group were more likely to enter labor during their optimal time of delivery. 2. Could the study’s findings be the result of some Hawthorne effect? It is highly unlikely that this study was affected by any “Hawthorne” effect, because during the four years that the clinical events of this study occurred there was no mention of, or planning for, this study. No one (patients, providers, nurses) knew at the time that this study was going to be performed. Hawthorne stated that you cannot study something without changing it, but his law does not usually apply to retrospective research situations. 3. The large group of obstetricians was considered a single entity. The large group of obstetricians was indeed considered as a single entity, but primarily for Table 1 where rates of induction, PGE2 usage, and preventive labor induction were estimated. In the analysis of outcomes, the findings within each entire study group (AMOR-IPAT exposed group vs. control group) were compared. Dr. Klein states that differences in continuity rates and epidural analgesia rates in the two main study groups “demonstrate” large differences in practice style, but he cannot prove that this caused the difference in group cesarean delivery rate for the same reason that I cannot prove that the difference was due to AMOR-IPAT exposure. In terms of epidural usage, is there a study that conclusively links epidural analgesia usage with less intimate care? I do not believe so, but even if there was I could argue that it might not apply to this study population. 4. Babies in this study were differentially admitted to the NICU. Babies in this study who were transferred to the NICU went there because they needed to, and not because of practice differences. Dr. Klein does not know anything about this hospital’s routines. I do. Pediatricians were not more likely to be present in the labor and delivery rooms of control group providers or obstetricians during vaginal delivery, although they were always present for cesarean deliveries (and cesarean delivery is linked to a greater need for NICU admission). The suggestion that babies were differentially admitted to the NICU is not supported by my understanding of the clinical events that occurred during this time period at this hospital. 5. The difference in blood loss was not clinically meaningful. Dr. Klein suggests that our report of the difference in blood loss in patients of the two groups was not clinically meaningful. We reported this data because of our historical concern that PGE2-assisted labor induction might increase the likelihood of post-partum uterine atony and post-partum hemorrhage. The data presented seems to argue against such a link. In fact, the reported data suggests that AMOR-IPAT exposure was associated with significantly less blood loss. I think this is important information. However, the information may be more valuable, and its inclusion in the study more understandable, when it is viewed in the context of this response. Had it not been included, some critics might have asked about the safety of AMOR-IPAT with reference to post-partum blood loss. 6. Demographic differences existed between the two study groups. Dr. Klein points out that many demographic differences existed between the two populations of the study. Actually, we (the authors) made these difference clear by presenting the data in the tables for review. Such differences are common in retrospective cohort studies, and are commonly dealt with, as we did in this paper, using logistic regression. Dr. Klein should not be surprised with such differences, and should understand that they were adjusted for in the study’s analysis. If a variable that was present at different levels in the two study groups was not included in the final logistic model, then that variable did not have a significant effect on the association between exposure [AMOR-IPAT status] and outcome [cesarean delivery] (i.e. the difference in levels of that variable was not statistically important). 7. Re-analysis without the high volume obstetricians. Dr. Klein requested that I remove the high volume obstetricians from both study groups and do a re-analysis. Removing the high volume providers leaves 309 women in the AMOR-IPAT group and 225 in the usual care group. Cesarean delivery rates in the nulliparous and multiparous sub-strata are noted above (nullip (3.7% vs. 10.9%) and multip (3.0% vs. 6.8%) cesarean delivery rates). Importantly, the AMOR-IPAT family physicians had larger numbers of nulliparous women in their practices. Taking all this into account, the AMOR-IPAT family physicians had a significantly lower cesarean delivery rate than the usual care family physicians (3.9% vs. 6.9%, aOR 0.37, 95%CI 0.17-0.83, p=0.015). The study effect does not vanish. The study effect persists.

I believe I have addressed Dr. Klein’s issues. However, as this debate goes on, cesarean delivery rates continue to increase in both the US and Canada. Maybe the lower cesarean delivery rates in this study were due to AMOR-IPAT or provider strategies… or some combination of both. In any case, it would seem advisable to study this method of care further. My final comment for this posting is that in all AMOR-IPAT studies spanning 14 years and two different settings I have always found single digit population cesarean delivery rates. Perhaps this is due simply to the increased use of preventive labor induction, or perhaps this is also related to subtle differences in prenatal care, in provider intimacy , or in variable methods of labor induction (e.g., the use of PGE2 rather than PGE1). However, the “whatever” that is represented by AMOR-IPAT has allowed for an increased rate of vaginal delivery and a decreased rate of NICU admission in settings of mixed parity and moderate obstetric risk. It therefore seems that the exploration and investigation of AMOR-IPAT should be encouraged and welcomed rather than vigorously challenged. Is anyone against safely lowering cesarean delivery rates? I hope not. I do not believe that others have been nearly as successful in improving birth outcomes to the degree found in the AMOR-IPAT studies. One thing I have learned in our life and medical career is that change is inevitable. My question to each reader is, therefore: How open are you to change? The answer to this question probably has significant impact on each person’s personal reaction to AMOR-IPAT.

Competing interests:   No new competing interests

Nicholson’s paper evaluating a system of preventative induction referred to as AMOR-IPAT provides an interesting exercise in the critical appraisal of its internal validity, external validity, interpretation of results, and conflict of interest.(1)

The study suffers a critical lack of internal validity. A retrospective cohort study, the designation of the exposed and non-exposed groups was arbitrary and subjective (the “authors’ first hand knowledge of attending physician attitudes concerning AMOR-IPAT” notwithstanding). Artificially divided into two distinct groups for the purposes of comparison, Table 1 clearly shows that in fact they represent a continuum. The exposed group was defined, a priori, as a group of physicians more open and committed to innovative ways (including inducing 50% of their VBACs) to lower their cesarean section rate. They were also defined by their predominant use of prostaglandin for induction, a practice known to be effective at avoiding cesarean section for failed induction, particularly in women with a Bishop’s score <6 (over 50% of women in both groups). The exposed group used less epidural analgesia (p<0.001), the strongest independent predictor of CS found in their study (OR 4.72), consisted of 45% family practitioners versus 25% of the non-exposed group (p<0.001) the next strongest predictor of CS (OR 0.41), and provided somewhat different prenatal care to different populations: more dating U/S (p<0.001); less excess weight gain (p=0.008); and less IUGR/oligohydramnios (p=0.002). These are not marginal differences explained by chance; the groups are fundamentally different in many ways besides AMOR IPAT: a level of confounding that precludes any comment on causality.

The external validity (applicability) of the study to current North American practice is also in serious question. Cesarean section rates of 5.3% and 11.8% are far below the U.S. national average even for the mid- 1990s. Perhaps the entire group was to some extent committed to keeping the rate low? Similarly, the rates of epidural analgesia (6.5% and 15.7%) were far below current national rates around 70%. Even if this paper were internally valid, its applicability to contemporary settings would be very tenuous.

The study results show identical CS rates for spntaneously laboring parturients. Lower CS rates are only observed in induced women and VBACs in the exposed group. The appropriate conclusion might be that the exposed group was more effectively induced and that their care providers had a different threshold for sectioning VBACs, but attributing the difference in CS rate to a global attitude toward care (i.e. AMOR-IPAT) is nebulous and fatuous.

Finally, the apparent conflict of interest, although disclosed, deserves to be highlighted: Forest Pharmaceuticals, the manufacturer of Cervidil, funded this study…

Can a committed group of maternity care providers, using a variety of methods, successfully lower their CS rate? Certainly. Michael Klein has also demonstrated a lower CS rate in such a cohort.(2) It is also well known that maternity care systems driven by family physicians and midwives are generally less invasive than those led by obstetricians.(3) A like- minded team of Family Physicians supported by a similarly minded obstetrician who use prostaglandins extensively for induction when the cervix is unfavorable, avoid epidural analgesia in favor of less invasive techniques, turn or labour their breeches, and maintain a high threshold for sectioning their VBACs is a good combination; the study group should be commended for bringing these factors (and certainly others) together. Simply being committed to lowering the C/S rate itself defines them as a group willing to go the extra mile for a vaginal birth: to wait the extra hour in the second stage, to avoid electronic fetal monitoring if possible; to do the fetal scalp gas instead of just a CS; to take the time to set up a trial of vacuum or forceps before going to CS; to not interrupt the vaginal birth of a multiparous breech at 7 cm with a laparotomy…

The low C/S rates demonstrated in Nicholson’s study is are laudable, but they are not due to AMOR-IPAT, per se. Instead, they are the result of a multifaceted approach that both Drs. Nicholson and Klein will recognize as essential to their common goal of bringing North American C/S rates back out of the stratosphere.

1. Nicholson JM, Yeager DL, Macones G. A Preventive Approach to Obstetric Care in a Rural Hospital: Association Between Higher Rates of Preventive Labor Induction and Lower Rates of Cesarean Delivery. Ann Fam Med 2007;5:310-319

2. Janssen PA, Klein MC, Soolsma JH. Differences in institutional cesarean delivery rates-the role of pain management. J Fam Pract. 2001 Mar;50(3):217-23

3. Enkin M, Keirse MJNC, Neilson J, Crowther C, Duley L, Hodnett E, Hofmeyr J. A Guide to effective care in pregnancy and childbirth, 3rd edition. Oxford University Press, Oxford, 2000. p.21

Competing interests:   None declared

I also argue for caution in the interpretation of Nicholson’s study which seemingly leads to the conclusion that elective induction of labor decreases Cesarean Birth rates. Although Dr. Nicholson later argues that he is not implying causation, the title of the study and the language used throughout the paper clearly does. Prevention means that an event does not occur due to the intervention mentioned. If this article was meant to stimulate thinking about a different approach to reducing Cesarean Birth rates it should have been more accurately called, “An Approach to Obstetric Care in a Rural Hospital: Association Between Higher Rates of Elective Induction of Labor and Lower Rates of Cesarean Delivery.” Then we could rightly have a debate about causation.

In view of the media’s response to the publication of this article, and the widespread adoption of interventions in the past such as continuous electronic fetal monitoring before their benefit and harm had been carefully studied, we need to be scrupulous in our use of language and attribution of causation.

Karen Buhler Assistant Head Quality, Department of Family Practice, BC Women’s Hospital, Vancouver

Competing interests:   None declared

First I would like to say that I am a clinician, not research faculty, so my editorial is not about the particulars of the study itself but rather about the idea of AMOR-IPAT as something I use in my practice on a daily basis.

I have been working at Penn Family Care for a little over a year now and came here specifically so I could do OB, and even more specifically because I believed in the idea of AMOR-IPAT. I use preventive induction in the majority of my own patients (those who have risk factors for either UPI or CPD) and encounter very little, if any, resistance from the patients themselves. These are patients from all different racial and socioeconomic backgrounds. They are aware that this is not “standard of care” and after a thorough explanation, most of them are all for it. I do not push it – rather I offer it – and those patients who prefer not to be induced under this protocol are in no way pressured. My individual c- section rate is considerably lower than the national average and my patients are very satisfied with their care and the outcomes. My outcome of a lower c-section rate is evident even within our own practice which has 2 sites. One site (where I practice) is an attending-only practice where AMOR-IPAT is used extensively. The other site is our resident clinic where relatively few providers induce labor in this manner. Between both practices we have a very similar patient population. The c- section rate between the 2 offices has been markedly different, though, with the attending practice section rate much lower than the resident clinic section rate. I know this is not in any way, shape or form a proper study, but the lower section rate when AMOR-IPAT is used seems to be consistently apparent.

Sometimes I am astonished at the intensity of the resistance to AMOR- IPAT, I suppose because I adopted it so early in my career and believe in it so whole-heartedly. We are family physicians and our passion is to prevent poor health outcomes in all realms of medicine.

How many times have you watched a primip pregnant patient between 36 and 38 weeks with blood pressures of 110/70, then 116/78, then 126/84 in each consecutive week when all of the previous blood pressures have been around or below 110/70? Say she has only trace protein in her urine, very swollen legs but no hand or facial swelling, and no other signs or symptoms of preeclampsia including normal labs. Add to the picture that she has a 40lb weight gain at this point and started out overweight. Each time I send a patient like this home with instructions to follow up in one week I have a moment of anxiety before I send them out the door about whether or not they are going to make it to 40 weeks without developing preeclampsia. This is not a patient whom you would induce by ACOG guidelines because she is not yet preeclamptic and she is only 38 weeks. But with all of the information you have about this patient, you can certainly tell she is trending in that direction and I feel it in my gut each and every time I send her out of my office to just ‘wait and see’. In all other areas of family medicine we try to stop an adverse outcome from happening before it happens. We watch a patients’ cholesterol start to creep up and we start medication before the patient has a heart attack, not after it has already happened. With AMOR-IPAT we induce labor in a safe and controlled setting before the patient develops preeclampsia and possibly even eclampsia or HELLP syndrome. Dr. Nicholson has documented several situations in which we know patients have adverse outcomes and he has applied this information to create a system in which we can do what family physicians do best – preventive medicine.

Sure, more studies have to be done and are currently in process. But this information needs to be out there and family physicians as well as obstetricians need to see that there is another option on the horizon, one that is possibly even better than what is already practiced.

Competing interests:   None declared

Nicholson in his response seems to personalize our differences, describing what he sees as my attempt to “suppress” dissemination of AMOR- IPAT. His “preventive induction” method continues to be both published and discussed, but in my editorial I raised a number of methodological issues in the current publication—none of which has he addressed, preferring, on the one hand to say that his method is only an hypothesis while on the other, claiming that it needs to be taken seriously.

Specifically Nicholson has failed to respond to my stated concerns as follows: 1. I make the case that the AMOR-IPAT physicians are very different in overall style and approach compared with the usual care group, while Nicholson claims that they differ only in the use of AMOR-IPAT. I claim that this overall approach cannot be disentangled from the study intervention. In effect, the actual intervention is this global approach, not just AMOR-IPAT. Some would call AMOR-IPAT an elaborate study or Hawthorne Effect. 2. I objected to the large group of comparison practitioners being considered as one physician entity when they are in fact 6. This artificially raised the continuity at birth for that group to 96.9% when individually it would be in the range of 16-20%. This and the lower use of epidural analgesia by IMOR-IPAT physicians clearly demonstrates the large difference in practice style and intimacy or continuity between the AMOR-IPAT physicians and the comparison group. 3. I pointed out the artificiality of the difference in NICU admissions, when this outcome is specifically affected by practice difference and not newborn condition. 4. I disputed the reporting of blood loss differences as clinically meaningful. 5. I pointed out the many demographic differences in the population cared for by AMOR-IPAT physicians vs. the comparison group. 6. I suggested removing high volume obstetricians from both study groups and doing a re-analysis. I contend that the study effect would then vanish. He has not done this.

It would be best for Nicholson to address these issues before claiming that his study is methodologically sound enough for us to take him seriously. Failing that, this study is at risk of being adopted by practitioners all to eager to add yet another intervention to an already seriously medicalized birth process.

Competing interests:   None declared

When I first read about AMOR-IPAT, I too was skeptical that cesarean deliveries could be prevented.

What I did know was that millions of additional cesarean deliveries have been done in the last few decades without improving the incidence of cerebral palsy or other outcomes. This motivated me to learn more about AMOR-IPAT. I have read the available articles on the topic and I spoke to Dr. Nicholson after attending one of his presentations.

I would like to review the discussion of the method as it appears to me. I have discussed the AMOR-IPAT concept with Family Physicans and Obstetricians. I have read editorials and Listserve comments.

No one has disputed the rising risk to mother and newborn from week 38 to week 41. The cesarean delivery rate goes up and the rate of NICU admission goes up along with several other measures of morbidity.

No one has disputed that prostaglandin/oxytocin induction is a way to effect delivery earlier. So there seems to be agreement that induction as prevention of morbidity at term is plausible.

In my opinion the resistance to considering the method takes the form of two main ideas. One, patients do not want interventions. The second is the idea that use of an intervention done to a patient who is free of disease must be based on good science.

Let me address the first idea. Patients will accept any intervention that works. The medicalization of pregnancy, labor and delivery has already occurred. Quad screens, epidurals, detailed ultrasounds and oxytocin use have exploded the myth of natural childbirth in the USA. How many patients resist having continuous fetal monitoring even though there is no evidence that it improves outcomes in average risk pregnancies?

This focuses attention on the second criticism of AMOR-IPAT. The studies to date suggest that AMOR-IPAT is a successful strategy for reducing the cesarean rate. The study in this issue indicates that for every 2 women who get a “preventative induction” one cesarean is prevented. That kind of effect size is unlikely to go away when the study is repeated by other investigators. The sample size in the current study has the power to detect increased risk to the exposed group and yet the data trends are toward decreased morbidity in this group. However, until the publication of a large multi-center randomized controlled trial, AMOR- IPAT opponents will set these kind of results aside as unconvincing.

We should not adopt an intervention just because it is plausible. But the work of Dr. Nicholson and colleagues pushes AMOR-IPAT past plausible and into the realm of probably beneficial and safe.

Competing interests:   None declared

Historically, Dr. Klein has been quite active in trying to suppress AMOR-IPAT related research, and he is clearly continuing in that mode. The paper published earlier this week claimed an association between AMOR-IPAT exposure and lower cesarean delivery rates, not a causal link. In his editorial, Dr. Klein make the case that we were claiming causation, and then goes on to explain why we did not meet criteria for proving a causal link. However, we were never trying to claim causation, but rather trying to open the question for debate and study: "Might AMOR-IPAT be a way to safely lower cesarean delivery rates?" Dr. Klein must know that small randomized controlled trials will not be funded unless retrospective studies have been published, and large randomized trials will not be funded until there is evidence that small randomized trials show benefit. To not publish articles like the one published earlier this week prevents the consideration of new therapies, and prevents funding of additional studies. Hence, the publication of this study was appropriate and timely. Furthermore, the evidence presented in this paper is not anomalous. Rather it is consistent with multiple studies spanning 14 years of clinical experience with multiple providers, different locations, different socio- economic mixes, and over 1500 patients.

Dr.Hutten--Czapski reflects appropriate concerns about the study, but his clinical example misses the point of the article. Should a 35 year old woman with anemia and cigarette use be delivered on or before 39 weeks 2 days? Our study would suggest that possibly her birth outcomes would be maximized if she were. To me, having practiced AMOR-IPAT for over 14 years, there is nothing strange about the idea of this kind of patient with an early term preventive induction. However, one of the take-away points of the article is that the patient listed above is really not "low- risk." She does have significant risk of utero-placental insufficiency. Granted, more likely than not she and her baby will do well if she does not develop labor until after 40 - 41 weeks of gestation. However, the question at hand is: "if you have 200 women just like her, and 100 were managed so as to ensure that she delivered on or before 39 weeks 2 days gestation, and the other 100 were not induced except for standard ACOG indications - which group would have better outcomes (lower cesarean delivery rates and lower NICU admission rates)." Currently the literature is not able to answer these questions.

Importantly, the study in under discussion uses a group-based analysis rather than a mode of onset-based analysis. Hence, women in both groups will have both spontaneous labor and induced labor - just different proportions of each. In prospective trials, this would also be the case. The proper way to study this question is not to compare 100 women who are induced to 100 women who develop spontaneous labor, because that is not the way our OB populations are. The proper was is to compare women exposed to AMOR-IPAT to those not exposed.

The bottom line is that there are currently no randomized controlled trials of AMOR-IPAT, but several retrospective trials that report major improvements in birth outcomes. I am all in favor of being cautious about the acceptance of AMOR-IPAT based on the currently available evidence, but I also believe that there is now more than enought evidence to take AMOR- IPAT seriously and to study its effect in the context of a adequately powered RCT. Its advocates, like me, need to be careful about overstating its effectiveness. However, its detractors need to be equally careful about overstating the likelihood that it is a method of care without significant potential.

Jim Nicholson

Email: phc@srpc.ca Title/position: Haileybury, Ontario Place of work: MD, Northern Ontario School of Medicine IP address: 24.18.181.32 Hostname: c-24-18-181-32.hsd1.wa.comcast.net Browser: Mozilla/5.0 (Macintosh; U; PPC Mac OS X; en) AppleWebKit/419.3 (KHTML, like Gecko) Safari/419.3

Parent ID: 5/4/310 Citation:

A Preventive Approach to Obstetric Care in a Rural Hospital: Association Between Higher Rates of Preventive Labor Induction and Lower Rates of Cesarean Delivery

James M. Nicholson, David L. Yeager, and George Macones

Ann Fam Med 2007; 5: 310-319 (Original Research)

http://annalsfm.highwire.org/cgi/content/abstract/5/4/310

http://annalsfm.highwire.org/cgi/content/full/5/4/310

----------------------------------------------------------------- "Intriguing but radical departure" -----------------------------------------------------------------

<!-- article ID: 5/4/310 -->

<P> Reducing unnecessary cesarean sections is a laudable goal and the authors raise an interesting hypothesis and provide preliminary data suggesting a possible way.

<P> One point that the authors did not make clear is that the scoring system they used would be a dramatic departure from current practise. My reading is that it would involve significant increases in induction rates (which may already, as Dr Klein mentions in his editorial, have been raised in association with an increase in cesareans)

<P> Consider the not uncommon (and not very well predictive) scenario of commonplace variables, with normal foetuses. By example 35yo smokers with a normal size foetus with nothing more sinister in the prenatal course than first trimester anaemia, get induced in the study group at 41 weeks less 12 days = 39 wk 2d!

<P> This is an intervention, that if it becomes widespread, with or without a decrease in cesarean, will involve a very large number of women at low risk.

<P>Competing interests:   None declared

Competing interests:   See article in question

As I expected would be the case, three of four US national news services have reported that "preventive induction" is a major breakthrough and holds great promise in reducing the epidemic of cesarean sections. Only one pointed out that there are methodological issues as yet unaddressed, and even that one put a positive spin on the new method.

Even as I prepared an editorial at the request of the editors, and did my best to outline what I thought were major issues with so-called "preventive induction," I assumed that the news would be spun this way. This is precisely why I am against premature publication of results before they are properly tested--no matter how intriguing and no matter how much debate and controversy might eventuate--and no matter how much press Annals might get in the process.

The public is all to ready to accept yet another intervention designed to further medicalize birth. We are a risk-aversive society and anything that might hold out hope for reducing risk, even marginally, will be eagerly snapped up. All the more reason to bend over backward to demand a high level of scientific proof before going down that pathway.

Competing interests:   None declared