When science is anti-science

In which she peers back the curtain of evidence to see what's there

Forty-five years ago, Archie Cochrane — the Scottish physician after whom the Cochrane Library is named — awarded obstetrics with a wooden spoon. Now, if you’ve never heard of the phrase “wooden spoon award,” like I hadn’t before learning this little factoid about obstetrics, let me elucidate it for you: it’s a “booby prize” (another word that is not part of my lexicon, but I digress) given to a person or team that has come last in a competition. (In the original usage, the phrase referred to an actual wooden spoons — which eventually got as long as 1.5meters in length — that was “awarded” to the Cambridge University student who had achieved the lowest marks but still earned a degree, a practice that evidently began as early as 1793 but was eventually ended in the early 1900s).

In the case of obstetrics, Archie Cochrane’s proverbial wooden spoon was given for the dubious honor of being the medical specialty with the least science underpinning its practice, and specifically, for being the least guided by randomized controlled trials (RCTs).

When you read about this moment now, in the 20th century, the common narrative is that it was a pivotal moment, a clarion call for obstetrics to shift from “opinion-based” to “evidence-based” medicine, and that practice shifted accordingly and promptly.¹ The narrative is an interesting one to me, though, given that another seismic shift was happening to the field of obstetrics in the seventies: the adoption of continuous fetal monitoring (cEFM), a technology that was never validated prior to its introduction (and which, importantly, has never been since).

In 1968, eleven years before Cochrane awarded obstetrics the wooden spoon, Hewlett Packard produced the first commercially available continuous monitor, which was presented as a scientific breakthrough that would end cerebral palsy. Two international conferences on fetal monitoring were subsequently held, but consensus about the benefits, rationale, or indications for use could not be found, given the lack of studies. Forester King, an OB practicing at the time, remembers the response his UK colleagues had to the tech: “‘those Americans are gadget crazy—it will never take off here, give me a good midwife anytime, and besides, I think cerebral palsy won’t be prevented by Caesarean section, it is more likely to be the result of antenatal brain injury.”² A large randomized study was designed shortly thereafter but was refused funding on the grounds that the technology should in theory do what it said it would (i.e. reduce neonatal morbidity and mortality), and that to deny half of patients a potentially life-saving tool would be ethically suspect.³ cEFM should also, obstetricians hoped, reduce liability by producing a record of what had happened (as it turns out, the exact opposite happened there, too). Most physicians accepted cEFM inventors’ claims and promises, without any of the rigors of the scientific method, without any RCTs, and without any clearly defined parameters for use. By 1976, cEFM had been subjected to 12 clinical trials versus intermittent auscultation (the practice of listening to the fetal heart rate not continuously, but intermittently in intervals dictated by the stage of labor). cEFM showed no benefit and instead resulted in a dramatic increase in surgical birth rate. But by 1979, the same year Cochrane awarded obstetrics the wooden spoon award, at least one-half of all laboring people were being continuously monitored in the United States⁴; today, around 89% of laboring people are subjected to it as a matter of course.⁵ Forester King’s colleagues were wrong too: 60% of UK birthgivers today also are subjected to cEFM.⁶

In other words, if you are going to argue that obstetrics has become more evidence based in the forty-five years since Cochrane’s damning “award,” you’d have to completely ignore the rapid integration of cEFM. And it’d be a pretty glaring omission given it is pretty much the foundation of all modern obstetric practice, a technology that obstetricians live and breathe by, the second most common reason they perform a cesarean section. In fact , if there’s a super villain of modern obstetrics, cEFM would definitely be in the running, harming more birthgivers and babies than it has ever helped.

Cerebral palsy is still the most common motor disability in childhood, and it’s now widely understood that the disability is more likely to be caused by prenatal factors than factors during labor anyway. In fact, not once since its rapid adoption has cEFM been shown to improve the health of babies in any measurable way (Apgar score, rates of brain damage, NICU admissions, or death). The data hasn’t changed since those first few studies in the late 1970s: the only thing fetal monitoring has ever been shown to do is increase the cesarean birth rate which, since its introduction, has soared from 5% to 32%. It’s nothing more than a talisman, a relic, an amulet. And yet it continues to be used as a matter of course, despite the fact that the obstetric community knows this and has published damning study after damning study.⁷

In my last couple of newsletters, I’ve been exploring binaristic thinking in midwifery and obstetrics — and in particular the popular⁸ idea that evolution of obstetrics represented a shift toward a rational, “scientific” understanding of birth whereas the “traditional midwifery” it replaced was irrational, faith- and intuition-based and so on. But I’d argue that you need only look at the tyranny of continuous fetal monitoring to understand this kind of binaristic thinking is false; how could anybody argue that its use is rational in any way? Is it rational to cling to a practice that has never been shown to decrease harm but has repeatedly been shown to increase harm? We seem to have equated the simple use of technology we have at our disposable — whether or not it’s actually effective or beneficial — as scientific. But if the basic pretense of science is testing hypotheses in a systematic way and then making conclusions based on those experiments, then I’d argue that the use of continuous fetal monitoring is decisively anti-science. As Sartwelle et al write in an article about cEFM and bioethics, “clinging to obstetrical illusions and myths is merely practicing nineteenth century medicine, which was based on whim, personal belief, and bias, rather than evidence-based, scientific-based medicine. It also violates every past and current concept in bioethics.”⁹

The truth is, obstetrics, even in 2025, even 46 years after Cochrane awarded it the wooden spoon award, is littered with so many examples of practice defying what we know via “the evidence” that you could throw said wooden spoon pretty much anywhere on a labor and delivery unit and immediately find an example. For example, there’s the fact that we know, without a doubt, that vaginal exams are a primary cause of chorioamnionitis, or infection, particularly in the setting of ruptured membranes¹⁰, and obstetric still routinely overuses vaginal exams as a tool even when someone’s membranes are ruptured — only 3 of 23 studies on premature rupture of membranes included in the Cochrane database involve people who were not digitally examined in labor.¹¹ There’s the fact that we know laboring and particularly pushing on one's back is a disadvantage in every way — reducing blood flow to the uterus and therefore oxygenation of the baby (increasing "fetal distress”), decreasing the effectiveness of contractions, closing the pelvis and obstructing descent, and causing more tears — yet obstetricians are routinely limiting people to this position (read more about this in a past newsletter here). There’s the wealth of evidence that “two-step” rather than “one-step” “deliveries” (something I’ve written about at length here) prevents shoulder dystocia, yet obstetricians routinely insist on “one-step” deliveries, thereby often causing the very emergency they are attempting to prevent. There’s antepartum fetal testing (non-stress tests and biophysical profiles), another unvalidated technology that the American College of Obstetricians and Gynecologists admits in its own position statement has “a lack of high-quality evidence from RCTs” to suggest it “decreases the risk of fetal death” yet concludes that nonetheless it’s the best technology we have, and therefore is de facto better than nothing.¹²

And these are only examples of where the obstetrics blatantly bucks what is demonstrated by its own “evidence.” But only 15-30% of evidence being used to justify obstetric practice is considered “high quality,” in the first place.¹³ And here’s where we get in really deep: obstetrics is by no means an “evidence based practice,” but even if it were, it would cause harm because what is studied, how it’s studied, and who is studied in and of itself reflects bias at the outset. We culturally hold up “evidence based medicine” as the gold standard — as the opposite of the wooden spoon — and yet, when it comes to birth, it the evidence itself is always-already ambiguous because of what is being studied and how.

A perfect example of this is the “evidence” we use to understand “labor curves,” or what is considered a “normal” rate of progress for a labor (something that is of enormous consequence because “failure to progress” — a completely subjective diagnosis that has no standard definition behind it — is the most common reason for a surgical birth in the United States today).

For much of the twentieth century, obstetricians were guided by “Friedman’s curve,” a study published in 1955 by Emmanuel Friedman of Columbia University. The study was based on 500 white subjects — 70% of whom were 20 to 30 years old — in a single hospital New York City, but became widely extrapolated and generalized to birthgivers generally for many decades.¹⁴ Its dominance began to change in the late 2010s, when a study by Zhang et al, based on more than 62,000 subjects from 19 hospitals across the U.S., intervened.¹⁵ Though Zhang et al’s study is certainly more generalizable than Friedman’s due to the size and diversity of its subject pool, it’s important to note that both studies are deeply limited by the fact that neither of them study physiologic labor specifically. In Friedman’s study, 56.8% birthgivers were “delivered” by forceps, 13.8% received pitocin augmentation, and a whopping 96% were sedated by demerol and scopolamine (117, or 23%, lightly so, 210, or 42%, moderately so, and 154, or 31%, deeply or “excessively”). All of these interventions inarguably affect labor time.¹⁶ In Zhang’s study, the confounders (as I would personally refer to them) are different, but no less significant; the team report that “nearly half” of birthgivers studied received pitocin augmentation and 80% utilized epidurals.¹⁷

What this means is that what obstetrics uses to understand how long a labor should take is based on labors that were artificially shortened — or at the very least, altered — by obstetrics in the first place. Or put another way: we don’t have studies on how long labor takes, but how long obstetrics allows labor to take, and then those studies are used to limit how long a labor can take in a never-ending loop. I remember the first time I explained this to my husband, an Economist, and the complete disbelief he expressed. “There’s no way,” he said. “There has to be a control group that had no interventions.” But there isn’t. The best Zhang et al do is admit that the limitations of their study “may be worth mentioning:”

given the very high frequency of obstetric intervention (induction and prelabor cesarean delivery) in contemporary practice, only a third of all births in our large population were comprised of women who were at term, had spontaneous onset of labor and vaginal deliveries. [As well,] since intrapartum cesarean deliveries were performed according to the prevailing definition of labor arrest, some cesarean deliveries may be performed too soon (before 6 cm), which can cause early censoring of observation. This censoring may have resulted in a bias towards a shorter labor, particularly at the 95th percentiles. [Lastly,] nearly half of the parturients included in our analysis were given oxytocin for augmentation, which may have altered the natural labor progression. Thus, findings from our study must be interpreted within the context of current obstetric practice¹⁸

But at no point, to the best of my knowledge, has anyone ever bothered to parse out labor lengths in a population that even approaches physiologic birth.¹⁹

I am something of an evidence nerd — mainly, I’ll admit, because, if you dig deep enough, you can find plenty of evidence on which to argue that the less we do to birth the better, and because if you go deep enough into study design, a lot of accepted obstetric practice (which I find routinely abusive) digs its own grave. And as someone who nerds out on evidence relatively regularly, one thing that becomes obvious when you look into obstetric evidence is the fact that birth itself has not been, in fact, meaningfully studied. What has been studied is the management of birth. We have pulled the cart before the horse — starting from what we do to birth rather than how birth itself behaves — and then we treat birth according to that logic. We limit what’s possible because we have never bothered to meaningfully investigate it. I don’t mean this as an indictment of obstetrics generally (well, not entirely anyway). But I do think we need to intervene in the idea that it is in any way “objective,” “rational,” “scientific,” or “advanced,” — or more importantly that it reflects any kind of truth. It does not. It is simply another belief system, not as unlike the belief system that I discussed among tradition Thai and Malay midwives in a prior newsletter as it would have us believe. Obstetrics is culturally constructed and reproduced, and a lot of its foundation is fragile. When we unravel or peel back the way obstetrics makes meaning about birth we begin to collapse the distinctions between what is commonly thought of as “modern medicine” and “tradition,” of “science” and “folk wisdom,” of “rational” and “irrational” practices around birth.

Obstetrics foundationally set a paradigm around birth — that it is pathologic, that it is risky, that it must be controlled by man-made technology lest it end it catastrophe and death that was not necessarily based in reality²⁰ — and has designed and produced research that reproduces that belief. In almost any obstetric study, in fact, it becomes clear that often “the risks” of the complication being studied cannot be separated from the risks of the things we do to try to avoid or resolve that complication. Returning to the topic of labor curves, for example: we routinely assume that labors that fall outside of the normal curve are somehow dangerous, and pathologize long labors. Yet we actually have little to no evidence on which to base this belief. If you look at the data in the United States,²¹ the risks of a long labor are risk of c-birth, risk of infection, risk of instrumental birth, and risk of severe tearing. (Importantly, there is little evidence long labors are linked to poor neonatal outcomes such as low Apgar scores, poor umbilical cord pH, meconium aspiration, infection, shoulder dystocia, or birth trauma.) Every last one of those “risks” are the result of a human’s decision to impose something onto a labor, not the consequence of the labor inherently: a c-birth or instrumental birth happens because a provider has decided a labor is taking too long; an infection almost always occurs because of one too many vaginal exams; severe tearing is almost always the result of instrumental delivery or episiotomy, which are actions providers choose to do in an attempt to shorten a labor.²² We talk about these as if they are natural occurrences, outcomes that just happen, inevitable consequences of how a birth progressed. They are not. They are impositions onto a birth, guided by belief systems that are often based on nothing but that belief system.

This doesn’t mean that obstetrics has no place in birth. This doesn’t mean it has no value. It doesn’t mean that the knowledge it has produced doesn’t include kernels of truth, or that some of its tools aren’t life-saving. What it does mean is that it shouldn’t be privileged as the only way of knowing, or as the most important way of knowing. Because it’s not. Because, study or not, I can guarantee you that if we were to bring the knowledges those Malay midwives held about birth to modern obstetric practice, we’d probably in many cases improve not only outcomes, but satisfaction. This is the tragedy of privileging things like modern obstetrics: when only one form of knowledge and knowledge production is promoted and validated, other forms of knowledge that may also have value and may also “save lives,” are eliminated. And what often gets eliminated is also that which doesn’t serve the state, or profit the ruling classes, or that bolster systems that exploit and oppress. cEFM, for example, wouldn’t have prevailed if it didn’t turn a profit, if it wasn’t being used on a marginalized group of people (largely female birthgivers) from within a field foundationally rooted in patriarchal beliefs.

And most of us don’t experience or understand the world and our bodies from within the mechanistic paradigm obstetrics claims to think about the world and bodies in (though I think I’ve proven there’s no there there). We do not experience or understand the world as a discrete series of scientific phenomena. Music isn’t meaningful because we understand the frequency of a G-flat and the mechanics of how our ear transmits messages to our brain; we do not experience our child falling in terms of equations and explanations of gravity. Love isn’t, to most people, a feeling that can be explained away by chemicals. There are limits to what “objective” and “rational" science can explain about the experience of being human. And, likewise, there are limits to what obstetrics knows about birth, limits that we should be more honest about. Obstetrics is one way of knowing birth. It is not the best way, nor is it the objectively “safest” way. To deconstruct it as a cultural belief system rather than a reflection of truth allows for a much more robust way of grappling with it and what it has to offer us. On its most simplistic, being honest about the limitations of obstetrics and what it knows de-stigmatizes questions such as where to give birth, or with whom; whether to give birth at home or in the hospital becomes less the simplified question of, “do you want to give birth in the safest place, or in a less safe place?” and more a conversation about what risks one is willing to take (because all settings confer risk), what priorities one has for the experience, what values one wants to be guided by. It becomes an essentially generative discussion that has the potential to get to know ourselves better and more deeply. And that is also worth understanding and investigating.

In solidarity and love,

Robina

1. See, for example:

Barrett JFR. The Cochrane Collaboration and the Rise (and Fall?) of Evidence-Based Medicine in Obstetrics. J Obstet Gynaecol Can. 2019 Dec;41 Suppl 2:S185-S188. doi: 10.1016/j.jogc.2019.08.041. PMID: 31785653.

Forrester King, J. (2005). A short history of evidence-based obstetric care. Best Practice & Research Clinical Obstetrics & Gynaecology, 19(1), 3–14. doi:10.1016/j.bpobgyn.2004.09.003

Thorp J. Wooden spoons and thromboprophylaxis in obstetrics. BJOG. 2018 Aug;125(9):1049. doi: 10.1111/1471-0528.14886. PMID: 30004645.

Tita, Alan T. N. MD, PhD¹; Stringer, Jeffrey S. A. MD^1,2; Goldenberg, Robert L. MD³; Rouse, Dwight J. MD, MSPH¹. Two Decades of the Safe Motherhood Initiative: Time for Another Wooden Spoon Award?. Obstetrics & Gynecology 110(5):p 972-976, November 2007. | DOI: 10.1097/01.AOG.0000281668.71111.ea

2. Forrester King, J. (2005). A short history of evidence-based obstetric care. Best Practice & Research Clinical Obstetrics & Gynaecology, 19(1), 3–14. doi:10.1016/j.bpobgyn.2004.09.003

3. For some of this history, see:

Schmidt JV, McCartney PR. History and development of fetal heart assessment: a composite. J Obstet Gynecol Neonatal Nurs. 2000 May-Jun;29(3):295-305. doi: 10.1111/j.1552-6909.2000.tb02051.x. PMID: 10839578.

4. Schmidt, J. V., & McCartney, P. R. (2000). History and Development of Fetal Heart Assessment: A Composite. Journal of Obstetric, Gynecologic & Neonatal Nursing, 29(3), 295–305.doi:10.1111/j.1552-6909.2000.tb02051.x

5. Declercq, E. R., Sakala, C., Corry, M. P., et al. (2014). “Major Survey Findings of Listening to Mothers(SM) III: New Mothers Speak Out: Report of National Surveys of Women’s Childbearing Experiences, Conducted October-December 2012 and January-April 2013.” J Perinat Educ 23(1): 17-24.

6. Watson K, Mills TA, Lavender T. Experiences and outcomes on the use of telemetry to monitor the fetal heart during labour: findings from a mixed methods study. Women Birth. 2022 May;35(3):e243-e252. doi: 10.1016/j.wombi.2021.06.004. Epub 2021 Jul 2. PMID: 34219033.

7. There are endless studies to cite here, but here’s a sampling:

Am Coll Obstet & Gynecol & Am Acad. Pediatricians, neonatal encephalopathy and neurologic outcome. (2nd ed.) (2014).

Badawi N, Keogh JM. Causal pathways in cerebral palsy. J Pediatrics and Child Health. 2013;49:58. doi: 10.1111/jpc.12068.

Grimes DA, Peipert JF. Electronic fetal monitoring as a public health screening program: the arithmetic of failure. Obstet Gynecol. 2010;116(6):1397–1400. doi: 10.1097/AOG.0b013e3181fae39f.

Maso G, Piccoli M, DeSeta F, et al. Intrapartum fetal heart monitoring interpretation in labour: a critical appraisal. Minerva Ginecol. 2015;67:65–79.

MacLennan AH, Thompson SC, Gecz J. Cerebral palsy – causes, pathways, and the role of genetic variants. Am J Obstet Gynecol. 2015;213(6):779–88. doi: 10.1016/j.ajog.2015.05.034.

Parer JT. Personalities, politics, and territorial tiffs: a half century of fetal heart rate monitoring. Am J Obstet Gynecol. 2011;204(6):548–550. doi: 10.1016/j.ajog.2011.04.012.

8. But I know you know better, right, reader?

9. Sartwelle TP, Johnston JC, Arda B. A half century of electronic fetal monitoring and bioethics: silence speaks louder than words. Matern Health Neonatol Perinatol. 2017 Nov 21;3:21. doi: 10.1186/s40748-017-0060-2. PMID: 29201387; PMCID: PMC5697350.

10. Gomez Slagle HB, Hoffman MK, Fonge YN, Caplan R, Sciscione AC. Incremental risk of clinical chorioamnionitis associated with cervical examination. Am J Obstet Gynecol MFM. 2022 Jan;4(1):100524. doi: 10.1016/j.ajogmf.2021.100524. Epub 2021 Nov 9. PMID: 34768023.

Soper DE, Mayhall CG, Dalton HP. Risk factors for intraamniotic infection: a prospective epidemiologic study. Am J Obstet Gynecol [Internet]. 1989 Sep;161(3):562-6; discussion 566-8. Available from: http://www.ncbi.nlm.nih.gov/pubmed/2782335

Soper DE, Mayhall CG, Froggatt JW. Characterization and control of intraamniotic infection in an urban teaching hospi- tal. Am J Obstet Gynecol [Internet]. 1996 Aug;175(2):304-9; discussion 309-10. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/8765246

Wagner M V, Chin VP, Peters CJ, Drexler B, Newman LA. A comparison of early and delayed induction of labor with sponta- neous rupture of membranes at term. Obstet Gynecol [Internet]. 1989 Jul;74(1):93–7. Available from: http://www.ncbi.nlm.nih.gov/ pubmed/2733949

11. Middleton P, Shepherd E, Flenady V, McBain RD, Crowther CA. Planned early birth versus expectant management (waiting) for prelabour rupture of membranes at term (37 weeks or more). Cochrane Database of Systematic Reviews 2017, Issue 1. Art. No.: CD005302. DOI: 10.1002/14651858.CD005302.pub3. Accessed 26 January 2025.

12. Antepartum Fetal Surveillance: ACOG Practice Bulletin, Number 229. Obstet Gynecol. 2021 Jun 1;137(6):e116-e127. doi: 10.1097/AOG.0000000000004410. PMID: 34011889.

13. Stanfield VR, Chauhan SP, Huntley BJF. References supporting recommendations in American College of Obstetricians and Gynecologists obstetrical practice bulletins. Am J Obstet Gynecol MFM. 2022 Sep;4(5):100669. doi: 10.1016/j.ajogmf.2022.100669. Epub 2022 May 27. PMID: 35644524.

14. Friedman EA. Primigravid labor; a graphicostatistical analysis. Obstet Gynecol. 1955 Dec;6(6):567-89. doi: 10.1097/00006250-195512000-00001. PMID: 13272981.

15. That curve’s most notable intervention was that it argued that active labor began at 6, not 4, centimeters dilation — an important finding given that it should theoretically reduce c-sections for “failure to progress” or “labor dystocia” prior to 6 centimeters dilation, which it hasn’t because obstetrics is not an evidence-based field — and by lengthening the average amount of time it took for the average person to dilate one centimeter, which should also theoretically reduce the rate of c-section for “failure to progress,” but hasn’t.

16. Friedman does not report how many clients had membranes that were artificially ruptured during labor — he does not distinguish this as being different than spontaneous rupture of membranes, even though it is a known tactic for augmentation.

17. Again, artificial rupture of membranes as an augmentation tactic are not reported within the study).

18. Zhang J, Landy HJ, Ware Branch D, Burkman R, Haberman S, Gregory KD, Hatjis CG, Ramirez MM, Bailit JL, Gonzalez-Quintero VH, Hibbard JU, Hoffman MK, Kominiarek M, Learman LA, Van Veldhuisen P, Troendle J, Reddy UM; Consortium on Safe Labor. Contemporary patterns of spontaneous labor with normal neonatal outcomes. Obstet Gynecol. 2010 Dec;116(6):1281-1287. doi: 10.1097/AOG.0b013e3181fdef6e.

19. It could be argued that there is no possibility of truly physiologic birth in a hospital setting, since setting foot in a hospital is itself the first intervention, and routine practices in the hospital just as cEFM or intravenous administration of fluids no doubt alter the course of labor, but certainly even a data set that simply excludes induction and augmentation, forceps and vacuum, epidural, and c-birth would be stunning. Of course, as both these studies make clear, it would be an exceedingly small data set.

20. Our understating of historical maternal death rates prior to obstetric is generally inflated; it’s thought that a rate of about .5% was common in late 18th and 19th century America, a rate that soared with the advent of obstetrics and the introduction of hospital birthing. To be sure this is higher than what we’d expect now, but the advancements that changed this rate had little to do with obstetrics and more to do with basic health improvements, such as antibiotic use and the ability to screen and treat for hypertension. The truth is that then, as now, even in complicated cases, gestation and birth are vastly more likely to proceed without crisis than go awry.

21. Global data would be a complicated discussion here because of confounders of poor nutrition and maternal health, lack of prenatal care, lack of close by medical facilities, and so on — all things that mediate both the causes and the outcomes of prolonged labor.

22. I’ve written about a similar dynamic when it comes to things like fetal positioning here.