Guest Post: Homebirth Safety and Risk
In today’s guest blog, Skepchick’s resident stats junkie/guru, Jamie Bernstein, breaks down Dr. Amy Tuteur’s analysis of a recent homebirth study.
Trigger Warning for talk regarding neonatal mortality.
The relative risks of homebirths versus hospital births is an extremely contentious subject. Previous research, mostly done in European countries has shown an extremely small or even nonexistent risk of newborn death from homebirths relative to hospital births, however until recently there have been almost no good studies looking at the risks of homebirths for mothers in the U.S.
This week on the Skepchick backchannel, Julia mentioned that a new study was released this month by the Journal of Midwifery & Women’s Health on the safety of homebirths in the US. This study is being touted by Midwifery organizations as confirming the safety of homebirths. Julia also shared with us this post by Dr. Amy Tuteur from the Skeptical OB where she makes the claim that this same study, which supposedly shows how safe homebirths are, actually shows a 450% increase in newborn death rates from homebirths and confirms how unsafe homebirths really are. Like Julia, I had trouble understanding how two different people can look at the same study and yet come to completely opposite conclusions as to the results, so I thought it might be an interesting area to turn my statistical eye.
The main crux of Dr. Tuteur’s argument is the following:
According to Citizens for Midwifery:
The overall death rate from labor through six weeks was 2.06 per 1000 when higher risk women (i.e., those with breech babies or twins, those attempting VBAC, or those with preeclampsia or gestational diabetes) are included in the sample, and 1.61 per 1000 when only low risk women are included. This rate is consistent with some published reports of both hospital and home birth outcomes, but is slightly higher than others.
No, it isn’t “slightly” higher. It is MASSIVELY higher.
According to the CDC Wonder database, the neonatal death rate for low risk white women at term from the years 2004-2009 is 0.38/1000. As Judith Rooks, CNM MPH noted in her review of Oregon homebirths, intrapartum death among low risk babies is essentially non-existent in the hospital, so the neonatal + intrapartum death rate for the hospital is still 0.38
As the chart above demonstrates, the MANA death rate for the same years was 5.5X HIGHER. In other words, the MANA death rate was 450% higher than the hospital death rate.
She also includes this chart comparing the various death rates:
I clicked the link Dr. Tuteur gave where she got the Citizens for Midwifery quote but didn’t see anything with that quote or numbers on the linked page. I did download and read the study which these numbers supposedly came out of, though it’s worth noting that Dr. Tuteur wrote her post prior to the full paper being released.
I’m not really sure exactly where this quote came from and these numbers don’t seem to match anything in the study. My best guess is that the quote came from a press release or results summary. The stats she mentions (2.06 per 1000 risk of death for all births and 1.61 per 1000 risk of death for low-risk births) are not mentioned anywhere in the actual study, though that doesn’t mean they are not accurate.
Page 7 of the study lists the following fetal and neonatal mortality rates for the homebirth sample:
Intrapartum fetal death (after onset of labor but prior to birth): 1.30 per 1000
Early neonatal death (death after birth in first 6 days of life): 0.88 per 1000
Late neonatal death (death between 7 to 27 days after birth): 0.41 per 1000
However, these numbers include all births in the sample, not just to those of low-risk women. The study also listed the following mortality rates for births excluding lethal congenital anomaly-related deaths:
Intrapartum fetal death: 1.30 per 1000
Early neonatal death: 0.41 per 1000
Late neonatal death: 0.35 per 1000
These numbers still include high-risk births such as breech or twins, though. I was unable to find a low-risk birth stat comparable to the one Dr. Tuteur mentions in her post. I was hoping the stats she mentioned would be in the study so I would get some good confidence intervals to work with, but instead I’m just going to have to take Dr. Tuteur at her word and have to interpret the numbers without confidence intervals.
It’s worth mentioning here as well that the study in question did not compare risks in homebirths to that of a comparable hospital birth cohort. Instead, all this study did is look at a sample of almost 17,000 planned homebirths in the US in which they had lots of detailed information and then report on various statistics about those births.
Let’s leave this study for a bit and go back to the numbers Dr. Tuteur cites in her post. According to Dr. Tuteur, 1.6 per 1000 low-risk planned homebirths from the recent study resulted in neonatal death within 6 weeks of birth. Using CDC data, she also determined that the risk of neonatal death for low-risk white women in the US during the same years was 0.4 per 1000 births. She then points out that OMGZ YOU GUYS THAT’S A 5.5X INCREASE IN BABY DEATHS!
First of all, can I first point out that it’s a 4x increase, not 5.5x (1.6/0.4 = 4). Even comparing the homebirth cohort that includes high-risk births (2.1 in 1000) to the low-risk only CDC cohort (which is not a fair comparison for obvious reasons) would only result in a 5.25x increase in mortality. I seriously have no idea how Dr. Tuteur came up with 5.5x or 450% increase in mortality from the numbers that she cited.
Ok ok ok, but even assuming that Dr. Tuteur screwed up some of the basic math here, we’re still looking at a 300% increase in neonatal death rates for low-risk births. That seems pretty huge. However, take a look at those numbers again. One of the reasons they are reported as deaths per 1000 births is because the risks are too tiny to report as percentages. What we’re really looking at here is a risk of death increasing from
0.0004% to 0.0016%. This is a difference of 0.0012 percentage points. 0.04% to 0.16%. This is a difference of 0.12 percentage points. (Ed. note: minor math correction is fixed, thank you all for spotting this!) In other words, we are comparing an extremely small risk to a slightly higher extremely small risk. Plus, the smaller the risk, the bigger the sample size you need to detect any change in the risk. When dealing with percentages this tiny on a sample size of only around 17,000 births, it’s not clear that we can get enough accuracy to the 1.6 per 1000 number to even determine that it’s truly “bigger” than the 0.4 in 1000 number from the CDC.
All this is even assuming that the numbers Dr. Tuteur cites are comparable. The homebirth study looked at births for women who were planning a homebirth (regardless of where they ended up giving birth). All we know about the info Dr. Tuteur got from the CDC website was that it was for white women with low-risk births. This includes hospital births, homebirths, and births in locations other than the home and hospital (though she labeled them as hospital births on the chart she posted). Women who choose to have a homebirth are likely very different from all white women giving birth, so it’s not really a fair comparison. Plus, Dr. Tuteur doesn’t provide enough information to determine that the definitions are similar for things like “low-risk” and “neonatal death.” It is just not clear to me that these numbers are coming from sources that are similar enough that they can be compared to determine relative risk.
Rather than worry about what Dr. Tuteur wrote in her post, let’s talk a bit about the actual research done in the new Journal of Midwifery & Women’s Health study on U.S. homebirths and what they found in terms of upsides and downsides to having a homebirth.
Neonatal mortality rates for low-risk births were very low.
This study found that neonatal mortality rate for infants of parents who planned a homebirth in the first 33 days of life was 0.77 per 1000. This is similar to numbers found in homebirth studies done in Europe and is very low. This could be slightly higher than planned hospital births, but because the numbers are so close we can’t really tell for sure without a study that compares the cohorts while controlling for anything that may make the homebirth population different from the hospital birth population.
Neonatal mortality rates for high-risk births were inconclusive but worrying.
The study specifically mentions that breech births, TOLAC, multiple gestation and maternal pregnancy-induced comorbidities all seemed to have neonatal mortality rates that are in ranges higher than are typically seen in a hospital setting. However, due to the small overall population and the rarity of these conditions, there were not enough births presenting these conditions in the sample to say for sure. It’s also unclear whether the higher mortality rates are caused by the homebirth or by something else regarding the population of women that choose to have a high-risk homebirth. It’s possible, for example, that the type of women who chose to have a homebirth even though their pregnancy is considered high-risk may also be more likely to refuse medical interventions or drugs that could have helped their baby survive. Additionally, many midwifes will not do homebirths for high-risk births and it’s possible that the midwifes who do choose to deliver a high-risk birth in a home setting may be different from the midwifes who refuse to do so. Either of these situations makes this group extremely difficult to compare with mothers who choose a hospital birth. The question of whether the higher death rates are due to homebirths or something about the women and midwives who do high-risk homebirths is unclear at this point, but is still a cause to worry. A future study using a bigger sample and matched comparison group will have to be done to really understand the relative risk between home and hospital settings in high-risk births, but until a better study is done, it’s best to go with the best evidence we have now (however flawed) and assume that having a high-risk birth at home is a bad idea.
Many women who plan homebirths end up giving birth in the hospital.
In the study sample, 10.9% of women transferred at some point during labor to a hospital. Additionally, 1.5% of the women transferred to a hospital after giving birth. Of the newborns born at home, 1.0% of them were transferred to a hospital after birth. If you’re considering a homebirth, it’s important to consider that you could be one of the 1 in 10 women that end up having to transfer to a hospital partway through the labor process. Additionally, women going through their first birth were 3 times more likely to transfer than women on at least their second birth. This is probably because women who had a difficult first birth may be less likely to choose a homebirth for their subsequent births.
Women who planned for homebirth had low rates of oxytocin augmentation, cesareans, epidurals, and other interventions.
There were low rates of pretty much every type of hospital intervention. However, like the other items I mentioned it can’t really be compared with hospital births without a good matched comparison group.
Lots of women paid for homebirths out of pocket.
Just about two thirds of the women in the study who chose to have a homebirth paid for it entirely out-of-pocket. It seems that not all insurance companies cover homebirth and not all midwives accept insurance.
Conclusions based on this one study: The upsides to a homebirth could be lower rates of medical and surgical interventions and it’s probably just more comfortable being at home rather than at a hospital. However, there is a rather large risk that you may have to transfer to a hospital at some point during labor (which seems less comfortable than just going to the hospital in the first place). For low-risk births the neonatal mortality rate of having a homebirth is either negligible or slightly higher than that of a hospital birth, though because we’re talking relative risk it’s still quite low. For high-risk births there is some evidence that neonatal mortality rates could be much higher in a homebirth setting compared to a hospital. Additionally, homebirths may be extremely costly even if you have insurance.
Again, just to be clear, this is not a review of all the literature and is just my reading of this one study. So, how do two different people read the same study and come to opposite conclusions? The “homebirths are so safe” review by the Midwife Alliance of North America (MANA) focused on the low neonatal mortality rates for low-risk homebirths. They did mention that there were factors that could increase risk, but in my opinion did not give enough of a warning about the possible dangers of having a high-risk homebirth. The “homebirths are killing babies” review by Dr. Amy Tuteur was less “focusing on some parts of the study while downplaying others” and more just a sloppy and unscientific attempt at calculating relative risk by using two completely non-comparable data sources in order to scare readers away from homebirths.