Prenatal Exposures Oral History Project
Post-war Obstetric Practice: A Discussion with Mark Klebanoff, MD, MPH
Post-war Obstetric Practice: A Discussion with Mark Klebanoff, MD, MPH
"It gave me a good healthy dose of humility because the between-the-lines
I learned about you through your work on the Collaborative Perinatal Project (CPP), which tracked outcomes in more than 50,000 pregnancies from the 1960s. Could you share a bit about the scope of your research and your work generally as an epidemiologist?
My interests are reasonably diverse. I'm interested very broadly in risk factors for pregnancy outcomes and associations with baby size at birth, and on intergenerational correlation of pregnancy outcomes.
Were there certain risk factors that you were more interested in than others?
I started doing work on the correlation between a mom's own birth characteristics and those of her children. I've done a lot of work related to reproductive tract infections as causes of premature delivery and I've been involved in several different clinical trials of outcomes to try either to prevent hypertensive disorders of pregnancy or to prevent premature birth.
What cohorts did you use in your research?
Not only the original CPP, but we did studies with the Philadelphia and Providence women in the CPP, finding them 20 odd or maybe 30 years later to look at their pregnancy outcomes. I've done some work with the 1969 Danish perinatal cohort. I've done clinical trials with de novo data collection. I was involved in the analysis of the vaginal infections and prematurity study cohorts. So a variety of different groups over the last few decades.
I'm curious, when you talk about adverse outcomes and risk factors, were you talking generally about birth defects and low for gestational age weight or were you talking about other outcomes as well?
Shortened pregnancy and premature birth, low size for gestational age, preeclampsia, hypertension during pregnancy. I've not done a ton of birth defects work.
Along the way, you familiarized yourself with patterns and practices in obstetrics in those eras, late '50s through today?
Could you talk a little about your observations about obstetric practices from the '50s into the '70s. I'm an outsider, and was shocked to learn about some of the practices quite pervasive in that era. Not only with respect to the synthetic hormone drugs like DES or the others I was exposed to, but the amphetamines, the sleeping pills, the barbiturates, the anti-anxiety drugs, the smoking that sometimes the doctors would recommend.
My clinical training is in pediatrics, so I’m an outside observer, too. But generally speaking much that clinicians do doesn't have a great empirical evidence base to support it. Instead what happens a lot of time is people act based on what at that point in time we think is the underlying mechanism of the disease that we're trying to treat or prevent. We're sort of victims or beneficiaries of what we think we understand at the time. Certainly even things in pediatric practice have changed in the 30-odd years since I did my training. Even how we treated diarrhea, for example, we later learned that to a fair degree it didn't help and probably did harm. Even how we manage conditions like childhood gastroenteritis or asthma has changed.
So we don't use those pregnancy drugs anymore because we understand the physiology better. From looking at obstetrics as a fly on the wall, if there's a unifying fallacy, it’s that they got it backwards. Look at gestational weight gain. Probably in the 1920s obstetricians noticed that women with hypertensive disorders of pregnancy, in particular preeclampsia or even eclampsia which involves seizures and is potentially life-threatening, would gain a lot of weight before they started getting protein in their urine or hypertension. So obstetricians concluded that if we restrict women's weight gain, maybe it will prevent preeclampsia. Well the problem is they got it backwards. It turned out that hypertensive disorders are often associated with acute kidney dysfunction and of course these women would retain water.
In other words, the observation was correct. But interpretation was backwards. In fact, restricting weight gain really didn't do much of anything to prevent preeclampsia but it created a lot of anxiety on the part of pregnant women. In fact, I’ve heard that in the 1950s a lot of pregnant women actually took up smoking so that they wouldn't gain weight.
That's what I found in my research.
My mother was told in the early to mid '50s, that if she gained more than 15 pounds, she might need a cesarean, and this was in the era where that was dangerous. So she gained 12 pounds when she was pregnant with me. I have no doubt women came to prenatal visits scared to death to step on that scale and get yelled at by their OB for gaining too much weight. In that era, obstetricians were still worried about obstructed labor so there was a thought that we want to keep the kids small. But it's not clear to what degree that restricting weight gain changes that.
The same is true, I think, though I don’t know for sure, with some of the original rationale of the Smith & Smith DES protocol. In the post-World War II years we developed better ability to measure estrogen levels in blood. Women came in with threatened miscarriage or had bleeding and doctors measured estrogens or progesterone in their blood and found lower hormone levels. So they figured, gee, low hormone levels cause miscarriage.
It seems this was an underlying rationale for giving women estrogen. Again this would be getting it backwards. The problem with miscarriage was that most of the time when a woman comes in bleeding or cramping at eight or nine weeks, the embryo is already dead and more importantly the embryo has probably been dead for a month already, and so what you're seeing is the dying embryo causing hormones to fall.
Of course we know there's always regression toward the mean which means if your first pregnancy is horrible, your next one is might be worse than average but it may be better than your first just on random chance. So if a woman miscarried in her first pregnancy, and you gave her hormones in her second, and she had a better outcome the second time, it might have been just “regression to the mean”, but the obstetricians of that era thought it was because of the hormones. I'm going to give obstetrics a tremendous amount of credit for having learned. It took them a while, like from 1920 to 1980, but they learned from their mistakes.
I think the last major commonly administered obstetric intervention based on less than stellar evidence is probably electronic fetal monitoring. It was developed the '60s or '70s and was never really subjected to a good randomized trial, but it seemed like it ought to prevent stillbirth, and cerebral palsy and it became a standard part of practice. By the time people finally evaluated it, they concluded that its benefits were certainly overstated and it's not even clear what they are.
It had become entrenched in practice, but it did certainly create a lot of false alarms that led to a rise in the cesarean rate. When EFM first came or just before it came in, the cesarean rate in America, in the early to mid '70s was along the order of 5%, which was not all that different from the maybe 4% of the CPP. Of course the rate rose steadily throughout the '70s and the '80s and got up to about 25% by the end of that time period, and now I think around a third of babies are born by cesarean.
Another intervention that was proposed, but that obstetricians wisely abandoned following better quality evaluation was ambulatory contraction monitoring. The thought was if we put ambulatory contraction monitors on pregnant women when they're walking around or home, we can identify who's going to have a premature birth and get them in early. This was going on in the '80s but people finally did clinical trials and they found that it didn't work worth a damn.
Obstetricians recently evaluated a new tool to monitor the health of the fetus during labor- fetal electrocardiogram ST-segment evaluation. Most randomized trials showed it didn't improve pregnancy outcomes, and at least so far it has not caught on. So I think now when someone is trying to introduce a major new technology or drug, the obstetric community has been burned so many times by practices becoming standard or without any good evidence, or even doing harm, that they become a lot less willing to just lock, stock and barrel, adopt these things. That's just not going to happen anymore. Well, I suppose I shouldn’t say never.
In the CPP era, the mentality was so different because they believed in the placenta as a protective barrier.
And they thought about birth defects in a very narrow way.
Oh another classic CPP intervention which nobody uses, again, it reflects their imperfect understanding of physiology, relates to pregnant women retaining water. Essentially, every woman at term is going to have swollen ankles. In the era of the CPP they used to say these women are getting edema, so they need diuretics.
Diuretics, yes, my mom was given those, and I saw those high rates in the CPP. It seems like the diuretics were prophylactically administered to a huge percentage of women from what I could tell.
Yes. It’s normal to retain fluid. In fact, one thing we learned from the CCP is a certain amount of fluid retention actually portends a good pregnancy outcome, although it is true if you get massive fluid retention, it could be a sign that you're developing hypertensive disorders. But again getting rid of the fluids doesn't change the hypertensive disorder. Hypertensive disorders cause fluid retention so you can get rid of the fluids with diuretics is not going to impact what happens to the hypertensive disorder itself. A lot of what we saw in that era was people saw that A and B were connected. But if A is a cause of B, rather than B causing A, then you can get rid of B, it's still not going to change A.
We saw this with hormonal supplementation, with restricting women's weight gain, with diuretics and edema. Here’s another example. It was of Williams who wrote the original Williams Obstetrics textbook in the 1920s. He was an advocate of what you might call an aggressive obstetric philosophy, picking principles he learned in surgery and applying them to management of normal pregnancy. That’s why it was common to do an episiotomy because any surgeon would have told you that a clean incision is likely to heal better than a ragged tear.
You saw that as a standard practice. Finally, when people actually started looking at it, they found that in contrast to everything that you learn today in General Surgery 101, in fact a ragged tear doesn't heal any worse than if you prophylactically put a slice on somebody. In fact in randomized trials they found that overall outcomes are better if you don't do episiotomies routinely, and if somebody has a laceration and repair, you’ll do more good than if you do an episiotomy on everyone. But Williams introduced this very aggressive approach to obstetric management. The problem is nobody never actually empirically tested the general approach to find out if it it's necessary whether it’s even worse than doing nothing.
Everything clinicians of that generation thought they understood to about general health carried into obstetrics. Robert Goldenberg, one of the leading obstetric investigators of '70s, '80s and ‘90s and through today, has publicly stated that there was a generation of obstetricians who came of age during the 1920s to the 1950s who were taught spectacularly bad obstetric practice. He has actually said that in public lectures on obstetric care and prenatal care. I don’t think he meant it as a personal criticism, but just as a comment on applying a lot of interventions without ever formally evaluating them.
Also, think also about twilight sleep used in labor, it was so toxic.
I don’t know too much about that, but yes it was quite common to give women scopolamine, which is chemically similar to locoweed. It can make you psychotic, at least for a short amount of time. In that era, there was a lot of stuff that was based on what well intentioned people thought would work but reality was a lot of more complicated than they understood.
I’m interested in some of the other drugs of the era, like barbiturates and amphetamines.
I was involved in a case many years ago in which a woman was given amobarbital. In researching it, I found there were a lot of women on barbiturates who also got amphetamines. My guess is that if you're giving a woman amphetamines to restrict weight gain, the amphetamines made her hyper and nervous, so you also give her phenobarb to take the edge off. Amobarb was marketed as a combination pill to counteract the amphetamines, given in the first place to restrict weight gain, which turned out to be of no value and at worst, harmful. It’s a classic obstetric story.
Did you do research on prenatal smoking? I’ve had a hard time getting numbers, though it seems female, though not necessarily maternal, smoking peaked at about 36% in the mid 1960s.
That sounds reasonable, it might have been about a third or more when it peaked. Among men it was easily half. About half of all men smoked back when the Surgeon General's report came out and it's been trending down ever since. The CPP was a lower income population, it might have been 45% of women. Interestingly enough even through the early 2000s at the prenatal clinic at Magee-Women's Hospital in Pittsburgh, they still had a 40% smoking rate.
Appalachian states, Kentucky, West Virginia, Western Pennsylvania, Ohio where I live , these are one of the last bastions of smoking in America.
I'm out here in California, where smoking has declined for sure.
Oh, California is a whole other world.
But even here now vaping, e-cigs, are catching on and that's just scary.
Oh don't get me started on that. That's just horrible. I know it's debatable as people argue about whether it helps smokers quit. I don't know. I'm not a regulatory expert, but it seems when the e-cig people themselves say this is not tobacco, you're giving your own answer—if it’s not tobacco then it’s a drug and it needs to be regulated like a drug, period. If it's a way of helping smokers quit then make it prescription-only. Clearly it's a vehicle that get kids hooked.
We tend to forget how pervasive smoking used to be.
The first real talk about the harms of smoking came with the Doll and Hill case control study of lung cancer in the late 1940s. They included smoking to cover their bases, but I think they were the two most surprised people in the world when smoking turned out to be the main risk factor for lung cancer. They weren't expecting that. And Ernst Wynder in the early '50s showed the same thing, and of course Doll and Hill did their cohort of British doctors showing the same thing.
But in the 1940s nobody was expecting that cigarettes would be harmful. In hindsight you say, how could you think it wouldn't be harmful? This is one of my favorite stories. When I was an MPH student at Hopkins in 1983, in a lab course we looked at the effect of head and neck radiation on thyroid cancer, in a study of child refugees from Europe who immigrated to Israel. They prophylactically radiated these kids' scalps to kill the ringworm that they almost all would have had. Somebody followed these kids through the '50s to '60s and was able to show an elevated risk of later-developing thyroid cancer.
We students thought, how could you radiate a kid for ringworm? We asked our professor, George Comstock, and he just gave a half smile and shrugged his shoulders and said, well, in 1940 we thought we were pretty smart too. It gave me a good healthy dose of humility because the between-the-lines message was let's wait another 40 years to find out how many things we do today do more harm than good.
Whenever I want to criticize what somebody did back then I try to remember that quote because at the end of the day we're always victims of what we don't know.
Radiation is one of those pregnancy exposures that was very common for a long time.
Yeah, through the 1970s. It was not rare for a first pregnancy to routinely have a pelvic X-ray to try to estimate if there were pelvic deformities or if the woman was going to have trouble at delivery.
My understanding was that the CPP was developed and funded in the wake of the thalidomide tragedy.
No. The CPP long predates the thalidomide tragedy. The original history in my understanding is that people started to see that we had a problem with epilepsy, cerebral palsy and mental retardation, These were seen as either becoming more common or at least not getting better as medicine and society advanced. I think when the Neurology Institute was founded the director believed that we needed to do a prospective study of pregnancy and delivery to find out the causes of epilepsy, cerebral palsy and mental retardation. This was really the original argument, probably around 1950 or 1951.
Some things fortuitously happened. One is thalidomide, but that was never that common in America. And in 1964 there was the huge German measles (rubella) epidemic. Things like thalidomide and rubella were used over the years as justification to keep the funding for the cohort going but I don't think those issues had anything to do with the original establishment. The CPP certainly had a share of powerful enemies from the get-go who thought it a waste of time and money. And as a result, the CPP on numerous occasions had outside evaluations and reviews and had to defend itself.
Well, go back to the original purpose. How ironic that we have such an explosion of people with intellectual disabilities and autism now. In spite of all of these endeavors to look for risk factors in decades past.
The causes of some of these intellectual disabilities very well may be during pregnancy. They may predate pregnancy. But the events that happened in the immediate delivery period are by and large not important. I think this is the ultimate irony — because that's what they were expecting to see. It turns out that probably about 10% of cerebral palsy can be traced to events at or around the time of birth. So at least the obstetricians who were kind of beleaguered were mostly off the hook for how they managed the delivery. For epilepsy, it turned out pregnancy events were pretty unimportant.
In your experience working with the CPP or other cohorts, did you ever come across any questions or concern about the fetal germ cells? Of course that's my main area of focus. Did anybody ever bring up the idea that drugs or interventions could be affecting egg or sperm of the babies?
Not that I remember.
My reading of the literature is of course limited, but I can't find anybody bringing up questions about that aspect of fetal development. Maybe because it's just so invisible and so little was known about or maybe it just assumed to be just a bunch of DNA and who cares? I mean, for example, in the Heinonen book on the CPP.
On the birth defects?
Yeah, the birth defects book. They looked at all manner of outcomes but it's not even mentioned that maybe there could be a germ line effect. I think at that time it just wasn't on the scientific radar.
I don't think people talked about it back then. I don't think the concept was there. It would be so beyond the scope of what they could do that even if they had thought about it, they probably wouldn't put any great deal of time and effort writing about it because it would not have been measurable.
I’ve struggled a little bit to see if the CPP could be used for generational purposes to look at grandchild effects. What's your general advice for people who are interested in looking at intergenerational outcome? Whether it's with the CPP or other cohorts?
Find a good data source and be persistent. People don't keep records so find a source where you can get pretty good data on the exposure and the factors associated with exposure and then keep plugging away to find a way to get money to do it. But finding a good data source is hard because trying to find a 50-year-old medical records or even 30-year-old is not that easy. Then you still need good data on why this treatment or drug was given in the first place.
It's hard to do in America because the records are nearly nonexistent, though there are some exceptions.
And a lot of times even when they're there, they don't have the kind of detail you might wish.
I guess you have to also live with data that might be of less detail than you wish you had simply because where else are you going to find 30, 40, 50-year-old records? I mean you just live with what's there and make the best of it and try to make good interpretation of your data being careful about overinterpreting it simply because the quality of the data might not be what you want it to be.
I really appreciate your time and sharing some of your historical wisdom and I appreciate the work that you do.
Thanks so much.