Friday, March 4, 2016

When evolutionary-minded medicine gets it (possibly) wrong about childbirth interventions

No one is saying that medicine isn't brilliant and hasn't saved lives. But it does intervene more than necessary when it comes to pregnancy and childbirth.

Part of that unnecessary intervention is driven by lack of experience. Part is an economically-driven disrespect for time. (Give childbirth some motherlovin' time.) Another part, related very much to experience, is how difficult it is to decide when intervention is and isn't necessary, especially when things are heating up. But another part of the trouble actually lies in the evolutionary perspective. Unfortunately it's not all rainbows and unicorns when M.D.s embrace evolution. Instead, evolutionary thinking is biasing some medical professionals into believing that, for example, birth by surgical caesarean is an "evolutionary imperative."

Here's one recent example in The American Journal of Obstetrics & Gynecology of how the evolutionary perspective is (mis)guiding arguments for increased medical intervention in childbirth.

link to paper
It's a fairly straight-forward study of over 22,000 birth records at a hospital in Jerusalem. The authors ask whether birth weight (BW) or head circumference (HC) is more of a driver of childbirth interventions (instrumental delivery and unplanned caesareans) than the other. Of course, the focus is on the biggest babies with the biggest heads causing all the trouble, so the authors narrow the data down to the 95th percentile for both. Presumably they're asking this question about BW and HC because both can be estimated with prenatal screening. So there's the hope of improving delivery outcomes here. And, of course, the reason they ask whether head size or body mass is more of a problem is because of evolution. They anticipate that they'll discover that heads are a bigger problem than bodies because of the well-known "obstetrical dilemma" (OD) hypothesis in anthropology.

OD thinking goes like this: Big heads and small birth canals are adaptive for our species' cognition and locomotion, respectively, but the two traits cause a problem at birth, which is not only difficult but results in our species' peculiar brand of useless babies. (But see and see.)

So, since we're on the OD train, it's no surprise when we read how the authors demonstrate and, thus, conclude that indeed HC (head circumference) is more strongly associated with childbirth interventions than BW (birth weight), at least when we're up in the 95th percentile of BW and HC. Okay.

They use this finding to advocate for prenatal estimation of head size to prepare for any difficulties a mother and her fetus may be facing soon. Okay.

Sounds good. Sounds really good if you support healthy moms and babies. But it also sounds really good if you already see these risks to childbirth through the lens of the "obstetrical dilemma" with that OD thinking helping you to support "the evolutionary imperative" of the c-section. Okay.

Too many "Okays" you're thinking? You're right. There's a catch.

When you dig into the paper you see that "large HC" heads are usually about an inch (~ 2.5 cm) greater in circumference than "normal HC" ones. (Nevermind that we chopped up a continuum of quantitative variation to put heads in arbitrary categories for statistical analysis.) And when you calculate the head diameter based on the head circumference, there is less than 1 cm difference between "large" and "normal" neonatal heads in diameter. That doesn't seem like a whole lot considering how women's bony pelvic dimensions can vary more than that.  Still, these data suggest that the difference between a  relatively low risk of having a c-section and a relatively high risk of having a c-section amounts to less than a centimeter in fetal head diameter. And maybe it does. Nobody's saying that big heads aren't a major problem sometimes! But maybe there's something else to consider that the paper absolutely didn't.

Neonatal heads get squeezed and molded into interesting shapes in the birth canal.

The data say that normal HC babies get born vaginally more often than large HC ones. But this is based on the head measures of babies who are already born! If we're pitting head circumference (HC) of babies plucked from the uterus against the HC of babies who've been through hello! then of course the vaginally delivered ones could have smaller HCs.

C-sected babies tend to have rounder heads than the ones squeezed by the birth canal. It's impossible to know but I'm fairly confident about this, at least for a subsample of a population: Birth the same baby from the same mother both ways, vaginally and surgically, and its head after c-section will have a larger HC than its squeezed conehead will after natural birth.

Measuring newborn head circumference (HC). source
When we're talking about roughly 2.5 cm difference in circumference or less than 1 cm difference in diameter, then I'd say it's possible that neonatal cranial plasticity is mucking up these data; we're sending c-sected babies over into the "large HC" part of the story just because they were c-sected in the first place. So without accounting for this phenomenon, the claim that large head circumference is more of a cause of birth intervention, of unplanned c-sections, than large body mass isn't as believable.

If these thoughts about neonatal cranial molding are worthwhile, then here we have a seemingly useful and very high-profile professional study, grounded in the popular but deeply flawed obstetrical dilemma hypothesis, that is arguing for medical intervention in childbirth based solely on the difference in head size measures induced by those very medical interventions. 

The circle of life!


Anne Buchanan said...

Oh my. Beautifully argued, Holly.

You know, an old friend of my family's recently died, at 94. He was an OB/GYN of some renown because he introduced the then radical concept of Lamaze to our area, under much protest from other doctors, hospitals and women. Until his last year he was attending obstetric conferences. His feeling was that a significant reason that the proportion of births by C-section has increased so much since his younger days is that medical students don't learn any more how to deal with 'difficult' births and so without those skills, they turn to surgical intervention much more readily than they would have if trained as older OB/GYNs were trained. Just a thought to throw in to the mix, though it has nothing to do with your argument today!

Holly Dunsworth said...

It definitely has everything, I think, to do with it!
It's anecdotal but it's in textbooks too. The longer you've been an OB, the lower your c-section rate.
One serious problem with all this is that new quantitative studies that create new knowledge or support medical/technological birth are valued over studies that aren't experiments (hard to do experiments on pregnant women!) and valued over qualitative studies too, and, even worse in many cases, valuded over working knowledge that exists already but outside the medical establishment. Midwives, people. I wish I was one of the people helping to bring midwifery together with medicine to improve the whole thing, rather than a bystander watching it get taken over by medicine. I'd love to do that but it will take some time for me to learn enough to jump in more actively... The blog and my other writings in journals, books, etc are all a start, an effort, something.

Holly Dunsworth said...

"valuded" (my typo) seems like a beautiful new word, or portmanteau. When you are deluded into valuing something!

Ken Weiss said...

These observations are very 'anthropological', that is, are about the culture of science. Somehow, the public image of science is that it is detached and objective and disinterested--just the facts. The truth is quite different, and you've been pointing out examples of that (as here). Emics and etics: what we say we believe and how we act are usually quite different.

Anne Buchanan said...

The valuding of Drumpf?

Sorry, couldn't resist.

Anne Buchanan said...

And, of course the other large social factor in C-section rates is legal pressure, preventing birth complications.

Holly Dunsworth said...


Holly Dunsworth said...

I bet there's a way to quantify that for many OBs it's objectively easier to c-section out a newborn than to guide the natural delivery of one. If that paper is out there, please someone guide me to it! If it is not, wouldn't that be a fun exercise to try?

Michael Finfer, MD said...

All I can say here is that if we discovered that one of our OB's was scheduling C-sections based solely upon a number like head circumference, there would be consequences.

Fortunately, these authors only want to study whether head circumference should influence prenatal counseling. I suppose that is hard to argue with as long as things don't go any further than that.

It seems to me that this is probably not going to be a terribly useful parameter, and that it could only be used to increase the C-section rate, which would obviously be undesirable.

One more thing to keep an eye on. Sigh.