Articles

120 thoughts on “A critique of the leading study of American homebirth

  1. Harriet:

    “Commenters seem to be missing the very valid overall points of Amy’s posts and getting bogged down in minor errors and minutiae.”

    Dr. Richard Dawkins has explained this style of “argument.” He was referring to creationism, but it can be extrapolated to other areas

    “Two rival theories, A and B, are set up. Theory A explains loads of facts and is supported by mountains of evidence. Theory B has no supporting evidence, nor is any attempt made to find any. Now a single little fact is discovered, which A allegedly can’t explain. Without even asking whether B can explain it, the default conclusion is fallaciously drawn: B must be correct …”

    “Natural” childbirth advocates have their own variation:

    Two rival approaches, modern obstetrics and the “midwifery model” are set up. Modern obstetrics explains loads of facts, is supported by mountains of evidence, and has dramatically reduced both neonatal and maternal mortality rates. Midwifery has virtually no supporting evidence for treatments and recommendations that are exclusive to it (as opposed to copied from modern obstetrics), nor is any attempt made to find any, and midwifery is not responsible for any discoveries, devices or procedures that have saved any lives. Now an error in the history of modern obstetrics is discovered. Without even asking whether midwives understood the error at the time or whether they discovered the solution, the default conclusion is fallaciously drawn: the midwifery model must be correct, and modern obstetrics is condemned as useless or even harmful.

  2. Zoe237 says:

    “Two rival approaches, modern obstetrics and the “midwifery model” are set up. Modern obstetrics explains loads of facts, is supported by mountains of evidence, and has dramatically reduced both neonatal and maternal mortality rates. Midwifery has virtually no supporting evidence for treatments and recommendations that are exclusive to it (as opposed to copied from modern obstetrics), nor is any attempt made to find any, and midwifery is not responsible for any discoveries, devices or procedures that have saved any lives.”

    and

    “Not only is there an extensive scientific literature to support everything that is done (new papers are produced at the rate of hundreds, if not thousands, each year), ”

    I can’t believe some of you can’t see what a false dichotomy this is. Dr. Tuteur is an extremist, and she claims that ncb advocates set it up , when she is the one asserting that modern ob is perfect and midwifery is useless. (But the other side is guilty of this somewhat as well). Dr. Hall, you made many arguments against homebirth and some for obstetrics in your “Business of being born” critique. You managed to do with without saying that you could support “EVERYTHING” that was done in hospitals. And you managed to do it without saying that midwives are useless and dumb (the VAST majority of whom practice in hospitals in the US and in areas of the world where there are very few doctors).

    Finally, the last ten or so comments from at least 500 from the last few weeks Dr. Tuteur has been here have been addressing her tone. I hardly call that attacking her. And it was only after multiple points were brought up that Dr. Tuteur refused to address. She might convince more people if she’s didn’t equate homebirthers with all women who would ideally like to avoid a c-section, if possible. Or villify midwives. Or defend 98% of modern obstetrics.

  3. apteryx says:

    Dr. Hall-

    If she doesn’t recommend episiotomy, why does she vocally denigrate people who recommend against episiotomy? Look, you say it is offensive that I point out she’s using threats of doom, but in the message she was writing at the very same moment (posted 4:46 PM), she was issuing another threat of doom! To wit, modern obstetrics is not a collection of practices to be judged individually, but a “system”; the standard of care is “hospital birth with all the ‘bells and whistles’” because this – by a straightforward reading, including “all the bells and whistles” – “produces the best and safest outcomes,” and there is “not one scintilla of doubt” about that. Handy, that – we can stop doing all these pesky controlled trials that sometimes so annoyingly find episiotomy, bed rest, etc. etc., actually do more harm than good.

    (Mixed with the threats, we have historical ignorance on a large scale: “In the past 100 years alone, modern obstetrics in the US has lowered the neonatal mortality rate 90% and the maternal mortality rate 99%. To my knowledge, midwifery has not made even a single contribution to this effort.” Well, we have already established that Dr. Amy’s knowledge is limited. This is just the common delusion that non-MD practitioners never learn or improve over time, which is simply fiction.)

    Anyway, what I want to know is not whether “the system” is The Best Possible (excluding furrin countries with lower death rates), but whether each particular “bell and whistle” make people any better off, because if the answer is no, they shouldn’t be insulted for saying they don’t want that bell or whistle. Fine, let’s all agree that you or your baby are more likely to die in a home birth (though the absolute risk is minuscule). I want to know, is there any evidence that refusing an epidural and using non-drug means of pain control, in the hospital, increases the chance that you or your baby will die or get gorked? Is there any evidence whatsoever that refusing an episiotomy increases the chance that you or your baby will die?

    If not, Dr. Amy is being very offensive indeed by chanting putative mortality statistics in response to questions about these procedures. It’s classic FUD. No, your not having an episiotomy will not kill your baby, and your choosing to suffer labor pains will not kill your baby, but if she can plant in your head the germ of the idea that refusing any “doctor’s order” could kill your baby, then you’re not likely to have the guts to say no when you’re lying on the table.

  4. IndianaFran says:

    Harriet:
    I find it curious that you say “I can’t believe how Amy’s words are being misrepresented and twisted”. and you accuse Plonit of “a gross distortion, to the point of being offensive.”

    And at the same time, you don’t seem to recognize that Amy has repeatedly used exactly those “debate” tactics to insult and denigrate researchers Johnson and Daviss. I have shown several specific examples where she has totally misrepresented their statements; where she has imputed a deliberate intent to deceive without justification; and where she has made accusations and assumptions about what data they possessed and when, which have been proven to be without merit. In short, to many readers, it is Amy’s style of “attacking the messenger” that is drowning out any “valid points” that might be present in Amy’s post.

    Harriet also says
    “But I fully support the right of a patient to choose a home birth as long as she is fully and accurately informed.”
    But every time that Amy repeats her claim that the Johnson and Daviss study “actually shows that homebirth has nearly triple the neonatal death rate of hospital birth for comparable risk women” she is not providing accurate information. If you believe that pointing out her intentional conflation of “neonatal mortality” with “neonatal plus intrapartum mortality” is nothing more than “minor errors and minutiae” – then I’m afraid we just don’t have the same definition of what constitutes “science-based.”

  5. IndianaFran says:

    Sorry, it was apteryx not Plonit that Harriet’s comments were directed towards.

  6. IndianaFran says:

    Amy says
    “Critics of modern obstetrics, particularly midwifery critics, have claimed that obstetrics is not evidence-based. However, they’re wrong about that.”

    Well, at least some mainstream medical academics hold that view:

    http://jdc.jefferson.edu/cgi/viewcontent.cgi?article=1001&context=obgynfp

    American College of Obstetricians and Gynecologists
    practice bulletins: An overview
    Suneet P. Chauhan, MD a
    Vincenzo Berghella, MD b
    Maureen Sanderson, PhD c
    Everett F. Magann, MD a
    John C. Morrison, MD d
    a Division of Maternal-Fetal Medicines Aurora Health Care, West Allis, WI
    b Department of Obstetrics & Gynecology, Division of Maternal Fetal Medicine, Thomas Jefferson University, Philadelphia, PA
    c University of Texas School of Public Health at Brownsville, Brownsville, TX
    d Department of Obstetrics & Gynecology, Division of Maternal-Fetal Medicine, University of Mississippi, Jackson, MI
    Presented at the 72nd Annual Meeting of the Central Association of Obstetricians and Gynecologists,
    October 19-22, 2005, Scottsdale, AZ.
    Objective
    The purpose of this study was to review the American College of Obstetricians and Gynecologists practices
    bulletins to quantify the type of recommendations and references and determining whether there are any
    differences between obstetric and gynecologic bulletins.
    Study design
    All practice bulletins published from June 1998 to December 2004 were reviewed. Odds ratios and 95%
    confidence intervals were calculated.
    Results
    The 55 practice bulletins contained 438 recommendations of which 29% are level A, 33% level B, and 38%
    level C. The 55 bulletins cite 3953 references of which 17% are level I, 46% level II, 34% level III, and 3%
    others. Level A recommendations were significantly more likely among the 23 gynecologic than 32
    obstetric bulletins (37% versus 23%, odds ratios 1.95, 95% confidence intervals 1.28, 2.96). The study
    types referenced in obstetric and gynecologic bulletins were similar (P > .05 for comparison of levels I, II,
    and III and meta-analysis references).

    Conclusion
    Only 29% of the American College of Obstetricians and Gynecologists recommendations are level A, based
    on good and consistent scientific evidence.

  7. Zoe237 says:

    Huh, I’d like to read what 29% of evidenced based recommendations those are.

    http://dinosaurmusings.blogspot.com/2009/11/tweaking-tail-of-home-birth-tiger.html

    Here’s another take on the homebirth studies from an MD (who links to SBM). She makes the point that even if the CDC numbers are correct, the difference is one between absolute and relative risk… very small.

    Interestingly enough, she runs the numbers on the CDC website for 2003-2004 and gets the certified nurse midwive (CNM- the nurses with masters degrees in midwifery) in hospital death rate as .37/1000, or 107 deaths out of 292,000 births, and CNM out of hospital births as 6 deaths out 11,853 births. The rate is supressed because the numerator is less than 20, but it’s .5/1000 if you do the math.

    .37/1000 (hospital) vs. .5/1000 (out of hospital).

    And for 2000, Dr. Tuteur claims that the rate for in hospital was .72/1000 for low risk women (or .92), and I found for CNM out of hospital, it was .9/1000. Don’t know if that’s statistically significant even, but it was about 7 out of 8000. .72/1000 vs. .9/1000.

    But, I would never claim that my own playing with the CDC database proves anything or is scientific.

    There will always be women birthing at home for whateve reason, be they Amish, having a bad hospital experience, or making a philosophical decision. They and their babies need to be low-risk, guaranteed well trained midwives, and have good, fast hospital backup in the case of an unexpected emergency. This is how it is in other countries in which the medical associations support homebirths (like CMA and NHS). These regulations are coming, no matter how emotional Dr. Tuteur gets about the prospect.

    And no, I don’t dispute that J&D could have used a better control group.

  8. fitzerald says:

    “Where fitzgerald sees a “personal dislike for unconventional birthing” I see a concern for giving patients accurate risk information so they can make informed decisions.” – Dr. Hall

    It could’ve been, if Dr. Tuteur didn’t make conclusions like “home births have three times the risk of neonatal death as hospital births.” That is by no means an accurate risk assessment. She makes the same mistake in her other articles like the water birth one. It’s one thing to critique studies, and another thing to twist the data to fit one’s own agenda. If the data suggests that the risks are similar, then the risks are similar. One method is not more or less safe than another. And if Dr. Tuteur’s bias has not been clear, it should be now:

    “Two rival approaches, modern obstetrics and the “midwifery model” are set up. Modern obstetrics explains loads of facts, is supported by mountains of evidence, and has dramatically reduced both neonatal and maternal mortality rates. Midwifery has virtually no supporting evidence for treatments and recommendations that are exclusive to it (as opposed to copied from modern obstetrics), nor is any attempt made to find any, and midwifery is not responsible for any discoveries, devices or procedures that have saved any lives.”

    This is supposed to be “science based medicine?” This dichotomy between modern obstetrics and midwifery exists only in Dr. Tuteur’s mind, the the minds of the unscientific people she’s criticizing. Saying that “midwifery has virtually no supporting evidence” is nothing more than a strawman argument created not to discuss the science of obstetrical practices, but to promote this false dichotomy between doctors and midwives. Midwives provide social support to laboring women, while being alert about possible complications that might require obstetrical interventions, at which point a doctor can take over. This system exists in most, if not all European countries, and overall, provides good outcomes for a reasonable cost.

    The studies on different methods of birthing and different interventions are not related to criticism of midwives or doctors, because for most Western countries, midwives and doctors work in the same medical system. This insult against midwifery and promotion of modern obstetrics (which apparently doesn’t include midwives, who knew) is not “science based medicine” but Dr. Tuteur’s own agenda. No wonder that with such a conflict of interest she’s not able to keep the discussion focused on science.

  9. Here’s an interesting new data point:

    According to their own newsletter, the licensed homebirth midwives of Colorado have a truly horrifying perinatal death rate of 8/1000! According to Karen Robinson, CPM, the President of the Colorado Midwifery Association:

    “In looking back over the past couple years of statistics, I see that there were 5 perinatal deaths reported each year for 2006 and 2007. This represents a perinatal death rate of 8 per 1000 for those two years, and that is too high for the low-risk population we serve. The state perinatal mortality rate for all births from 2003 to 2007 was 6.4 per 1000.”

    In other words, the perinatal death rate of LICENSED homebirth midwives in Colorado, caring for low risk patients, exceeded the perinatal death rate of 6.4/1000 for the entire state (all races, all gestational ages, all birth weights, 2003-2007)! Homebirth was the most dangerous form of planned birth by far.

    Ms. Robinson continues:

    “I don’t believe we have a poor perinatal mortality rate, but if solid data shows we do, then I will be at the forefront of the effort to improve our practices and lower the perinatal mortality rate for homebirth in Colorado.”

    If she’s going to be at the forefront, then she had better get out there. The just published statistics for the year 2008 are even worse. Last year, licensed Colorado midwives had a perinatal mortality rate at homebirth of 8.6/1000 (7 deaths among 806 patients). These numbers are nothing short of horrifying.

    I also find it interesting that Ms. Robinson doesn’t “believe” the data. That is the typical response of many homebirth advocates.

    Curiously, these statistics are not mentioned on the website of the Colorado Midwifery Association. The state is aware that perinatal data rate for homebirth is extraordinarily high. The midwives themselves are aware that the perinatal mortality rate for homebirth is extraordinarily high. The only people who haven’t been informed, it seems, are the patients.

    How can a woman in Colorado make an informed decision about homebirth with a licensed midwife if she has no idea that the homebirth death rate exceeds not only that for low risk births, but exceeds the rate that includes high risk births as well?

  10. apteryx says:

    It sure does seem that Colorado midwives have an unacceptably high perinatal death rate, enormously higher than the national average, although when absolute numbers of deaths are very small, just one or two random extra deaths in a year can greatly change the percentages. If it’s not an artifact of chance, it seems to lend credence to the claim somebody made in an earlier thread that if you deduct a few states where there seem to be notable problems, homebirth in the remainder looks about equally safe as hospital birth.

    It also raises a social justice question. Dr. Amy tells us repeatedly that this <1% death rate is "horrifying … horrifying." Well, I find online the claim that in 2004, infant mortality rates were 13.2 per 1000 in Baltimore, and 16.3 per 1000 in Detroit. I'm sure nobody would say that they felt less horrified by the death of a lower-class black woman's baby, but as a society, we act as if that is how we feel. I estimated earlier that shunting all low-risk women who might choose to homebirth (outside Colorado, I guess!) to hospital birth with "all the bells and whistles" would incur millions of dollars of economic costs per healthy baby saved. Suppose we were to spend the same millions on providing better nutrition, healthier environment, prenatal care, drug treatment for those who need it, to a whole lot of poor urban pregnant women. How many babies would be saved?

  11. edgar says:

    “commenters seem to be missing the very valid overall points of Amy’s posts and getting bogged down in minor errors and minutiae. ”

    Huh. This is stunning to me.
    Amy based her premises on data from the J&D study and other data sources. When I point out that her numbers, logic & conclusions are faulty and explain WHY (from an EPIDEMIOLOGICAL) point of view, it is considered ‘minor errors and minutiae. When in fact the validity of her claims rely on these differences.

  12. fitzerald says:

    “How can a woman in Colorado make an informed decision about homebirth with a licensed midwife if she has no idea that the homebirth death rate exceeds not only that for low risk births, but exceeds the rate that includes high risk births as well?”

    An informed decision would require considering whether the death rate for home births for differences that small indicates higher risk of death or random chance. If it can be considered higher risk of death then why was this so? What was the problem in those births?

    “Curiously, these statistics are not mentioned on the website of the Colorado Midwifery Association.”

    That would be wrong, unless the State ACOG or the hospitals in the state also didn’t mention their stats, not just for deaths, but also for medical interventions. I’m not sure if they do. I know the ACOG guidebook for pregnancy doesn’t comment on any stats. I guess they don’t consider it important enough for informed decisions?
    Both ACOG and Midwifery Associations are professional organizations who’d be careful to mention anything that reflects negatively on them, especially if the data requires analysis.

    “I also find it interesting that Ms. Robinson doesn’t “believe” the data. That is the typical response of many homebirth advocates.”

    I think she said that “I don’t believe we have a poor perinatal mortality rate.” I don’t see it as a rejection of the data but questions about how to interpret it. Perhaps it is typical of homebirth advocates not to “believe” data. But again, you’re showing the same attitude, just in the opposite viewpoint.

  13. “An informed decision would require considering whether the death rate for home births for differences that small indicates higher risk of death or random chance. If it can be considered higher risk of death then why was this so? What was the problem in those births?”

    This is not a small difference; this is a huge difference. For any 3 years period, we would expect the number of perinatal deaths to be approximately 3. Instead, it is 17.

    Why is the risk of death higher? Because homebirth midwives are grossly undereducated and grossly underqualified and homebirth increases the risk of neonatal death.

    As to what the problems were in these births, you’d think that the Colorado Midwives Association might want to find out, and since the midwives who submitted the statistics are their members, they could. They don’t seem to have made any effort to do so.

  14. pmoran says:

    fitzerald: “If it can be considered higher risk of death then why was this so? What was the problem in those births?”

    I think you asked this before. Of course, the specific obstetric problems will not be unique to home birth. Deaths (and added distress) will be due to delays — from lack of continuity of care, and inevitable, unpredictable delays in getting to hospital. Ambulance services have their bad nights too.

    There are also many possible reasons why there may be poorer judgment in lone midwives in the middle of the night, in some countries/states including inferior training and experience, and a reluctance to refer patients at the first sign of difficulty if hospitals are not seen to fit in with natural childbirth philosophy. An added factor may be the determination of some mothers to succeed in giving birth at home even in the face of a midwives’ misgivings.

    Good medicine can lie at the mercy of outside factors.
    This is why we can be so sure that optimal home birth midwifery may approach the standards of safety that can be offered by optimal hospital care, but never quite reach it.

  15. Zoe237 says:

    “Curiously, these statistics are not mentioned on the website of the Colorado Midwifery Association. The state is aware that perinatal data rate for homebirth is extraordinarily high.”

    Actually, the newsletter is on their website… they are hardly trying to hide anything.. What I have had difficulty finding is c-section rates and mortality rates for physician practices in my area.
    http://www.coloradomidwives.org/Newsletters/CMA%20Fall%202009%20Newsletter.pdf

    There is also information on the Colorado public health website about certain counties in Colorado that have astronomically high perinatal mortality rates, some well over 9/1000 deaths. These are vey sad. And I can’t figure out why Dr. Tuteur ignores these hundreds of hospital deaths for the five homebirth ones, except that the latter fit in with her midwifery witchhunt.

    http://www.cdphe.state.co.us/ps/mch/mchadmin/mchdatasets2009/Perinatal%20Mortality%20map%20and%20list%202009%20Rel.pdf

    My next question is about the statistical significance of the data. Is 6.4/1000, or 453 deaths in the hospital group out of 70,804, significantly different than 8/1000, or 7/806? (And yes, I realize that 70,804 is everybody, not just low-risk mothers). 8 vs. 6.4, with such a small numerator and denominator for the latter? I seriously doubt it, but I could be wrong. I agree that the Colorado Midwives Association needs to be looking into this.

    http://www.cdphe.state.co.us/hs/mchdata/vs2007/Colorado.pdf

  16. “Is 6.4/1000, or 453 deaths in the hospital group out of 70,804, significantly different than 8/1000, or 7/806?”

    That’s the wrong question. The perinatal mortality rate for low risk birth in Colorado is not 6.4/1000. You’d need to compare the perinatal mortality rate for homebirth with that of low risk white women, a rate that is substantially lower.

    In the homebirth group, we should have seen 1 or 2 perinatal deaths per year. Instead they had 5-7 deaths per year for 3 years in a row. That’s an extraordinarily high number, as Robinson acknowledges.

  17. Plonit says:

    You’d need to compare the perinatal mortality rate for homebirth with that of low risk white women,

    ++++++++

    Do we have reliable information about the homebirth population in Colorado with regard to risk and race?

  18. Ken Hamer says:

    apteryx said “we independent observers”

    How do we know you are “independent?” (Even if that does NOT necessarily mean unbiased or objective.)

    Drs. Tuteur and Hall have stated their case, and put their own, real names (and reputations) on the line. You, on the other hand, have chosen to remain anonymous behind a psuedonym. Is it a fair inference to say you don’t want your name (and/or reputation) attached to your comments?

    If you were truly inpdependent, and I presume unbiased and objective, perhaps you could identify yourself. ‘Cause in the mean time, no one can know if you truly are indpendent, or in fact are somehow related to the study and/or the authors.

    For all I know, you could be one of the authors.

  19. fitzerald says:

    @pmoran,

    The risks of home births that you mentioned are controllable to a large extent, and that was kind of my point in asking the question. I don’t think there’s any research evidence to conclude surely that optimal home birth care cannot reach optimal hospital care safety. The risk of home birth, and any form of birth can vary from woman to woman, depending on several factors. The latest study from Netherlands points out some of the issues to consider:

    http://www.rcog.org.uk/news/bjog-release-new-figures-safety-home-births

    In a science based blog, we should be discussing research methods, research results, the reasons for the results, and what implications that might have. If the Colorado stats suggest that the neonatal deaths for low risk white mothers is higher for home births than hospital births, then we need to know why those deaths took place to understand the risk factors. Data is particularly complicated for America, because of the variability in midwifery training, various laws against midwifery and barriers for training, socioeconomic status, race, ethnicity, and of course, quality of pre-natal care and access to emergency services.

    Discussions about these issues would be complicated enough, even without the unscientific accusations and conclusions from Amy. Even the latest comment, “Because homebirth midwives are grossly undereducated and grossly underqualified and homebirth increases the risk of neonatal death,” is simply unnecessary. It has already been pointed out that qualifications of midwives in America vary, and that the risk of neonatal death in home birth is firstly, not directly comparable, and secondly, the methods Amy uses are not much better than the study she’s criticizing. Even more unnecessary are suggestions about Colorado midwives not disclosing data, when similar data regarding hospital births are also hard to come by. And I’ve already pointed out the strawman argument about midwives vs. obstetricians.

    The moderators of this blog really need to think about the objective in Amy’s articles. Dr. Hall thinks this is a case where the messenger is being attacked for the message. What is the message, first of all? If the author could avoid attacks on midwifery, attacks on study authors, and commenting on theoretical issues like why women choose non-hospital births, then perhaps the message would have been clear.

  20. Anthro says:

    Oh gosh, I had three babies at home. Has it ever occurred to any of you that some people are willing to take the risk involved (although it is indeed small for a well-screened individual–and I’m an individual, not a study group) in order to avoid medicalization of birth? I had one baby in a hospital and it was enough for me. I am a hater of all things woo and not at all religious, but I maintain my right to birth as I please and I deeply resent doctors (especially male doctors) reducing this particular topic to a pile of statistics. It’s a case by case decision and I was fully prepared to accept the consequences of my decision. I can grasp that not every single baby can survive and that this might even be for the best in individual circumstances and that it is for me (the one who has to raise the child) to make that decision.

    Now to really shock you–I had no midwives, certified or otherwise. Of course I don’t go around preaching that others should do this, but it was right for me (who wouldn’t have made it to the hospital anyway) and I stand by it as do the adult children involved.

    I will look at the study because I would like to know WHY the babies died in the cases in the study. What could the hospital have done that a competent midwife could not? If it’s all about extreme technology, it gets into a gray area for me–where simply maintaing life takes precedence over possible quality of life. I expect to be attacked for this, but I just had to have my say on this. I really am not a nutter, but I take this one thing very personally.

Comments are closed.