Friday, July 8, 2011

Pre-Modern Death in Childbirth

More than once, I've seen homebirth advocates remark that homebirth is good and safe because it's natural: before there were hospitals, every woman gave birth at home, and most of them came through it just fine!

And just as many times, I've seen homebirth opponents reply: Yeah, umm, a lot of those women died. Giving birth at home is like giving birth in the middle ages, when mothers were dying left and right! (The Middle Ages gets mentioned a lot in this sort of argument. I guess we're to picture homebirth as if it were taking place in the most exceptionally ignorant, disease-ridden situation we can imagine.)

Of course, both arguments are wrong. The evidence of the past doesn't support or preclude homebirth. Things were very different then. Looking at the past, those of us in developed countries should be very glad that birth is so much safer now (for us, in our countries). In wealthy countries, our comparative good health and ease of access to information and skilled birth attendants makes birth remarkably safe, regardless of where we do it.

What, I ask myself, were the biggest killers of women back in the day? My suspicion is (1) infection, the deadly "childbed fever," and (2) postpartum hemorrhage. If I'm right that those were the major killers, well, that really speaks for the safety of homebirth today! Modern understanding of hygiene helps to prevent infections; antibiotics can (for now) take care of any infection that does occur. As for hemorrhage, it's scary, but any good midwife should see it happening and get the woman to the hospital in time, where the mother will most likely be saved. The literature for the LifeWrap notes that a woman can bleed to death in less than two hours. That's very bad news if you're hours away from a hospital, but it's kind of encouraging for women giving birth in developed countries who can get to a hospital in just a few minutes.

Anyway, I'm trying to do some research to see if my ideas here are correct. Here's what I have so far.

This website has a graph midway down the page of leading causes of modern maternal mortality. This is worldwide, not just U.S. Ignoring the many deaths from unsafe abortion, pregnant and birthing women are most likely to die of:
  1. Hemorrhage.
  2. Sepsis (infection).
  3. Obstructed labor.
  4. Eclampsia.
This WHO page gives different percentages, and it puts eclampsia down as a more frequent killer than obstructed labor, but the top four killers remain the same. That particular page cites a much lower percentage of death from unsafe abortion, though still 13%.

Maternal mortality is going to be dominated by places where we could characterize medical care as "pre-modern": places where women have poor pre-natal care, little or no access to trained and skilled birth attendants, and long distances or no access to hospitals. According to the WHO, "99% of all maternal deaths occur in developing countries, where 85% of the population lives." Thus, we can take worldwide causes of death in childbirth as representative of "pre-medical" birth.

The most deadly countries for mothers have maternal mortality rates of 1,000 maternal deaths for every 100,000 live births. That's about a 1% chance of dying each birth. (Well, probably greater, when you consider that some women are going to die after a stillbirth, or from an obstructed labor with no live birth.) In the countries with the worst maternal mortality, 1 in 6 childbearing women die from pregnancy-related complications in that country. That's in Afghanistan and Sierra Leone. It's 1 in 7 in Niger. (For comparison, in Europe, it's 1 out of 30,000 women. The worldwide average is 1 in 74 women. Source - Lancet article.)

That Lancet article, "Maternal mortality: who, when, where, and why," backs up the above-cited article and says that postpartum hemorrhage is the leading cause of maternal mortality worldwide.

That same article also notes the disparities in maternal deaths between rural and urban areas. On average, worldwide, a woman giving birth in a rural area is half again as likely to die during birth as an urban woman. (About 640 deaths per 100,000 in rural areas, 447 per 100,000 in urban areas.)


Okay, so, that's modern statistics that may be somewhat representative of past conditions for women. How about actual past conditions? That's not so easy because people weren't exactly doing big medical studies in the 1500s, but there is some evidence, of course!

Here's a cool article available as a PDF online that anyone can access. Published in 1982, it's called "An Attempt to Estimate the True Rate of Maternal Mortality, Sixteenth to Eighteenth Centuries." (Found via this article, which contains a fascinating account of childbirth in ancient Rome!) I will spend the rest of this section discussing that paper.

The article notes first that, of course, the evidence is flawed and incomplete; any estimate of maternal mortality from the 16th to 18th centuries cannot be any better than a very rough, error-riddled estimate. That said, here's what they found. In one parish in England, church registers counted 23.5 maternal deaths per thousand baptisms (so, per thousand births, assuming each birth results in a baptism). The London Bills of Mortality count an average of 15.9 maternal deaths per thousand baptisms from 1666 to 1758, not counting plague years. That's a maternal mortality rate comparable to that in modern Afghanistan. The paper notes that these are "certainly underestimates." For example, deaths from ectopic pregnancies or early miscarriage complications might not have been counted if they couldn't be recognized.

Continuing on, death rates in the mid-1800s were apparently lower, on the order of 5 maternal deaths per 1000 live births. That's a bit higher than Bangladesh's rates today. Odds of the mother dying were much higher when the baby was stillborn, ranging from 57 to 137 maternal deaths per 1000 stillbirths. That's as many as 13% of women dying while giving birth to a still baby. Sort of an intuitive result: unknown pregnancy complications, on which we can only speculate, mean a much higher chance of both maternal and fetal death.

Overall, the paper estimates about 25 deaths per 1000 live births from the 16th to 18th centuries. That's a 2.5% chance of death per birth, or 2500 in 100,000 live births. That's quite a bit higher than the rate in Afghanistan today, which is 1800 maternal days per 100,000 births.

As far as causes of death: a male midwife who lived 1596-1768 described postpartum hemorrhage as one of the leading causes of death in his patients. He also noted that women who delivered on their own had infections less often. Infection was nonetheless a great danger. The paper notes that "even in normal cases, the vagina was repeatedly smeared with materials such as butter, goose grease, capon's or hen's fat, or whole egg. Operative procedures almost inevitably meant infection." Not so hygienic, back in the day.

Now, this was just one paper, so take it with a grain of salt. However, from it, we can glean that pre-modern childbirth was more dangerous than it is in the most dangerous-to-birth-in countries today. Some evidence from New England suggests an average maternal mortality rate of 2.5%. That is, for every 1000 births, there would be 25 women who died. In countries with the maternal mortality closest to that, 1 in 6 childbearing woman will die from complications of childbearing; we can expect that the rate was similar in pre-modern times.


What's my take home message here?
  • Birth in the past was very dangerous. 
  • In developed countries, birth is very safe: we're generally healthy and well-nourished, making complications less likely. When complications arise, we can generally treat them quickly, effectively, and safely.
  • In many countries of the world, birth today is nearly as dangerous as it was hundreds of years ago. In far more countries, maternal mortality rates are 10 to 15 times as high as the rate that appalls us here in the States. We are right to care about having a rising mortality rate in the U.S., but it is also right to care about the staggering maternal mortality in the countries we don't think about very often.


  1. you seem to have a mistake in your calculations:
    in one place you write: "In the countries with the worst maternal mortality, 1 in 6 childbearing women die from pregnancy-related complications in that country. That's in Afghanistan and Sierra Leone."
    in another place you write: "The London Bills of Mortality count an average of 15.9 maternal deaths per thousand baptisms from 1666 to 1758, not counting plague years. That's a maternal mortality rate comparable to that in modern Afghanistan."
    15.9/1000 is closer to 1 in 60 than to 1 in 6.

    1. Most childbearing women have more than one child, so she has multiple chances to die. Your calculation assumes each woman only has one child.

  2. Thanks for taking the time to talk about this! I am in the process of looking for information on maternal mortality from premodern times and found your article very helpful. And thanks for citing your sources - they've been a big help to me, too!