Home birth, its merits notwithstanding, may be a bigger risk to the baby than some parents would tolerate in other decisions.
With many friends getting pregnant these days, one question I get a lot is some version of
“We’ve heard a lot about home births. What do you think about them?”
Remarkable new-ish research from Oregon helps explain why it’s not something I’d choose for my family.
Relative Risks
We live in an uncertain world. For some, kite surfing is in but free diving is too much risk. Some ride a motorcycle through Nairobi at night with no headlights.(1) Others set their risk tolerance lower.
Most parents I know are, understandably, quite risk averse around the safety of their new baby. Thus, it surprises me how many such parents contemplate home birth. I suspect this means those of us in the baby-delivering business (and the media around us) are doing a poor job communicating these risks.
The Oregon data showed a little more than double the risk of infant death in home births relative to hospital ones. In absolute terms, home birth conferred an extra 2 infant deaths per thousand births. Considering the rate of all deaths before age 5 in the US is just 7 per thousand, that is a big incremental risk.
For some perspective, the risk of SIDS, while every case is tragic, comes in at 0.4 per thousand.
I am a huge believer in individual liberty and autonomy in medical decision making. Clearly, many people value a home birth tremendously, and it can be a safe, viable option in the right setting. I have the utmost respect for them. On average, home birth without careful backup plans, is probably beyond the risk tolerance of most parents I see for their babies.
Footnotes
- Yeah…I did that. But in my “defense”, I hadn’t figured out that my headlight, unlike on my American bike, had an on/off switch. Not boldness so much as stupidity.
Details, details…
The Oregon Study
Starting in 2012, the state Oregon did something totally awesome to make this analysis possible.2 They began to ask parents registering a new baby where they had intended to deliver.
(Huge thanks to Dr Chitra Akileswaran for pointing this article out. Check out her astute obstetric perspective here.)
Before this (and still in the rest of the US), birth records asked where the baby was born but not where it was intended to be born. This created a huge confounding problem. If a couple planned a home birth, but something went wrong & they rushed to the hospital, that counted as a hospital birth instead of the home birth it was intended to be. Every midwife, OB-gyn & pediatrician had horror stories of such home-births-gone wrong, but the statistics were confounded by this problem.
Thanks to Oregon asking “Did you go into labor planning to deliver at home or at a freestanding birthing center?”, we now have the answer.
In the Oregon data, perinatal death (defined as death in the first 28 days of life) occurred in 3.9 out of 1000 live births. Compare that to 1.8 out of 1000 among planned hospital births with a p-value of 0.003.
(There was also an increase in deaths in the first year of life, by about 0.6 per 1000, but, as it wasn’t statistically significant, I’ll leave that out – even though it would be another major downer if real. Likewise, these data only report deaths. It seems plausible there may also be significant differences in disability and neurodevelopmental outcome that are not captured in these data.)
Rounding down, that’s an extra 2 perinatal deaths out of 1000 associated with home births.
More Relative Risks
For the average American baby, her birthday is, by no small margin, the most dangerous day of childhood. Fortunately, only about 7 out of 1000 American newborns die before they turn 5 years old
(While this is a huge improvement on decades past, as you may know, it’s not our most impressive national health metric by global standards. The US ranks behind Bosnia, Montenegro, Bahrain and Slovenia, to name just a few, on this metric, with no offense to those countries intended. But that’s a topic for another blog.)
Of those 7, 6 die in their first year. Per CDC data, Most of those 6 deaths in infancy are related to issues at birth (i.e., prematurity, birth defects & birth trauma). If a baby makes it through her birthday without trouble, inn average risk terms, the next 50 years are a piece of cake by comparison.
So birth risks really stand out. And yet we worry, appropriately, about other risks to babies. Whole industries are built on minimizing such risks. I am glad they do, but those risks pale in comparison to birth complications.
Relative to What
Sudden infant death syndrome (SIDS) is tragic. Every case I have seen come through the ER has made me cry. Parents understandably go to great lengths to minimize the risk. (A quick Google search so as not to pick on any particular product.) But for all its tragedy, SIDS is way less likely than birth complications – consistently ~0.4 out of 1000 newborns per year. I wish every one of those deaths could be prevented, but it seems inconsistent to worry about getting just the right breathable mattress cover while forgoing a hospital birth.
Kidnappings are certainly a terrifying thought. Data on this turn out to be fuzzy, but the best estimate is that ~33,000 children in the US each year are “caretaker missing”, i.e., not where they are supposed to be for 1 hour or longer. (Only 115 cases per year are “stereotyped”, i.e., what we see in TV police dramas.) With ~72 million kids in the US, that’s ~0.4 per 1000 children (<18 years, not just newborns) per year.
Childhood gun injuries, accidental or otherwise, get much attention in the media and from advocacy groups (including the American Academy of Pediatrics, of which I am a member). In 2013, 1,670 children (<18 years) died by gunshot and an additional 9,718 were injured. With 73.6 M children in 2013, that’s 0.15 per 1000 children.
Deadly bee stings? Best estimate I found is ~50 cases total, adults & children combined, per year nationwide.
Biases & Caveats
On the Oregon Data
All kinds of selection bias are surely baked into that number. Home birthers may have a higher baseline risk because they also shun prenatal care. Or they may have a lower baseline risk because the known high-risk mothers (age-related, known pregnancy complications, known fetal anomalies, etc) are strongly advised to go to the hospital. I wouldn’t claim to know which bias is stronger.
On SIDS
SIDS is by definition death without an explanation. It’s very likely some of deaths in the Oregon study were “SIDS” and so we are, sort of, counting them twice. But that’s potentially part of the same problem – because SIDS is just an unexplained death, complications of home birth could well increase the risk of apparent SIDS. Even if they are totally unrelated, the magnitude of the home birth effect dwarfs SIDS.
On Pediatricians
Especially as one who does lots of neonatal ICU work, I fully admit to having a skewed view of the world. I see & remember the sad cases that go wrong and none of the successful home births.
