Articles

257 thoughts on “The Tragic Death Toll of Homebirth

  1. AlexisT says:

    Windriven:

    Look, you are free to choose a midwife for your childbirth, coffee enemas for your pancreatic cancer or colloidal silver to amp up your immune system.

    Please be more careful with your terminology. Certified nurse-midwives (or fully trained direct-entry midwives, as used in other countries), working with proper backup and support, have an excellent safety record. There’s no reason why a woman having a normal pregnancy should not choose one if it suits her.

    As for the dogma that “home birth is safer than hospital birth”: There are no studies showing that home birth has a lower mortality rate. There are studies showing that it is about as safe. (Some home birth advocates will then add the lowered likelihood of certain interventions and the lowered likelihood of CS to equal “safer”.) The catch? They were not done in the US and reflect standards and conditions that do not exist here. There were recent studies out of the Netherlands and Canada that showed good outcomes. However, they involved strict criteria, good transfer arrangements, and midwives with hospital privileges. The transfer rate in the Netherlands study was 30%.

    Mark: It’s not that simple. If I choose one of my local hospitals, my chance of getting a section is 50% higher than if I choose another. Your place of birth influences your likelihood of undergoing surgery, independent of whatever is wrong with you.

  2. beatis says:

    First of all, I am from the Netherlands, where home births are much more the norm than in many other countries. Also, the link between a preference for home births and alt med is much less strong than it seems to be in the USA.

    Our midwives are very well trained and they always work closely together with MD’s. The Dutch health insurance system differs vastly from the USA: in the Netherlands both home births and hospital births are fully funded in the basic insurance package, provided the midwife is a fully qualified professional. A maternity nurse taking care of mother and baby for a miminum of 5 days after the birth is also standard.

    In april 2009 a new study was published on the safety on home births in the Netherlands:
    http://www.bjog.org/details/news/182410/New_figures_from_the_Netherlands_on_the_safety_of_home_births_.html

    I ‘m just a linguist, not a scientist like many on this blog, so please bear with me should my questions be stupid, but could it be that both level of training as well as logistic factors (hospital always close by) have anything to do with the Dutch outcomes, compared to those in the USA?

  3. overshoot:

    “I’ll readily concede that homebirth has liabilities, but it’s not always a matter of realistic choice either.”

    Homebirth with a homebirth midwife is a fringe choice. Less than 1/4% of women opt for planned homebirth with a direct entry midwife. The studies done on those who plan homebirth show that it is almost always chosen for ideological reasons, not because of distance issues or insurance issues.

    The purported “safety” of homebirth rests almost entirely on the notion that in the event of an emergency, you can simply transfer to the hospital. So being far from a hospital would be a reason not choose homebirth, not a reason to favor it.

  4. Mark P:

    “The whole vaguely fascist thing of births needing to be “natural” really bugs me.”

    Ah, yes, the naturalistic fallacy.

    The irony is that “natural” childbirth bears no relationship to childbirth in nature. “Natural” childbirth activists hark back to the past that never existed. I have written about “Birth Fantasyland”:

    In Birth Fantasyland, women were healthy, strong, in meaningful relationships, and had chosen to have a child. There are no unwanted children in Birth Fantasyland, no teenage mothers, no rape or sexual violence. Every pregnancy is desired.

    In Birth Fantasyland, maternal mortality was rare and neonatal mortality only slightly more common. There were no eclamptic seizures, no ruptured ectopics, no retained placentae. The few problems that existed could be prevented or treated with “good nutrition”.

    Since there were only positive experiences in Birth Fantasyland, the midwives and mothers approached birth as a time of joy and had no fear. They viewed labor as safe, enjoyable, and possibly even orgasmic. They planned to be and were empowered by the experience.

    Not surprisingly, Birth Fantasyland bears a striking resemblance to childbirth among homebirth advocates. Every baby is a wanted baby. There are few complications (because all the women who have risk factors are directed elsewhere) and birth is viewed as empowering and as an “achievement”.

    The reality, of course, was quite different. In nature, the average woman became pregnant shortly after menarche and gave birth within the year (generally between ages 16-18), gave birth to 8-10 children over her lifetime, faced a 1 in 13 lifetime risk of childbirth death and had a life expectancy of 35 years.

    We’re supposed to believe that in nature teenagers who had 1 in 13 lifetime chance of dying in childbirth, who could expect to have 8-10 children, who had no control over their fertility approached birth without fear. We’re supposed to believe that societies with a 1% maternal mortality rate and a 7% neonatal mortality rate “used to birth fine.” We’re supposed to believe that these same teenagers who left no record of any kind indicating that they found birth empowering, painless or ecstatic, actually approached birth in the same way as 30 year old Western, white women in the 20th century. The entire premise is absolutely absurd.

  5. beatis:

    “could it be that both level of training as well as logistic factors (hospital always close by) have anything to do with the Dutch outcomes, compared to those in the USA?”

    Almost certainly, and in addition to the differences in training and logistical factors, there are important differences in philosophy.

    The mantra of American homebirth midwives is “trust birth.” The studies of homebirth in The Netherlands make it clear that Dutch midwives don’t follow that philosophy. In fact, the fundamental premise is exactly opposite: birth is inherently dangerous and great care must be taken to prevent, diagnose and manage complications. The Netherlands mandates rigorous education and training of midwives; the eligibility criteria for homebirth are strict; and a dedicated transport system is in place.

  6. overshoot says:

    The purported “safety” of homebirth rests almost entirely on the notion that in the event of an emergency, you can simply transfer to the hospital. So being far from a hospital would be a reason not choose homebirth, not a reason to favor it.

    Bear in mind that I live in the Western USA. Quite a few of the towns I visit are hours away from the nearest hospital, which means that getting there once labor starts is likely to just substitute “backseat birth” for “home birth.” Having a backseat brother and knowing a fair number of “backseat babies” I can understand people making the choice to deliver at home. In an emergency, a helicopter is going to have an easier time finding a house in town (or a ranch with GPS coordinates) than a milepost along US Highway 70 — out of cellular range, no less.

    As for the prospect of birth complications, you’re preaching to the choir. Tomorrow will be the 26th anniversary of my emergency C-section 34-week sons. Their mother worked NICU for most of their childhoods, so I’ve heard plenty of the stories. Given a choice, my grandchildren (if any; I can hope) will be born with an obstetric resident seconds away.

  7. tractrix says:

    I have several friends who have opted for homebirth, and the one thing they all have in common is this idea that “Your body knows what to do” or “Your body is made to do this.” So this is really a religious position (“God made me”), with related magical thinking. They are anti-evolutionists. Even if they’re not religious per se, they are “spiritual” to the point where they think the “universe” takes care of things. Garbled nonsense like this leads many women to risk their lives and their babies lives in this homebirth stupidity.

  8. micheleinmichigan says:

    Thanks for this. Of course the other statistic I’d like to see is poor outcomes. A good friend of sisters’ had a home delivery and the midwife missed the signs of fetal distress. Then once it was clear that the baby wasn’t doing well it took time for the EMT to arrive and get to the hospital. The child is permanently disabled for reason that would have been totally preventable in a hospital. The parents thought they were choosing a “safe” delivery.

    Hospital deliveries are very different than they were 25-50 years ago. Yet many home delivery supporters still tell stories horror stories from the 1950 and 70′s.

  9. midheleinmichigan:

    “Of course the other statistic I’d like to see is poor outcomes.”

    To my knowledge, no one has looked at this. MANA (the Midwives’ Alliance of North America) has a very detailed collection of data on homebirths, and it may include data about adverse outcomes besides death, but they will not release their data to the general public.

  10. october27 says:

    For a website titled science based medicine, I am surprised at the number of anecdotal stories pillorying women who choose homebirth as sanctimommies, selfish, fanatics, idiots, uninformed, or [substitute negative adjective here].

    Since so many deem it to be appropriate to contribute their anecdotes, I will add mine in the hope of providing some balance. I am currently pregnant with my second child and hoping to have a homebirth. With my first, I had natural childbirth in a hospital setting with no complications. My pregnancy was as uncomplicated as a pregnancy can be, and I feel very fortunate. Thus far, this pregnancy shows all signs of being the same, and I hope it cotinues this way. I am under the care of a CNM, who provides prenatal services indistinguishable from those I received from my OB/GYN during my last pregnancy. I will have an ultrasound and genetic testing (using a doctor whose related miscarriage rates I have researched and am quite comfortable with). If at any time it appears during my pregnancy that there will be any danger to me or my child if I give birth at home, I will change my plans and give birth in one of the two hospitals that are within a five minute drive of my house. Similarly, if in the course of my labor any shadow of a complication arises, I will transfer to a hospital with my midwife.

    My decision is based on several factors, none of them religious or “magical.” The first is my medical history with my first birth. The second is the number of unnecessary hospital protocols I had to submit to (for myself and my child) because I gave birth in a hospital setting. The third is simply because, after having reviewed the available data on homebirth with CNMs and my own midwife’s past performance, I have concluded that the risk of additional complications from homebirth is infintesimal and one I can live with, and that all things being roughly equal I would simply prefer to give birth at home in comfort and not leave my toddler for a couple days.

    If you think I am selfish, so be it, but I certainly don’t think I can be viewed as misinformed. Frankly, I”m not sure I understand where the line should be drawn for most people who spend their time clucking over the risks of homebirth with a qualified health care professional.* Amniocentesis has been raised as an example. Why not simply have women give birth in operating rooms as well, so that C-sections can be performed that much faster if warranted? Why not put all women on a government mandated prenatal diet to ensure their fetuses thrive? Why not require all women with newborns to quarantine themselves within their homes for the several weeks following birth to protect their infants’ fragile immune systems? Why not require selective abortions for women carrying multiples to ensure the health of the surviving siblings? Most homebirth advocates (and I admit, there are some zealots out there) are simply concerned with preserving choices for women, so that they can make educated decisions regarding their medical care without the constricutions of a “protocol” that may make little or no sense given their particular situations. Just my two cents.

    * I do agree, however, that direct entry midwives are problematic and don’t seem to have the required training for the task.

  11. Kylara says:

    lizdexia:

    “It’s also nice to have confirmation that there are OB’s who recognize the education and training of CNM’s, and that the relationship between professional, certified nurse-midwifery and obstetrics need not always be adversarial.”

    My OB practice has four doctors, and I think FIVE CNMs now (four when I was pregnant, but I believe they’ve expanded again). It’s widely considered the best practice in town, and it serves primarily a well-educated professional clientele. For routine care (yearly pelvics, etc.), you can choose a doctor or CNM; during pregnancy, you generally rotate through two doctors and two CNMs so that you’re familiar with several people and someone you know will be on call in case of emergency. We have the choice of laboring the whole time with the CNM of your choice, laboring with the hospital nurses and having the OB come at the end, or laboring with the CNM and having the OB come at the end. In case of emergency C-section, the CNM usually stays to hold the mother’s hand during the process.

    The standard of care, both medical and emotional, is extremely high, and they’re able to cater to a variety of types of “birth plans” while ensuring that emergency OB care is always available and that transition from CNM to OB is seamless if it becomes necessary.

    (These are all hospital births; in my state it is illegal for a licensed professional to attend a homebirth. One of the hospitals has a “birth center.”)

    I don’t live in an area known for being on the cutting edge of medicine (or the cutting edge of woo, either) — pretty traditional in medical practice. But almost every Ob/Gyn practice here has at least one CNM on staff, and all the hospitals allow CNM deliveries, doulas, natural births, rooming in, immediate breastfeeding; and cater to most “birth plans” as long as the delivery remains safe.

  12. windriven says:

    @ Alexis T

    I agree that it was inappropriate to conflate midwifery with coffee enemas and I apologize for it. But I stand unconvinced that home birthing, even with a CNM, is the best choice for even low risk mothers. In my limited experience physicians and hospitals have made tremendous efforts to make the birthing experience more comfortable and less ‘clinical’. The last two of my children were birthed in a room that was very much like a suburban bedroom – except that an operating room and NICC unit were right down the hall.

  13. gregladen says:

    How are these stats affected by either patient being rushed to the hospital or a clinic for emergency response and being removed from the “home birth” count, or being rushed to the hosptical or clinic and being saved from death and thus lowering the death rate count?

    The reason I ask is that I am very surprised that the number is so low.

  14. apteryx says:

    Amy Tuteur says:

    “The first rule of citing studies is you have to read them in order to know what they actually show. … That raises the possibility that allergy in the parents is a risk of C-section (strange, I know, but other studies have suggested similar associations)….Second, though the authors don’t mention it in the abstract, among children to non-allergic parents, there was no association between mode of delivery and development of asthma. The first rule of citing studies is you have to read them in order to know what they actually show. The abstract is not enough; that’s what the authors claim they show, but they are required to publish an entire paper in order for everyone else to evaluate whether the data supports the claim. So the most that one could say is that for the children of parents with allergies, mode of delivery might predispose to asthma, but it is equally likely that parents’ allergic status predisposes to needing a C-section.”

    Good job of sneering, but a very poor job of interpreting scientific papers. Roduit et al. DID indicate in the abstract that if neither parent was allergic, the increase in asthma was not statistically significant. However, the significant increase in asthma in C-sectioned children with one or two allergic parents was compared to vaginally birthed children with the same number of allergic parents. Thus, trying to pin the excess asthma on those families’ genetic flaws simply does not work. You also ignore the fact that the risk of allergic sensitization was statistically significant only in children with no allergic parents (being, in that case, more than double), probably because the children of allergic parents had a higher baseline rate. Moreover, this is hardly the same study that shows similar effects; there are dozens. When you deny facts that do not suit your preferences, you raise doubts about your credibility on a broad scale.

    Windriver:

    “Yutz? Now who is resorting to ad hominem attacks? I engaged in no name calling.”

    Yeah, I regret that. When I’m on the receiving end of a lot of ad homs at once, sometimes I let my buttons get pushed. You really got my goat with the following:

    “Look, you are free to choose a midwife for your childbirth, coffee enemas for your pancreatic cancer or colloidal silver to amp up your immune system.”

    Contrary to your repeated claim above, I have argued – so far irrefuted, and I think correctly – that the question of whether homebirth is ever a legitimate choice depends upon value judgements, rather than simply statistical analysis of one set of data, and thus cannot be answered by science alone. By contrast, the question of whether coffee enemas treat pancreatic cancer is one of fact, which science can (if never quite 100%) answer by the accumulation of data – and there is now enough information that anyone who is still relying on coffee enemas either is completely ignorant of recent scholarship, or isn’t interpreting it rationally.

    Now, I ask myself, why do you assume – or pretend for the entertainment of your orthodox buddies to assume – that if someone thinks it may be legitimate to disagree with you on a question of values, they are not just wicked, but also stupid, ignorant, or delusional? The flip side is that if people differ with you on a question of fact – right or wrong – they may be accused of being not just ignoramuses but emotionally disturbed and possessed of ulterior motives. That is not how human character traits are distributed in real life, but it falls nicely into an Us vs. Them worldview where the expression of a single heresy puts one into the category of Bad People, who must be inferior to you not just in one way but in every way. The modern Western term for this is called “projection,” and you guys are great at it.

    Here’s a hint: many of the people who are getting called preggos, nuts, woo-crazies, fascist, “these people,” and so forth actually already agree with you on most things, do most of what you would order them to do and don’t do most of what you would order them not to. If they show up here and see that their one doubt or dissent has caused you spokespeople for conventional medicine to move them into the subhuman category, it’s NOT going to make them embrace your philosophy wholesale. Don’t believe me if you want, but humans as a group are not that easily cowed.

  15. apteryx says:

    Sorry about the doubled partial quote. The interface isn’t working very well this morning.

  16. Harriet Hall says:

    October27,

    You are depending on the probability that your home delivery will be uncomplicated or that a 5 minute trip to the hospital will be a negligible delay. Probably all will go well, but it would scare me because of OB complications I have seen that you probably have not. Ignorance is bliss, and not having experienced OB emergencies may bias your judgment.

    (1) I have seen the most normal delivery turn to impending disaster in a heartbeat.

    (2) 5 minutes transport time does not include the time necessary for the hospital staff to evaluate the situation and get you to the OR. That process is much faster with a patient who has been in labor in the hospital.

  17. october27 says:

    Harriet Hall,

    I appreciate your response. I have discussed emergency situations with my midwife and feel quite confident with my plan of action should any develop. My midwife has explained to me in detail how she handles hospital transfers, including the timing of a homebirth transfer v. the timing of a woman entering surgery from a hospital, and I am comfortable with proceeding with homebirth.

  18. gregladen says:

    We are about to deliver (within a few weeks). Last week, we lived 35 minutes from the hospital. Then, we went there yesterday for an appointment and discovered that we are now 45 minutes away because a major road has been cut from three to two lanes for refurbishing of drainage and sewer systems.

    I’m just sayin’

  19. skepchick says:

    MOI:

    There is a right way to co-sleep?

  20. Dr. Skeptizmo says:

    ADR150- Sorry about the late response but I just got to this posting. I am a DO student and would like to respond to your question. Just like MD students, we have some serious challenges with the encroachment of CAM in our studies. And just like our MD counterparts, we have choices to make about whether or not to accept that as truth. Beyond that the educational tracks are very similar and we end up with the same amount and depth of training as an MD physician. Let me know if you have any other questions about this… Thanks for the interest.

  21. Joe says:

    october27 on 06 Nov 2009 at 12:19 pm “I appreciate your response. I have discussed emergency situations with my midwife and feel quite confident with my plan of action should any develop.” [italics added]

    Do you “appreciate” the meaning of the word emergency? I think you and your midwife do not. I wish you the best. My best wishes for you have no effect on the high likelihood that you will do well, or the small likelihood that you will wish you had obstetric care in a hospital if you suffer the worst. Are you prepared to shrug it off “We took a chance and lost” if your procedure ends in disaster?

    Yes, there are bad results in hospitals; but the response to emergencies are better than yours.

  22. “You are depending on the probability that your home delivery will be uncomplicated or that a 5 minute trip to the hospital will be a negligible delay.”

    I strongly agree with Harriet.

    Basically, choosing a homebirth is a gamble with a baby’s life. Most of the time, everything is fine. However, if there is a life threatening emergency that requires an immediate C-section or an expert neonatal resuscitation with intubation, the baby will die at home. It doesn’t matter where you live; by the time you transfer, it will be too late.

  23. MOI says:

    @ Skepchick – Yes, yes there is a right way to co-sleep! However, since I do not (kids are noisy and they move a lot, it’s bad enough I have to share a bed with my husband and cat), I cannot give you off the top of my head the proper guidelines. I do know that some co-sleepers do not use pillows or blankets when their infants are small and others have their mattress on the floor. Obviously the parents abstain from alcohol and any other drug that could affect their ability to be roused from sleep if there is a problem. I don’t believe it is recommended for “heavy” sleepers or those who thrash about when sleeping. Not everybody “can” or “should” co-sleep.

    It appears that the infant mortality rate for home birth in Canada is the same as hospital births and it appears that the difference between Canada and the US is the education and experience of the mid-wives, the guidelines involved in chosing whether or not one is a low risk for complications and having quick access to a hospital. Has anyone looked at those stats (from Canada) to assess whether a proper comparison was made (ie, low-risk, full-term pregnancies) between the home birth and hospital births?

    I for one would never chose a homebirth. I feel very comfortable in a hospital. I am also chosing one that is a little farther away from my doctor because it’s at a children’s hospital and is the best in the area (if not the state, right behind Madison). If there is a problem, I want the best available to him (or her) and I want it to be readily available.

    Dr. Amy, you have been gracious and prompt in your responses. I hope I can use this information to better inform the many women I encounter on-line regarding this issue. I have already used your articles/blogs as resources to educate other women on the various subjects of pregancy and birth (this article, water birth and infant mortality in the US). LOL, it looks like I’m sucking up but I’m simply grateful.

  24. storkdok says:

    Anecdote:

    Seven years ago I was called stat to the OR of my hospital (community hosp.) for a delivery. I was in attendance when the ambulance arrived. It was a transport for attempted home delivery by a lay midwife. She “thought the baby might be breech but wasn’t sure” and assured the mother she could perform breech deliveries. At home.

    I observed a fetal head entrapment and immediately performed a standard OB maneuver, delivering the baby. Apgars 0,0. Resuscitation was unsuccessful.

  25. MOI says:

    storkdok – No need to bum everyone out! Ugh, what are the stats on breech babies in regards to vaginal deliveries? I suppose it depends on the type of breech.

    Oh, OT but I need you storkdok on CM!!

  26. lkw says:

    apteryx makes an interesting observation. The choice to home birth appears to involve an application of values and not just data. (Dr. Tuteur also rightly points out that the best data possible should be used in making that choice.)

    Here is another such example. Suppose a physician advises an (adult) cancer patient that treatment has a high probability of extending his life for 1 year. The patient then makes the choice not to accept treatment based on his values (for example, because his quality of life will not be acceptable).

    It appears to me that the critical difference here is that we are talking about two patients (mother and infant) and the risk of death at home is far greater for the infant than for the mother. Further, the infant is not the one making the choice. Oh, there it is. The “choice” word. I knew something about the vehemence of this debate seemed familiar.

  27. IndianaFran says:

    I think that some skepticism is warranted about making a statement of causality (“Homebirth with a homebirth midwife dramatically increases the risk of neonatal death”) based on loosely matched observational data involving a self-selected subgroup.

    A history lesson: For several decades, mainstream MDs were strongly recommending the use of menopausal hormone replacement as a preventative measure against cardiovascular disease, based on observational studies that showed such benefits. Indeed, the evidence was believed to be so strong, that many believed an RCT would be unethical by denying the actual drugs to the placebo group. But now we know the rest of the story – not only was the earlier conventional wisdom not confirmed, it was strongly refuted. The earlier observational data was clearly contaminated by confounders, which were not adequately controlled or adjusted for. And the current consensus is to use HRT only in as low a dose for as short a time as needed for relief of menopausal symptoms, and explicitly NOT for prevention of cardiovascular disease.

    Will there ever be a valid RCT comparing home birth to hospital birth? Highly doubtful that enough women could be recruited who would be willing to be randomized to gather data of sufficient power. But the unavailability of a stronger data set is no reason to exaggerate the conclusions that can be drawn from self-selected observational data.

  28. IndianaFran says:

    And gregladen asks an interesting question here:
    “How are these stats affected by either patient being rushed to the hospital or a clinic for emergency response and being removed from the “home birth” count, or being rushed to the hosptical or clinic and being saved from death and thus lowering the death rate count?”

    The answer is, we don’t know.
    We do know that the total number of attempted home births is perhaps 10 – 30% higher than the number actually registered on birth certificates (or maybe more!). How many of these hospital transfers are good outcomes versus bad outcomes? We don’t know, that data can’t be derived from the CDC data. Dr. Tuteur believes the home birth death rate would be higher if these transfers were accurately accounted. That’s possible. But it’s also plausible that home-to-hospital transfers are more common among the clients of the most prudent risk-averse homebirth midwives. So we can’t really say with certainty which way the numbers would move.

    And, it must be noted, the exact same data uncertainty exists for CNM hospital deliveries. A significant percentage of labors which begin under the care of CNMs are also transferred to MDs for eventual delivery. Which attendant is ultimately recorded on the birth certificate can be a matter of hospital policy and local custom. (How do we know that? – because the CDC data includes some cesarean deliveries which are recorded as CNM-attended births). Again, we have no way of knowing for sure whether including these CNM-to-MD transfers would make the CNM numbers look better or worse.

    So both cohorts as reported by birth certificates are known to be incomplete. This should make us especially wary of drawing conclusions from cohort death rates that have rather small absolute differences.

    Even though “triple the risk” sounds more definitive.

  29. Harriet Hall says:

    These statistics may not be definitive, but they are consistent with everything we know about childbirth and consistent with common sense. Emergencies happen unexpectedly even in the most apparently normal labors, and even a short delay can be life-threatening. Even a home birth with an obstetrician in attendance cannot possibly be as responsive to emergencies as a hospital birth adjacent to an OR – unless you set up an operating suite in the home with appropriate personnel standing by.

    The HRT story is frequently misrepresented. There was early evidence that made us think that HRT reduced cardiac risk. BUT I don’t think any doctor was prescribing HRT solely for cardiac risk; it was an additional factor that helped tip the balance towards prescription in women with peri-menopausal symptoms who also wanted to reduce their risk of osteoporosis. We recognized all along that there were risks associated with HRT. The studies that got all the attention did not show any increase in overall mortality with HRT, and it showed a protective effect for some coniditons like colon cancer. A new study shows a mortality benefit and improved quality of life for younger postmenopausal women. http://www.sciencedaily.com/releases/2009/10/091028162632.htm

  30. IndianaFran:

    “based on loosely matched observational data involving a self-selected subgroup.”

    I did not mean to imply that this is the only data on the subject. It is consistent with the results of virtually every study of American homebirth. Although such studies are few, poorly done, and misrepresent their data, all show that homebirth increases the risk of neonatal death and (possibly coincidentally) all show that homebirth with a homebirth midwife has approximately triple the neonatal death rate of hospital birth for low risk women.

    The most widely cited study on American homebirth (Johnson and Daviss, Outcomes of planned home births with certified professional midwives: large prospective study in North America, BMJ 2005;330:1416), which claims to show that homebirth is as safe as hospital birth, actually shows that homebirth has triple the neonatal mortality rate of hospital birth for comparable risk women. The story of that paper is worthy of its own separate post.

  31. october27 says:

    Dr. Amy,

    You keep mentioning that homebirth has triple the neonatal mortality rate of hospital birth for comparable risk women. You are slightly misrepresenting your statistics. As you noted up thread,

    “I did not include the information on the chart, but homebirth with a CNM has double the neonatal mortality rate of hospital birth with a CNM.”

    Not all hombirths are with lay midwives or DEMs. If homebirth with a CNM has double the neonatal mortality rate of hospital birth with a CNM (and I see no reason to doubt your calculations, although I haven’t doublechecked them), that would put the rate of neonatal death at 0.74 per 1,000 live births, significantly less than the rate for DEMs and only slightly higher than the rate for MDs (of course, I recognize that MDs will have a higher risk group of patients).

  32. october27:

    “You keep mentioning that homebirth has triple the neonatal mortality rate of hospital birth for comparable risk women.”

    Using a DEM or a lay midwife is the typical scenario for American homebirth. So when American women choose homebirth, that is usually what they are choosing.

    CNMs have better outcomes for homebirth, but that is probably more a reflection of philosophy than skills. CNMs are very skilled, but they can’t treat life threatening emergencies at home, either. The difference in neonatal mortality rates reflects greater rigor in eligibility requirements for homebirth.

    At first glance, it seems that homebirth is simply about the place of birth. In reality, though, it reflects a philosophical orientation toward birth. The fundamental issue is whether birth is viewed as inherently safe or inherently dangerous.

    American homebirth advocates insist that birth is inherently safe, complications are rare, and generally announce themselves with plenty of time to address them. Everyone else, and that includes midwives in countries besides the US, knows that birth is inherently dangerous, complications are common, and an immediately life threatening situation can arise from one moment to the next. Therefore, the utmost care is taken in excluding anyone who has any risk factor, no matter how apparently minor.

    The end result is that fewer complications occur at homebirth with CNMs in the US or with midwives in The Netherlands, Canada, etc. However, when those fewer life threatening complications occur, the baby still dies.

    So while CNMs may appear to be safer homebirth practitioners, the reality appears to be that they are simply less likely to gamble. Unfortunately, though, gambling is always a part of homebirth. You gamble that a life threatening emergency will not occur. And if you bet wrong, the baby dies, regardless of attendant.

  33. IndianaFran says:

    Harriet:
    Common sense is not the same as science. Obviously, there are risks associated with home birth, and responsiveness to unanticipated emergencies is the greatest one. Yes, there are cases of death associated with home birth that would have been averted by being in the hospital. But you are not looking at the full picture if your common sense tells you that there are zero offsetting risks inherent to hospital based birth. In the real world, some babies are actually harmed by hospital procedures, like the common overuse and abuse of induction and augmentation agents. Some neonatal deaths occur in hospitals that would not have occurred in a less aggressively “managed” environment. To what extent do those risks offset each other? That is still an unanswered question. It is certainly a common belief that hospital birth is the overall safer option, but that has never been “proven” to the standards of “science based medicine”.

    If the issue of emergency responsiveness was an overwhelmingly large issue, then there wouldn’t be studies from Canada and the Netherlands that show that homebirth can be practiced safely.

  34. IndianaFran:

    “In the real world, some babies are actually harmed by hospital procedures, like the common overuse and abuse of induction and augmentation agents.”

    It is axiomatic among homebirth advocates that the hospital increases the risk of neonatal death and those deaths offset the risks posed by homebirth. There’s a big problem with that analysis: there’s no evidence that hospital procedures increase the risk of neonatal death.

    Moreover, as in the case of the data I presented, and in studies like those done by Johnson and Daviss, any purported deaths due to interventions are already included in the hospital group. So if you think the neonatal death rate in the hospital group could be lower (and you’d need to present evidence for it) that would make the homebirth group look even worse by comparison.

    American homebirth advocacy is a mixture of mistruths, half truths and outright lies. American homebirth advocates insist that homebirth is as safe or safer than hospital birth when the reality is tragically different.

  35. IndianaFran says:

    Dr. Tuteur:

    You have made it very clear in many different venues that you disagree with the conclusions of the Johnson and Daviss study.

    My observation is this:
    The Johnson and Daviss study was submitted to peer review, published in a scientific journal with an excellent reputation for scholarship, and the original study has never been retracted.

    Your personal re-analysis of the study has been widely disseminated on the internet, but has never been peer-reviewed or accepted for publication in an appropriate scholarly environment.

    Your analysis and presentation of statistics extracted from the CDC database also does not meet the standards for publication in a scientific journal.

    If it is true that “all the scientific evidence” leads to a clear and definitive answer about relative risk, then that evidence should be submitted through the accepted channels of scientific inquiry.

    Until then, what you have is a hypothesis.

  36. IndianaFran says:

    “American homebirth advocates insist that homebirth is as safe or safer than hospital birth when the reality is tragically different.”

    A more valid restatement would be
    American homebirth advocates insist that homebirth is as safe or safer than hospital birth when the reality is that we do not have enough evidence to answer the question either way, in a definitive scientific manner.

  37. IndianaFran:

    “Your personal re-analysis of the study has been widely disseminated on the internet, but has never been peer-reviewed or accepted for publication in an appropriate scholarly environment”

    On their own website, Johnson and Daviss have acknowledged the validity of my analysis (without mentioning my name) and have actually performed another faulty calculation attempting to address the criticism.

    I plan to devote a separate post to the issue, but for those unfamiliar with it, here’s a brief summary. Johnson and Daviss looked at the intervention rate and neonatal death rate for all homebirths attended by CPMs in 2000. The compared intervention rates for homebirth to low risk hospital births in 2000. Then they compared the neonatal death rate to … a bunch of out of date hospital studies extending back to 1969.

    A comparison of the CPM death rate in 2000 to the death rate for low risk hospital birth in 2000 shows that planned homebirth with a CPM had a neonatal death rate approximately triple that of hospital birth for comparable risk women.

  38. Harriet Hall says:

    IndianaFran said,

    “But you are not looking at the full picture if your common sense tells you that there are zero offsetting risks inherent to hospital based birth.”

    You are putting words in my mouth. I never suggested any such thing.

    “Some neonatal deaths occur in hospitals that would not have occurred in a less aggressively “managed” environment.” Do you have data to support that claim?

  39. IndianaFran:

    “A more valid restatement would be
    American homebirth advocates insist that homebirth is as safe or safer than hospital birth when the reality is that we do not have enough evidence to answer the question either way, in a definitive scientific manner.”

    That’s not true, either, but the key point is that homebirth advocacy websites and publications don’t temper their claims. They insist that “studies show” that homebirth is as safe as hospital birth.

    The new mantra among “natural” childbirth and homebirth advocates is that obstetricians ignore the scientific evidence on this issue and many others. Just like they’re wrong about that claim in regard to homebirth, they’re wrong about that claim in regard to almost every obstetric intervention and practice.

  40. StatlerWaldorf says:

    apteryx, I appreciate your excellent comments. I have taken two quotes that sum up major flaws in Amy Tuteur’s pundit pieces on homebirth here and on her other blogs.

    “… the question of whether homebirth is ever a legitimate choice depends upon value judgements, rather than simply statistical analysis of one set of data, and thus cannot be answered by science alone.” SO TRUE.

    “You have a habit – I’ve gone and read some of your blog – of assuming that those who disagree with you can only be doing so because they are mentally or emotionally inferior in some way.” SO TRUE.

    I have said before on Amy’s blog that even if the questionable statistics did show a 3x higher mortality rate, there would still be families choosing homebirth. What is the solution? Ban homebirth and jail anyone who does it? How is it that people can support abortion, but not a woman’s right to birth where she chooses?

    Instead of spending time criticizing homebirthers as being intellectually or emotionally defective because they don’t make birth decisions based on a small set of statistics, why don’t people spend time trying to understand the complexities of birth choices and come up with ways to make birth better and safer for all women and babies?
    ————————————————————————–
    lkw said, “It appears to me that the critical difference here is that we are talking about two patients (mother and infant) and the risk of death at home is far greater for the infant than for the mother. Further, the infant is not the one making the choice. Oh, there it is. The “choice” word. I knew something about the vehemence of this debate seemed familiar.”

    Well, you can use the term “patients”, but I think “individuals” is more appropriate. What you discuss is exactly the core of the issue. In society, the baby’s rights seem to have more value than the mother’s rights, after x number of gestational weeks that is. Trauma to the mother is not factored in as she is simply seen as the vessel that brings forth the baby. No matter if there is damage to the vessel, the goal is a live baby at the end. A c-section, even if done as defensive medicine rather than in a clearly indicated emergency situation, is seen as acceptable because the outcome was a live baby. There is little consideration for the fact that a c-section affects the mother’s reproductive future. In the US, vaginal births after cesarian (VBAC) is being denied to women all over the country, which means all future births will be done by repeat cesarian. Reading more on the complications of c-sections and repeat c-sections should make it clear to anyone why some women want to do their best to avoid an unnecessary c-section. The c-section rate is too high, despite what Amy Tuteur says (and her own c-section rate never bore out).

    There are many factors influencing a woman’s birth in hospital. There are staffing issues, resource issues, hospital protocols based on legal protection, not to mention the schedule of the ob-gyn and his/her preferred practices. That is a lot of interests influencing a birth that don’t necessarily have anything to do with the parents’ wishes or the health and best interests of mother or baby… Some women have had terrible birth experiences in hospital due to these factors, and therefore turn to homebirth to be left in peace to birth unhindered. Sure, some women choose a homebirth because of their own personal or spiritual beliefs, but there is a large number of women who homebirth because of dissatisfaction with previous hospital birth experiences. Something that needs consideration.

  41. Plonit says:

    Just like they’re wrong about that claim in regard to homebirth, they’re wrong about that claim in regard to almost every obstetric intervention and practice.

    +++++++++++++++

    I will look forward to Dr Tuteur leading some science-based discussions of those obstetric interventions and practices, e.g. routine use of cEFM, valsalva pushing, drinking and eating in labour, routine IV access, mobility in labour, early cord-clamping.

  42. october27 says:

    I really don’t think this is fair or appropriate for this website:

    “American homebirth advocacy is a mixture of mistruths, half truths and outright lies. American homebirth advocates insist that homebirth is as safe or safer than hospital birth when the reality is tragically different.”

    Some homebirth “advocates” may hide or misrepresent data, but you are painting with a very broad brush here. I understand you are passionate about this issue, but when you resort to these types of statements you start sounding like the “all obstetricians are evil / out to cut you open / don’t care about their patients” crowd. There are plenty of midwives who are honest about the risks and benefits of homebirth and want their patients to be fully informed about their options. Obviously on some level they “advocate” homebirth or they wouldn’t attend homebirths. They work closely with other medical professionals, including obstetricians, and are valued members of their medical communities. Suggesting that all homebirth midwives are little better than snake oil salesmen is frankly pretty appalling.

  43. gaiainc says:

    Birth is not a competitve sport.
    Birth is not a competitive sport.
    Birth is not a competitive sport.

    A woman doesn’t get extra special points for having a vaginal delivery without medications or other interventions or gets penalized for having an epidural or a c-section or whatever. Each birth is unique, has its own risks that are known or unknown, and and it’s not a competitive sport.

    And did I mention that birth is not a competitive sport? Seriously. A c-section is not a failure. A vaginal birth is not a win. A win is a healthy mom and baby however that comes about. Labor and delivery can throw some serious curve balls. It also can go smoothly without any problems. Our ability to predict which is going to happen is not great. It’s not terrible, but it’s far from perfect.

    And really, birth is a not a competitive sport. Anyone who tries to make it into that needs a reality check.

  44. “Some homebirth “advocates” may hide or misrepresent data, but you are painting with a very broad brush here.”

    As I intended.

    Homebirth in the US is not really about place of birth. It is a philosophy with specific empirical claims, almost all of which are false. In is akin to “natural” childbirth, a philosophy that bears little or no relationship to childbirth in nature.

    I have written extensively about the empirical claims of both “natural” childbirth and homebirth advocates and I plan to write about them here in the future.

  45. phren0logy says:

    As others have noted, this post seems far less even-handed than most of the others on SBM.

    Where does the claim that most home births happen with DEMs and not CNMs come from? I’m not saying it’s wrong, just unsupported by the presented data.

    Also, the “two or three times higher” thing is taking a page from the other team’s playbook: it makes the difference sound larger than it is, because double a small number is still a small number.

    I am a physician, and I don’t think it’s too controversial to say that medicine in America sometimes uses procedures in a way that has limited benefit compared to the risk/cost (i.e. bypass). The idea that in a specialty as loaded with liability as OB/GYN might have some procedures that are overused (to demonstrate that all possible measures were taken at the slightest indication of risk) seems reasonable.

    Also, the author seems to have an agenda about lumping many different kinds of “natural childbirth advocates” in with the numbers presented. These particular numbers speak to location and attendant of birth, and other extrapolations seems to be a bit of a reach.

    With all of that in mind, I’m really glad to see these issues getting some activity on this blog. The “natural childbirth” movement plays fast and loose with their claims, and need to be held to account. Hopefully that can be done in a constructive way that acknowledges there are improvements to be made in the way babies are delivered in the US.

  46. MOI says:

    I see an emphasis on “American” homebirth. What about the cited Canadian study? Apparently it showed no difference in death rate. Are they doing it right?

    Oh and I have a list of procedures that I would like to see covered in this blog as well: episiotomies, pushing at 10 cm and no sooner, being told when to push…

  47. I have some concerns about both the recent Canadian and Dutch studies. Both are actually bad news for American homebirth midwives because they make the American homebirth statistics look dismal.

    As regards various obstetrics interventions, I have written extensively about them in the past and will recap my observations here in the future.

    The bottom line about homebirth remains the same. When nothing bad happens, everything turns out fine. When there is a life threatening emergency requiring immediate C-section or expert resuscitation with intubation, the baby almost always dies.

  48. Simonw says:

    “… it makes the difference sound larger than it is, because double a small number is still a small number.”

    Where you are dealing with preventable mortality in neonates, small numbers matter.

    The UK NHS was criticized in the US health care debate, but at ~20,000 GBP per quality adjusted leap year, avoiding a single preventable neonatal death of this kind is worth spending of the order of 1 million GBP.

    A million might not sound much spread over the 1200 births needed to prevent a single death in the US DEM. But midwives are recommended to have a maximum of 27.5 births a year here, assuming they are overworked (probably a safe bet) we’d be talking about a million pounds a year to improve training for ~50 midwives if these figures happened here. So you should be spending ~40,000 USD per DEM per year on improving the performance of these midwives by my back of the envelope calculation. I’m guessing that is more than enough training budget to turn them all into the more qualified midwives.

    Note also mortality is usually just a marker in medicine, you can probably assume if (preventable) mortality is higher, other adverse outcomes are also higher.

    If I understand Amy’s point, it is not home-birth as such (which works out well in my part of the UK according to the best evidence available), but that home birth with inadequately trained midwives, and a lack of acceptance of evidence based midwifery (i.e. when not to arrange home births, and how to spot issues as early as possible).

  49. heyunyi says:

    It looks like there are 13 states that are causing the home birth death rates to be skewed. If you’re giving birth in the other 37 states or D.C., the mortality rate is 4 out of 17,238 or .23, so much safer than hospital birth. (I included 2003-2005 data). If you’re giving birth in one of these 13 states, the rate is 2.25! The 13 states with the high rates are Arizona, Idaho, Indiana, Maine, Michigan, Missouri, Nevada, Pennsylvania, Tennessee, Texas, Utah, Vermont, and Wisconsin. In half of these states, home birth midwifery is either illegal or unregulated, so that could be why the numbers are so high. It does appear that the states with the worst stats tend to be the ones where it is illegal or unregulated. It could also be that homebirths are vastly underreported, especially in states where it is illegal for a midwife to attend a birth, many home births could go unreported to the CDC, except for ones where something goes wrong, thus skewing the statistics. Indiana, for example, had 171 homebirths in three years so the one death that they had is of course going to make their numbers look very bad. But the fact that 37 states had over 17,000 homebirths with only 4 deaths clearly shows that the rate you gave of 1.15 does not apply on a national level.

    So maybe instead of being all “OMG HOME BIRTH TRIPLES THE CHANCE OF YOUR BABY DYING!” you should spend your energy finding out why the rate is so bad in 13 states and what we can do to make it as safe in those states as it is in the other 37 where it is apparently safer than hospital birth.

    I am curious if anyone can give me a good reason why I personally should not homebirth. I live in Washington state which has 3,244 homebirths (twice the national average) and not a single fatality among homebirth babies during the first 27 days during 2003-2005. If you were right about homebirth being as risky as it is, there should have been at least two or three deaths in my state, but there’s not. So how can you say that I would be putting my baby at risk by homebirthing?

  50. jane2008 says:

    I agree that going to a hospital improves your odds of a good outcome, but it doesn’t always mean everything will work okay, either.

    Anecdote: All three of my kids have been delivered at internationally known tertiary care centers. Children #1 & #3 were delivered under the supervision of a CNM. With child #2, however, my only option (insurance limitations) was a very well-known OB group. Because child #1 had been a fast delivery & I was concerned that #2 would be even faster, I dutifully called to let them know I was on my way to the hospital once it was clear I was in labor. Unfortuately, no one in the office notified Dr. On Call, who decided to go home. After being stuck in Friday afternoon rush-hour traffic, I finally arrive and am pushing within 5 minutes of walking in. (Triage nurse couldn’t be bothered to check me because I couldn’t possibly be that composed and be in hard labor.) Residents Tweedle dee and Tweedle dum haven’t a clue what to do and are panicking because there’s meconium and baby’s heart rate is decelerating with every contraction (there was a knot in the umbilical cord)–they want to know whether they can wait for Dr. On Call to show up—oops! too late, baby is crowning. Tweedlee dee decides to yank on baby’s head to hurry things along. I yell at him to stop (remember, no pain killers). Baby then flops onto the table (because no one remembers that you should actually catch the baby) and everyone just stands there, staring the child and doing nothing, as if they have never a baby being born before. Ten minutes later, Dr. On Call shows up, walks in, and says, “What happened?”

    And this from an internationally acclaimed hospital.

    Although everything seemed okay at first, this child has turned out to be a special needs child. I’ll never know for sure whether the delivery played a role in that. But it just goes to show that a hospital setting is no guarantee of competence.

  51. TimMills says:

    First, let me point out that I agree with several of the points raised by Amy and others here: “natural” does not automatically mean “better”, anecdotes do not equal evidence, and there is an appalling amount of anti-science woo surrounding many people’s choice to have a home-birth.

    Second, I’d like to point out that, after over 100 comments, the only actual peer-reviewed scientific studies mentioned in this discussion have shown planned homebirths to be as safe as planned hospital births, where a competent midwife is present and the woman is considered low-risk before labour. (Amy suggests that Johnson & Daviss 2005 and other studies that claim to support homebirth as no more dangerous than hospital birth have important failings; I’ll reserve judgment until she has made her case. Perhaps she could link to posts on SkepticalOB where she does this, if she has done it already.)

    In case anyone is interested, here is the Cochrane review of studies comparing home to hospital birth: http://www.cochrane.org/reviews/en/ab000352.html.

    The result: only one study met their criteria for quality (randomized controlled, etc), and it included only 11 women. Thus, no reliable conclusions could be drawn.

    Upshot: by the strictest rules of medical scientific evidence, we cannot say that either homebirth or hospital birth is safer or more dangerous for women who are assessed before labour as being low-risk.

    This review was assessed as up-to-date as of 30 April 2006.

    The next-best paper I’ve come across, by one of the same authors, is Olson 1997. I can’t currently access the full article, but here are the highlights from the abstract:

    “This study examined the safety of planned home birth backed up by a modern hospital system compared with planned hospital birth in the Western world.”
    meta-analysis of 6 studies
    controlled observational (ie, not randomized, but known variables controlled for)
    combined population observed = 24092 women
    “Perinatal mortality was not significantly different in the two groups.” (no maternal deaths)
    APGAR scores and maternal lacerations were worse in the hospital group.
    Several interventions measured were more common in the hospital group.
    Conclusion: “Home birth is an acceptable alternative to hospital confinement for selected pregnant women, and leads to reduced medical interventions.”

    When my wife and I were investigating birth choices here in the UK before our daughter was born, the research we came across seemed pretty clearly to support the safety of homebirth. That is, they seem to establish that homebirth is as safe as hospital birth, where the attendants have good training and are integrated into the health system, and where a well-equipped hospital is accessible in case of complications. If you’re interested, here are two other discussions of homebirth that I’ve seen and participated in on skeptical blogs: Bad Science and Rational Moms.

    Looking forward to seeing where this discussion leads.

  52. heyunyi:

    “I live in Washington state which has 3,244 homebirths (twice the national average) and not a single fatality among homebirth babies during the first 27 days during 2003-2005.”

    That’s not what I find when I look at the database. During 2003-2005, there was 1 neonatal death: a postdates baby who died of asphyxia at home.

    Moreover, that might not be the only death. The database defines the birth attendant as the person who signs the birth certificate. Therefore, any homebirth transfers end up in the MD group. There could be additional deaths that are not reflected in the database. Since the transfer rate for direct entry midwives is 10-20%, that’s a very real possibility.

  53. My elder daughter is staying with us for a few days. She would not be here if she had chosen a home birth or refused a cesarean when her son was born.

    My wife is here but had she chosen a home birth for our second daughter she would not be here.

    My younger daughter will turn 21 in a couple of weeks. She would not have survived a home birth.

    Perhaps you understand why I was so annoyed when our local member of parliament announced that the Katoomba hospital would have 24/7 anesthetists and obstetricians in the birthing unit and the response of the women campaigning for a better service rejected this because it was just a “medicalisation” of childbirth and the MP should have arranged to spend the money on more doulas and home birth support.

    I don’t believe it’s a coincidence that my local area has the second lowest rate of vaccination in the state.

  54. TimMills,

    In June 2006, attempting to address the issue of homebirth safety, NICE (National Institute for Health and Clinical Excellence) reviewed the existing homebirth studies. Their findings:

    Janssen et al, Outcomes of planned home births versus planned hospital births after regulation of midwifery in British Columbia, CMAJ, February 5, 2002; 166 (3):

    A cross-sectional study conducted in British Columbia … compared 862 women who planned a home birth with a matched control group of 571 women with a planned midwifery-led unit birth (MLU) and 743 women with a planned consultant-led unit birth (CLU)… Perinatal mortality involved 3 babies in 860 planned home births, 1 baby in 733 planned CLU births and no babies in 563 planned MLU births. No further details were given regarding the perinatal deaths… Furthermore, there were 5 infants out of 860 in the home birth group who received assisted ventilation for more than 24 hours, but none in the MLU and CLU groups.

    Bastian et al, Perinatal death associated with planned home birth in Australia: population based study, BMJ 1998;317:384-388:

    A cross-sectional study (N=1,502,756) was conducted in Australia between 1985 and 1990. This … included a comparison of planned home birth with data for the whole country, including details of perinatal deaths for home births. … birth weight specific mortality for babies > 2500g showed a higher mortality rate with home birth … The intrapartum-related perinatal mortality rate, excluding perinatal mortality associated with congenital malformation and/or extreme immaturity, was higher for babies born at home … Intrapartum asphyxia was responsible for about half (24 out of 50 deaths) of infants dying after a planned home birth in Australia between 1985 and 1990.

    Chamberlain et al, The report of the 1994 Confidential Enquiry by the National Birthday Trust Fund:

    A UK questionnaire-based cohort study conducted in 1994 compared 5971 women who had a planned home birth and 4634 women who had a planned hospital birth. Women were matched for age, parity and area of residence. A further 1337 women in the planned hospital birth group were not matched, but were included in the analysis. The complications in current pregnancies reported by women, differed between groups [higher risk in the hospital group] … The characteristics of the babies showed a marginal but significant difference between groups [higher risk in the hospital group] Perinatal mortality rates were 10.7 per 1000 livebirths for planned home birth and 15.6 per 1000 livebirths for planned hospital birth.

    Durand, The Safety of Home Birth: The Farm Study, Am J Public Health, 1992;82:450-452:

    A large cross-sectional study conducted in the US compared outcomes for 1707 planned home births from the Farm midwifery service in rural Tennessee (between 1971 and 1989) with 14033 comparators from the 1980 US National Natality/National Fetal Mortality Survey… The 2 groups cannot be compared directly and the study should be regarded as a case-series.

    Johnson and Daviss, Outcomes of planned home births with certified professional midwives: large prospective study in North America, BMJ 2005;330:1416:

    A large case-series was conducted in the US and Canada upon 5418 women expecting to give birth at home in 2000 supported by midwives with a common certification…. There were 18 perinatal deaths in the 5418 planned home births (3.3 perinatal deaths per 1000 planned home births).

    NICE concluded:

    “The uncertain evidence suggests intrapartum-related perinatal mortality (IPPM) for booked home births, regardless of their eventual place of birth, is the same as, or higher than for birth booked in obstetric units.”

    and:

    “When unanticipated obstetric complications arise, either in the mother or baby, during labour at home, the outcome of serious complications is likely to be less favourable than when the same complications arise in an obstetric unit.”

  55. IndianaFran says:

    @harriethall

    **************************************************“But you are not looking at the full picture if your common sense tells you that there are zero offsetting risks inherent to hospital based birth.”

    You are putting words in my mouth. I never suggested any such thing.”
    ***************************************************

    Sorry, I wasn’t trying to put words into your mouth. You made a statement which I read to say something like “common sense tells us that home birth is risky because of the inability to respond to emergencies promptly”. I was just making the point that reaching a science-based answer requires looking at risks and benefits in both environments. I’m sure that you know that. Your earlier response just seemed incomplete, and I was trying to point that out.

    **************************************************“Some neonatal deaths occur in hospitals that would not have occurred in a less aggressively “managed” environment.” Do you have data to support that claim?”
    ***************************************************

    Well, if you are looking for anecdotes like those provided in support of hospital birth, there’s Tatia Oden French and her baby Zorah – a case of both maternal and neonatal death from AFE after induction with misoprostol : http://tatia.org/index.html
    Can we say for a fact that she and her baby would have survived a non-hospital birth? We don’t have a crystal ball, but we can certainly say that the hospital procedures used in that case most likely led to two deaths.

    As for statistical numbers, there’s this:
    An average of 195,000 people in the USA died due to potentially preventable, in-hospital medical errors in each of the years 2000, 2001 and 2002
    http://www.medicalnewstoday.com/articles/11856.php
    Medical Errors Cost US $8.8 Billion, Result In 238,337 Potentially Preventable Deaths, Study Shows
    http://www.sciencedaily.com/releases/2008/04/080408085458.htm

    While maternity patients typically younger and healthier than other hospital patients, given that approximately 1/4 of all hospital admissions are for maternity care, the odds are that at least *some* of the preventable deaths occurred in obstetrical units. You can’t deny that nosocomial infections do exist, and that maternity departments are not immune from them.

    Based on the inclusion categories used in the original post above, there were 4,360 neonatal deaths in hospital between 2003 – 2005. Is is possible to say how many of these were preventable? Not from the data provided. How many could have been prevented with “less aggressive treatment” versus “more aggressive treatment” versus “typical standard of care treatment”? Again, no one can say.
    The same is true for the 44 neonatal deaths that occurred out of hospital, with an attendant listed as “other midwife”. We don’t know which or how many of them may have been prevented by additional medical care, or more competent midwifery.

    My conclusion is to agree with TimMills:
    Upshot: by the strictest rules of medical scientific evidence, we cannot say that either homebirth or hospital birth is safer or more dangerous for women who are assessed before labour as being low-risk.

  56. Harriet Hall says:

    IndianaFran,

    Good lord, no, I’m not looking for anecdotes!!! The plural of anecdote is not evidence. For my comments on “Death by Medicine” see http://www.sciencebasedmedicine.org/?p=136

    You might as well just admit that you don’t have any credible evidence to back up your belief. And you certainly can’t argue that the delay required to transport a home birth gone wrong has no impact on the outcome.

  57. TimMills says:

    Amy, thanks for offering that.

    It is certainly not the case that it’s always safe to have a homebirth. Certain maternal risk factors likely contra-indicate it. Lack of adequately-trained homebirth attendants would, too. As would living somewhere with a health system that does not adequately integrate homebirths into overall health care (ie, if it makes hospital transfer more complicated than necessary).

    When the system, the attendant, and the maternal risk factors all line up (as they often do in the UK, Netherlands, and other places), then home birth is as safe as hospital birth. The obvious corollary of this is that, for studies that find an increased risk in homebirth, perhaps we should look to these other factors (the system, the attendant, maternal risk) to explain it.

    I’d also like to explicitly address this point that is repeatedly raised in these discussions: “When unanticipated obstetric complications arise, either in the mother or baby, during labour at home, the outcome of serious complications is likely to be less favourable than when the same complications arise in an obstetric unit.”

    While true, this is a bit of a red-herring. The overall mortality (or other outcome measure) is the same (given the right attendant, a suitable health care infrastructure, and low maternal risk). In that case, any increased risk for homebirth mothers due to emergency complications must, statistically-speaking, be offset by decreased risk for other mothers. There are plausible-sounding suggestions for what risks would be reduced (suggestions, I know, which the woo-peddlers also like to trot out), generally revolving around increased rates of interventions.

    The evidence I have seen published tends to lead me to believe that the overall outcomes are, in fact, similar for homebirth and hospital birth (given ideal conditions).

    Whenever someone says, “what if…” followed by some frightening scenario, the proper (skeptical) response is to ask for evidence. Is home birth more dangerous overall? Apparently not – unless you introduce further risk factors (incompatible health-care system, incompetent attendants, specific maternal contraindications).

  58. IndianaFran:

    “We don’t have a crystal ball, but we can certainly say that the hospital procedures used in that case most likely led to two deaths.”

    You keep missing, or ignoring, the main point. Even when you include hospital errors, even when you include malpractice, homebirth has nearly triple the rate of neonatal death. So as “bad” as hospital birth may be, homebirth is 3X as bad.

    Hospitals are not perfect; there is plenty of room for improvement, but homebirth is not as safe as hospital birth.

    People who are cognizant of the risks may want to choose homebirth anyway, and they have the right to make that choice. But they are gambling with their baby’s life and if they bet wrong, the baby will die.

  59. TimMills:

    “The overall mortality (or other outcome measure) is the same (given the right attendant, a suitable health care infrastructure, and low maternal risk).”

    No, I’m sorry, but that’s simply untrue. All the scientific evidence published prior to August of this year showed that homebirth has an increased risk of neonatal death. All the available national and state statistics published thusfar indicates that homebirth has an increased risk of neonatal death.

    Recently, a large Dutch study of homebirth was published that purported to show that homebirth is as safe as hospital birth. However, the homebirth and hospital birth groups differed in at least one important factor: race. Since race in The Netherlands is a risk factor just as it is in the US, it is not clear that the comparison was an appropriate one.

    The Canadian homebirth study also had some problems. The authors were quite vague about the circumstances of death in the two groups. They have not provided information about the deaths in the hospital group, but in subsequent communications revealed that the death in the homebirth group was exactly what we would expect. A serious problem occurred in labor, the patient was promptly transferred, but by the time the baby was delivered it was too late.

    As I’ve said repeatedly in this thread and elsewhere, homebirth is a gamble. The parents gamble that a life threatening emergency will not occur during labor. If they guess wrong, the baby dies. It’s just that simple.

  60. IndianaFran says:

    Thank you Tim for more clearly stating the point that I was trying to make in my earlier reply.

  61. lkw says:

    Well-stated, Tim Mills. Many risk factors for infant mortality associated with home birth clearly differ from risk factors for hospital births. The data above demonstrate that one risk factor for home births in the US (that hospital births do not share) is the presence of a DEM (and absence of a CNM or MD). Seems a worthy public health cause to eliminate this particular risk factor in the US.

    Absent this particular risk factor, it appears to be difficult to tease out the relative risk between home birth and hospital birth (in the US), given the current data available, though perhaps they are roughly equal. More data needed.

  62. edgar says:

    Science based medicine, huh?
    I am truly stunned at the interpretations in this post and subsequent responses. It seems to me that a truly scientific inquiry would say “Direct entry midwives have three times the neonatal mortailty rate in 2003-2004 why could that be?”.

    1. What were the rates in other years? What was the trend across time?
    2. Could it be that they are untrained/undertrained? Could it be that DEM includes many types of midwives? Are there different outcomes between CPMs and other midwives?
    3. Is this increased rate a function of the small numbers, in which typically fluctuate wildly?
    4. Could the rate actually be higher and thus even more dangerous because DEM’s refer to docs?
    5. Could the rate be high because the Amish are a large contigent of homebirthers and have a higher rate of congenital issues?
    6. Is using type of attendant an acceptable proxy for place of birth?

    It it stunning to me that medical professionals do not ask these questions (and seek to find them, if they really care), and do not see that this single snapshot in time, cannot possibly give the answers that they are attributing to it. Stunning that they shout “Well, here is PROOF!” Stunning that they know so little about health surveillence.

    And this is a site called SCIENCE BASED MEDICINE?

  63. heyunyi says:

    Amy,

    I didn’t have postdate pregnancies included in my dataset so that’s how I missed that one. It’s not even legal in WA state for a homebirth midwife to attend a woman past 42 weeks (Though apparently it happens according to the data. The midwives I saw for my last pregnancy didn’t allow it though). I think that whether homebirth is safe between 37-41 weeks 6 days and whether it is safe after 42 weeks may be different issues that need to be examined separately, so I’m not really interested in what happens to homebirth postdate pregnancies as it relates to the relative safety of homebirth, though it would be interesting as an entirely different subject.

    Transfer rates would not make up for the discrepancy in our stats. If 20% of the WA homebirthers had transferred, this would have been around 800 transfers. According to your rate of 1.15 per 1,000 this could have only added one more death, if even that. Let’s look the 37 states with good homebirth records (again, this is 37-41 weeks, not sure how much including postdates births would affect this.) With 17,000 homebirths, a 20% transfer rate would have meant starting with around 21,000 with 4,200 transfers. At most, that would add 4 deaths. So 8 deaths out of 21,000 gives us .38, still just as good as in hospital with a CNM.

  64. heyunyi says:

    edgar,

    5. Could the rate be high because the Amish are a large contigent of homebirthers and have a higher rate of congenital issues?

    I think this definitely needs more looking into.

    The states with the worst stats also tend to be ones with Amish populations:

    Ohio: 1/1277 #1 in Amish population
    Pennsylvania: 8/3795 #2 in Amish population
    Indiana:1/201 #3 Amish
    Wisconsin 5/1845 #4
    Michigan 4/1617 #5
    Missouri 3/507 #6
    Tennessee 2/419 #12

    I believe there are only two states that have more than one death and are NOT Amish and that would be Texas and Utah. All the rest of the states that have had more than one death over three years all contain significant Amish populations (though when I ran the data initially, I didn’t include postdate pregnancies so there could possibly be more states with multiple deaths if you look at that.)

  65. cgh says:

    Lady, you’re a tornado of bullshit. Go away and come back when your bullshit dies over somebody else’s house.

  66. Jenrose says:

    The huge problem with using this as the “definitive” homebirth sample is that it is not accurate.

    1. It does not distinguish between direct entry “lay” midwives and certified professional midwives (CPM). Oregon has both, for example. There is a HUGE difference in the training.
    2. Not all states include non-nurse midwives in birth certificate data. In my county in Oregon, I know for a fact that I gave birth with a CPM at home in 2005. Yet there are ZERO “other midwife” births listed in Oregon in 2005, I assume they are all lumped in “other”… Oh, and my baby lived, which falls in the category of anecdote, but is a specific anecdote which actually DOES prove that the data is not accurate.
    3. In states where direct entry midwifery (certified or not) is not legal, there is a disincentive for the birth certificates to be filled out accurately by attendant type.
    4. The numbers of direct entry midwifery births are not a large enough sample to compare accurately.

    I know that in my personal situation, with more certainty that most, that if my daughter had been born in a hospital she would have spent a significant time in the NICU for feeding issues with the confounding issues of iatrogenic infections looming large, and breastfeeding likely would have failed, as it barely succeeded with me birthing at home, and took heroic measures and constant kangaroo care for 6 weeks to get her so she was gaining consistently at the breast. The outcomes for babies with her condition who are given formula are pretty dismal, she has a rare chromosome disorder, she doesn’t tolerate citrates, all formulas have citrates… As it was she was born at home, I was able to be with her 100% of the time, nursing her, pumping for her, feeding with a bottle, expressing milk into her mouth, everything we could do to get more milk into her. Every other child with her condition spent time in NICU, of the 6 or so other kids with her exact deletion, 2 have died. None have language…except her.

    When it comes down to it, people don’t generally choose their birth attendant for the absolute margin of statistical safety. If they did, CNMs would be a lot busier, they have the best stats by those figures.

  67. Jenrose says:

    Oh, and a second data point for Oregon, my sister, in 2009, was planning a homebirth, but started to develop pre-eclampsia, which the CPM detected, and went in for an induction, which turned into a VERY necessary c-section. She had had an ultrasound, and the usual number of prenatal visits, and blood tests, CPMs can order those in Oregon. Her little girl was born at 4 pounds 4 ounces, with a tiny placenta and short cord. Does this make me think homebirth isn’t “safe”? She’d planned one and transferred before things got dangerous, her little girl was tiny but able to go home with Mom and is now a petite but thriving 8 month old.

    I don’t vaccinate my kids, my sister does, but my reasoning isn’t about “evil doctors bad” (I have a lot of medical issues myself and do get care from MDs for that, as do my children).

    My reason is that on my older daughter’s third round of shots, she went lethargic and non-responsive for 12 hours, and lost developmental milestones that it took about 3 months to recover (she was getting up on her knees and rocking before the shot, and did not even try to crawl until she was closer to 10 months old.)

    We’ve had pertussis (despite me being fully vaccinated and that being the thing she was vaccinated the most with) and it was scary but survivable. She’s had a tetanus booster since, and tolerated it well. Now that my younger child isn’t getting viral croup every 6 weeks, I might consider vaccinating her.

    You can’t assume that people who make choices different from yours are “idiots”. My IQ is 145, I was a National Merit Scholar, I’ve spent most of my adult life reading about pregnancy and birth. I’m not uninformed, and I’m not cavalier about my children’s health.

    I’ve seen a lot of doctors in births use completely inaccurate statements to bully women into interventions they did not want or need. One friend was actually screaming, “I do not consent” as they dragged her into a c-section for a baby who would have died anyway of a severe birth defect. And I know that some midwives do not give optimum care, and I *have* lost a baby. My mother has lost a baby. I have friends who have lost babies, and honestly, they weren’t situations the doctors had any control over at all. I understand both what it is to have a severely special needs child (which has nothing to do with how she was born, giving birth at home did not delete her genetic material), and to lose a child, and I STILL feel more comfortable with homebirth.

  68. heyunyi:

    “According to your rate of 1.15 per 1,000 this could have only added one more death, if even that. ”

    Whoa, that’s not correct.

    The death rate for 1.15/1000 is for patients who experienced no complications. The death rate for those who were transferred is higher, possibly several orders of magnitude higher.

    In the Johnson and Daviss BMJ 2005 study, the authors provided information on transfers and cause of death. In their study of all CPM attended homebirths in North America in 2000, they found 8 deaths in 63 emergency transfers for a death rate of 127/1000! At that rate, there could have been one hundred additional deaths.

    We don’t know what the death rate was for transferred homebirth patients, but it is undoubtedly much higher than the death rate for patients who experienced no complications, possibly dramatically higher.

  69. “Could the rate be high because the Amish are a large contigent of homebirthers and have a higher rate of congenital issues?”

    There’s no evidence of that. Indeed, the studies that have looked at cause of death in the homebirth populations revealed that most of deaths were the result of preventable causes that could have been addressed in a hospital.

  70. Jenrose:

    “It does not distinguish between direct entry “lay” midwives and certified professional midwives (CPM).”

    The Johnson and Daviss study was restricted only to CPMs and the neonatal death rate in the homebirth group in 2000 was almost triple the neonatal death rate for low risk hospital birth in 2000.

    The Midwives Alliance of North America (MANA) the national CPM organization is in possession of a very large database of CPM attended homebirths. According to MANA president Geradine Simkins:

    MANA has collected data on planned midwife attended home births since 1993.

    Data collection includes “evaluation of all aspects of midwifery care in terms of safety, optimal maternal, fetal, and family outcomes,and cost effectiveness.

    Data collection “uses a very extensive data form! ~360 questions.”

    MANA estimates approximately 20,000 cases will be in the database by the end of 2008.

    MANA has no intention of releasing the statistics to the public. Only “qualified researchers” will be granted access based on “an application process/review.”

    What is that application process/review?

    “The association [requesting permission to view the data] needs to contact the Director of Research on association letterhead, with the following:

    a. A statement that the decision has been made by the group
    b. A list of participating members
    c. The name of a contact person who has been chosen to manage the account
    d. The name of the association official authorized to sign the contract for the account

    4) The DOR will then send a contract which contains two parts:
    a. The agreement between the association and the Midwives Alliance for the account
    b. A Non-disclosure Agreement which prohibits inappropriate use of the data…”

    By its own admission, MANA has a database of 20,000 CPM attended deliveries evaluated on 360 parameters … but you can’t see it, I can’t see, no one from the general public can see it.

    It does not take a rocket scientist to surmise that the data shows that homebirth is not nearly as safe as hospital birth.

  71. Jenrose:

    “My IQ is 145, I was a National Merit Scholar, I’ve spent most of my adult life reading about pregnancy and birth. I’m not uninformed”

    You are grossly uninformed. These are not issues intuited by “common sense” and they certainly don’t depend upon IQ. They depend solely on understanding immunology, virology, statistics and the scientific method. And sadly, although you don’t intend it, you are jeopardizing the health of your children AND mine.

  72. IndianaFran says:

    “It does not take a rocket scientist to surmise that the data shows that homebirth is not nearly as safe as hospital birth.”

    Well, a rocket scientist probably wouldn’t be interested. But a medical scientist wouldn’t just surmise – he or she would write up a validly designed research proposal and then approach the owners of the data for appropriate access.

    By the way, in the interest of public disclosure, are you also calling for all hospitals to release the details of all their births and deaths to the general public, so that they can make fully informed choices about their options? Since 99% of American women choose to give birth in a hospital, wouldn’t that data be far more useful to them as consumers?

  73. We’re not talking about data identifying individuals or practices. We’re talking about data about the safety of practitioners as a class. C-section rates, neonatal death rates, maternal deaths rates, you name it; they’re all available for doctors. And in many places they are already available for individual hospitals.

    CPMs are currently in the midst of a big campaign for recognition and licensing in all 50 states. It seems that a minimal requirement ought to be to hand over their own safety data. I can’t imagine a valid reason for hiding it, can you?

  74. Tsuken says:

    My irony meter just got a good thump. I tweeted about this, and have just got an @reply from someone who tells me my ignorance is staggering, and that women should be allowed to make their own informed decisions … ummm, yeah .. that’s kind of the point of providing information – at least I thought so, anyway. Apparently “informed decision” actually means “informed of the good things about what we do and the bad things about anything else”. ;)

  75. TimMills says:

    I’m currently going over some of the papers you have mentioned, to see if they truly support your conclusions. (You’ll understand, based on our exchange re: C-sections, that I want to check for myself.)

    In the meantime, I just wanted to voice my agreement that any dataset which comes with such strings attached as the MANA data is effectively useless to researchers. Any publications that do reference it won’t be checkable by peer reviewers, and so nobody (except insiders) will be able to truly know whether those publications meet scientific standards.

    (I think that’s a damning enough criticism, and I don’t think we can speculate further that the data are in fact damning on their own. We all know how easy it is for cranks or media to distort data, and it could be that MANA is simply being over-paranoid about their data being misused in such a way. Unfortunately, we’ll never know.)

  76. TimMills says:

    Having now read the Johnson & Daviss (2005) paper that Amy Tuteur keeps mentioning, I’d like to address her claims about it in detail. Apologies for length, but I want to be thorough.

    She says that “in studies like those done by Johnson and Daviss, any purported deaths due to interventions are already included in the hospital group.” This is a common red-herring from the anti-homebirth side of these debates. The Johnson & Daviss study, like most of the others that find home birth to be safe (specifically, that fail to support the claim that it’s more dangerous than hospital birth), was conducted on an intention-to-treat basis. The “home birth” group includes all women who intended to give birth at home when labour began. Therefore, it includes all of the transfers – the births that (as Amy points out) turned out to be higher risk than anticipated.

    “In the Johnson and Daviss BMJ 2005 study, the authors provided information on transfers and cause of death. In their study of all CPM attended homebirths in North America in 2000, they found 8 deaths in 63 emergency transfers for a death rate of 127/1000! At that rate, there could have been one hundred additional deaths.

    Where do you get these numbers from? Table 2 in the paper reports 185 urgent transfers, and 655 transfers overall in the homebirth group. I was unable to find an explicit reference the number of deaths among either urgent or overall transfers. However, the box near the end of the paper laying out the cause for all 14 intrapartum and postpartum infant deaths in the home birth group mentions transport during or soon after labour for 6 of them. This (very informal) count gives a death rate of 6/655=9.2/1000, but see my next comment:

    “We don’t know what the death rate was for transferred homebirth patients, but it is undoubtedly much higher than the death rate for patients who experienced no complications, possibly dramatically higher.”

    This is true but largely irrelevant. The key question is whether planning a home birth is safe for a woman who appears low-risk before the onset of labour. To cherry-pick those planned home births after the fact where things didn’t go well (those with urgent transfers to hospital during or soon after labour) and say that they carry high risk is to state an obvious but irrelevant fact, in terms of decision-making by parents, medical staff, and policy-makers. What is important is the overall result for the prospectively-identified group (low-risk pregnant women being attended by CPMs).

    “On their own website, Johnson and Daviss have acknowledged the validity of my analysis (without mentioning my name) and have actually performed another faulty calculation attempting to address the criticism.”

    From another blog of hers, I infer that Amy is either referring to this document, where they address questions about their paper (but don’t do any re-calculations), or this conference abstract. In the abstract, they do seem to take on board the criticism (below) that the original article used out-of-date hospital mortality rates. It is not clear that the calculations they perform are faulty. (Based on the abstract, their reanalysis seems to be motivated by a desire to make the home birth and hospital figures more directly comparable – to match for the known confounding factors of race.) What is clear in Amy’s post is a tendency to attribute nefarious motives to those who disagree with her – soemthing that I tend to treat as a red flag in science-based debates. Amy, it is possible that someone can honestly disagree with you, can be wrong, and yet not be “[trying] to scam people”.

    “Johnson and Daviss looked at the intervention rate and neonatal death rate for all homebirths attended by CPMs in 2000. The compared intervention rates for homebirth to low risk hospital births in 2000. Then they compared the neonatal death rate to … a bunch of out of date hospital studies extending back to 1969.”

    This is, I think, the most substantive of Amy’s criticisms. The authors acknowledge point out in their paper that “the main study limitation was the inability to develop a workable design from which to collect a national prospective low risk group of hospital births to compare morbidity and mortality directly. Forms for vital statistics do not reliably collect the information on medical risk factors required to create a retrospective hospital birth group of comparable low risk“. (emphasis added) In other words, this data isn’t perfect, but the authors think it’s better than the alternatives. They seem to have tried to address this problem in this more recent analysis mentioned above (thanks to Amy’s other blog for the link). I look forward to a more detailed analysis from Amy on exactly how this re-analysis is misleading.

    So, the data is imperfect. When is it not? But from looking at just this one paper that Amy has discussed, I find that her analysis is misleading, irrelevant, and/or inaccurate in several places. I think I’ll join IndianaFran in trusting the consensus of peer-reviewed research papers over the conclusion of someone with such clear shortcomings in her ability to objectively present other people’s research.

  77. TimMills,

    The criticism of the Johnson and Daviss study is straightforward. Instead of comparing homebirth in 2000 with low risk hospital birth in 2000, the authors compared it to a bunch of out of date papers extending back to 1969, because that was the only way to make homebirth with a CPM look safe by comparison.

    Johnson and Daviss performed two separate comparisons in their study. First they compared intervention rates for homebirth to intervention rates for low risk hospital birth in 2000:

    “We compared medical intervention rates for the planned home births with data from birth certificates for all 3 360 868 singleton, vertex births at 37 weeks or more gestation in the United States in 2000, as reported by the National Center for Health Statistics [Births: final data for 2000. National vital statistics reports. Martin JA, Hamilton BE, Ventura SJ, Mencaker F, Park MM. Hyattsville, MD: National Center for Health Statistics, 2002;50(5)]”

    When comparing neonatal death rates, they should have used the exact same hospital group. What was the neonatal death rate in that group? The hospital neonatal death rate for white babies at term of 0.9/1000 is not corrected for congenital anomalies, pre-existing medical conditions, pregnancy complications or multiple births.

    Johnson and Daviss reported a neonatal death rate at homebirth with a CPM of 2.7/1000 (uncorrected for congenital anomalies, breech or twins).

    Johnson and Daviss have publicly claimed that the neonatal death rate for low risk birth in 2000 was unavailable when their paper was submitted for publication in 2004. That’s simply untrue. The data was published in 2002, but, more importantly, they had it in their own hands even before publication. It was the exact same dataset they used to calculate intervention rates.

    As an aside, it is worth noting that Johnson and Daviss are not independent researchers. Johnson is the former Director of Research of the Midwives Alliance of North America (MANA) and Daviss, his wife, is a homebirth midwife. The study was commissioned by MANA and was funded by the Foundation for the Advancement of Midwifery, a homebirth advocacy group.

  78. TimMills says:

    Amy,

    While some of my criticisms of your comments stand, you have convinced me that the Johnson & Daviss paper is not reliable. Particularly, I hadn’t picked up on the (now obvious) fact that they used current data for other measures, but ignored that data for mortality.

    Thanks for pointing that out, and for mentioning the authors’ conflicts of interest. (Just so everyone knows, there is a spot in such papers where conflicts of interest can be noted. In theirs, it says “None declared”.)

    For the record, I accept your claim that homebirth as generally practiced in America is dangerous and goes against the best scientific evidence.

    I remain skeptical about whether this claim can be generalized to countries where midwifery in general, and home birth in particular, are part of the mainstream health system, and where practitioners at both home and hospital births have comparable qualifications. The 2009 Dutch study is impressive in terms of raw numbers (321307 planned home births, and 163261 planned hospital births). It shows similar risks of mortality intrapartum, in the first 24 hours, and in the first 7 days, for home and hospital births, after adjusting for race and other known confounding factors.

    Do you have some positive evidence that Dutch or British home births are less safe than hospital births? Or are you operating under the assumption that they are until proven otherwise?

  79. TimMills:

    “Do you have some positive evidence that Dutch or British home births are less safe than hospital births? Or are you operating under the assumption that they are until proven otherwise?”

    The two new Dutch and Canadian studies appear to show that under rigorous guidelines, homebirth can be as safe as hospital birth. The Dutch situation is very different than anywhere else because of the short transport distances and the dedicated transport system. No other country can match those safety features.

    Prior to those studies, there were none that showed homebirth to be as safe as hospital birth and quite a few that showed the opposite. Australian studies have repeatedly shown that homebirth has an increased neonatal death rate (approximately triple hospital birth) and their state and national statistics confirm the study results.

    The bottom line, though, is pretty straightforward. When no complications occur at home, everything is fine. When certain complications occur and there is time and an efficient method of transport, the baby can be saved. When an immediately life threatening situation occurs (for example, the baby requires immediate intubation) the baby will die.

  80. Plonit says:

    You are not allowing for the possibility that the incidence of complications requiring intervention may vary according to place of birth.

  81. Plonit:

    “You are not allowing for the possibility that the incidence of complications requiring intervention may vary according to place of birth.”

    That’s is the wishful thinking that undergirds homebirth advocacy. There’s no evidence for it.

  82. TimMills says:

    “That’s is the wishful thinking that undergirds homebirth advocacy. There’s no evidence for it.”

    It may be a favorite mantra of the woo-birthers, but it’s also a plausible interpretation of the empirical data. Intervention rates are lower for planned home births than for planned hospital births in low-risk women (Janssen et al 2002 – the Canadian study – and others). You can either attribute that to uncontrolled confounding factors, or to the known variable “place of birth”.

    If the latter, it may be because (a) when you’re not right next to the operating theatre, you’re more ready to try less intervention-based methods, (b) attendants with home birth experience are more aware of such alternative methods, or (c) labouring at home better enables the physiological conditions (hormone production, etc) that lead to an intervention-free birth.

    I don’t know of any study that has probed this particular question more deeply, but the evidence suggests it’s worth researching.

  83. lkw says:

    I’d like to join the chorus calling for Dr. Tuteur to publish her analysis of the Johnson & Daviss in a peer-reviewed journal. At this point, any valid criticisms are smothered in Internet noise and have little chance of getting through to those who need the information most.

  84. TimMills:

    “Intervention rates are lower for planned home births than for planned hospital births in low-risk women”

    Among homebirth advocates, an “intervention” is anything that homebirth advocates don’t approve of. For example, medication for pain relief is an “intervention” but acupuncture for pain relief is not. Electronic fetal monitoring is an “intervention,” but auscultation with a fetoscope or doppler is not.

    Moreover, homebirth advocates like to ignore the role of patient choice. Some women want pain relief. The fact that they request is and then get it does not mean that they endured an “intervention.”

    The idea that homebirth could be safer than hospital birth is counter-intuitive. Obviously hospitals have the personnel and equipment to deal with life threatening emergencies and homebirth midwives do not. Therefore, it is imperative for homebirth advocates to claim (falsely) that hospital birth increases the risk of neonatal and maternal death, and those putative deaths offset the deaths caused by lack of emergency care. There’s simply no evidence for that claim.

    Moreover, homebirth to hospital comparisons always include these putative “hospital caused deaths” and homebirth has triple the neonatal mortality rate anyway.

  85. lkw:

    “I’d like to join the chorus calling for Dr. Tuteur to publish her analysis of the Johnson & Daviss in a peer-reviewed journal.”

    This betrays a lack of understanding of how science works. Being published in a peer review scientific journal does not make something true, and a refutation does not have to be published in order to be correct.

    Johnson and Daviss committed the scientific equivalent of a bait and switch. The evidence that they had in their possession told them that homebirth with a CPM in 2000 had nearly triple the neonatal death rate of low risk hospital birth in 2000. They simply withheld that information.

  86. Plonit says:

    The idea that homebirth could be safer than hospital birth is counter-intuitive.

    +++++++++

    We’re talking about science-based medicine here not intuition-based medicine, so it doesn’t much matter whether the idea is counter-intuitive.

    There are studies (e.g. Chamberlain et al 1994) that show less neonatal resuscitation required at home births than in hospital births (matched for risk, parity, socio-economic status). In other words, complications that prompt neonatal resuscitation were less frequent in home births. It may be counter-intuitive, but that is sometimes where the science leads you.

  87. lkw says:

    Dr. Tuteur: “a refutation does not have to be published in order to be correct”

    You missed my point. It should be published to be *heard* above the din of the Internet. Without peer review, from a research point of view, it unfortunately stands on equal footing with the woo.

  88. edgar says:

    I am not seeinf where they used data from 2002 in the interventions It states quite clearly in the intervention table that they are usung 2000 data from NCHS. There is no mention of 2002 data in the references, either. Please clarify.

  89. edgar says:

    Amy,you could very well submit a refutation…At a bare minimum, an editorial.

  90. lkw says:

    Dr. Tuteur: “This betrays a lack of understanding of how science works.”

    I find myself rather offended by this statement and see no reason for it. Is there an editorial policy on SBM regarding this type of thing?

  91. apteryx says:

    I think SBM’s editorial policy on “this type of thing” is “the more the merrier.”

    The reason electronic fetal monitoring is an “intervention” and occasional use of a stethoscope is not is that the former does not require the woman to stay motionless on her back to avoid dislodging the leads, which means she cannot walk to relieve pain, thus perhaps needing more medications, and her labor may be slower.

  92. apteryx says:

    argh – of course, I means “the latter does not require…”

  93. Plonit says:

    The reason electronic fetal monitoring is an “intervention” and occasional use of a stethoscope is not is that the former does not require the woman to stay motionless on her back to avoid dislodging the leads, which means she cannot walk to relieve pain, thus perhaps needing more medications, and her labor may be slower.

    +++++++++

    Actually, it is better to understand both intermittent auscultation and continuous monitoring as interventions, with potentially different harms and benefits that can be tested against each other. While it’s true that CTG interferes more with a woman’s labour, in terms of position – it is not clear to what extent the negative outcomes associated with CTG are mediated by factors such as position (which could be overcome with telemetry or greater efforts on the part of caregivers to achieve mobility with CTG monitoring) and to what extent they are the result of the low specificity of the technology.

    Both forms of monitoring could, of course, also both be tested against genuine non-intervention (i.e. “don’t monitor”) provided you could get an ethics committee to approve such a study and women to participate in it – which I doubt.

  94. apteryx:

    “the reason electronic fetal monitoring is an “intervention” and occasional use of a stethoscope is not is that the former does not require the woman to stay motionless on her back to avoid dislodging the leads, which means she cannot walk to relieve pain, thus perhaps needing more medications, and her labor may be slower.”

    But cEFM does not require a woman to stay motionless on her back. That’s just another canard made up by homebirth advocates. Fetal monitoring can be done wirelessly and a woman is entirely free to move.

  95. Plonit:

    “Chamberlain et al 1994″

    That study, like almost all homebirth studies, resorted to comparing homebirth to hospital birth of high, moderate and low risk. Obviously hospital birth of high and moderate risk women is going to have higher neonatal resuscitation rates. The only valid comparison is with low risk hospital birth, and that’s a comparison they neglected to perform.

  96. Plonit says:

    Not entirely a canard, since movement can result in loss of contact (if using external transducer) and therefore may be discouraged by a caregiver whose priority is consistency of the trace over the woman’s mobility. Personally, I’m happy to move with the woman, respositioning the transducer as necessary to ensure a consistent trace. Of course, that does require 1-2-1 care – something that is apparently too expensive.

  97. lkw:

    “It should be published to be *heard* above the din of the Internet.”

    What difference does it make whether it is heard above the din on the internet? The point is that careful examination of the paper shows that rather than demonstrating the safety of homebirth, it demonstrates that homebirth with a CPM increases the risk of neonatal mortality. The authors deliberately made a deceptive comparison instead of the appropriate comparison and subsequently they lied about it, claiming publicly that the data was not available when it was in their hands all along.

    What bothers me, and this is an ethical issue, not a scientific issue, is that professional homebirth advocates like Johnson and Daviss, Henci Goer and others KNOW that the J&D paper shows homebirth increases the risk of neonatal death; they’re are hiding the truth about homebirth and they are knowingly misleading women.

  98. Plonit says:

    hospital birth of high, moderate and low risk.

    ++++++++++

    That’s not an entirely fair judgment of the study. The study groups were matched as far as possible for obstetric history, as well as age, parity etc…

    The home birth group was all women who chose home birth, and then matched with a woman in the same locality, close in age, socio-economic factors, While it is true that some obstetric risk factors were higher in the hospital group than the home birth group, the hospital group were still lower risk than the total population (e.g. previous low birth weight baby: 2% of the home birth population studied, 3% of the hospital population studied, as compared with 7% in the population overall). Looking at the comparison of the groups, it would be fair to say that the hospital group were much lower risk than the general population (as you would expect from the matching by obstetric history) but on several measures not as low risk as the home birth group.

    That’s the difficulty with a case-control matched pair study. The difficulties of doing an RCT on place of birth are discussed by Chamberlain. These are some of the difficulties that the NPEU study has presumably been designed to resolve.

Comments are closed.