Episode 8.1 Overuse of HCGs, Famous Studies in OB, and the History of Mag Sulfate
n the season 8 premiere, we discuss over-utilization of serial quantitative HCGs. Then we discuss several famous or influential studies in obstetrics, including the WOMAN trial, the CLASP trial, the ARRIVE trial, MagPIE, and several more. Finally, we discuss the history of how magnesium sulfate came to be used in obstetrics.
00:00:02 Overuse of HCG
00:13:17 Aspirin for Preeclampsia Prevention Debate
00:18:51 Debate on the ARRIVE Trial
00:29:10 MagPIE
00:39:05 Fetal Fibronectin Trials
00:44:53 Induction at 41 Weeks Discussion
00:55:16 History of Magnesium Use in Obstetrics
Links Discussed
Tranexamic Acid for the Prevention of Blood Loss after Vaginal Delivery
Low-Dose Aspirin to Prevent Preeclampsia in Women at High Risk
Labor Induction versus Expectant Management in Low-Risk Nulliparous Women
Randomized Trial of Labor Induction in Women 35 Years of Age or Older
Fetal Fibronectin in Cervical and Vaginal Secretions as a Predictor of Preterm Delivery
THE MANAGEMENT OF PRE-ECLAMPSIA
Transcript
Announcer: 0:02
This is Thinking About OB-GYN with your hosts Antonia Roberts and Howard Herrell.
Howard: 0:18
Antonia.
Antonia: 0:18
Howard.
Howard: 0:19
What are we thinking about on today’s episode?
Antonia: 0:21
Well, it is season eight for us, so we must be crazy, but we do have a lot of things planned for this season and we would love more listener questions and ideas and maybe suggestions of articles that any of you all out there would like us to discuss, or any other feedback, of course. And while you’re at it, share some of your favorite episodes with your friends and hopefully, all those new OB interns out there did well with power listening to our podcast right before they started this year.
Antonia: 0:54
And hopefully it’s been a strong start so far. But anyway, today we are going to talk about some famous and influential trials in obstetrics. You could call them landmark trials, perhaps, and we’ll tell you what they say, but also what they don’t. Say, that maybe a lot of people think they say that they don’t. But first, what is the thing we do without evidence?
Howard: 1:16
Well, how about routinely getting quantitative HCGs for patients who present to the ER or clinic with bleeding or other symptoms, who are beyond six weeks or so, who’ve already had an ultrasound showing intrauterine structures?
Antonia: 1:29
Okay, and I’m sure this is preaching to the choir. I’m sure anyone who hears this is like man. I hate it when people do that, and we do see that a lot with ER follow-ups, and I think we talked about that more in episode 10 of season 5. That was our episode of 10 things we wish ER doctors knew about OBGYN, but I suppose sometimes OBs might do this as well. So why don’t you quickly explain this to us and then we’ll get on to some of those landmark articles we wanted to discuss.
Howard: 1:59
Well, the idea, of course, is that patients are often getting some serial HCGs every 48 hours and expecting to see them double. But that doubling plateaus and it doesn’t hold true that trend after about six weeks, so really beyond five weeks. It just isn’t as informative as you might think it would be, unless you’re suspecting a partial mole or something like that and you have a reason. So you should just be doing an ultrasound and look for progression of the features of the pregnancy there on ultrasound.
Antonia: 2:28
Yeah, so you might see it if there is a known intrauterine pregnancy but there’s some question about the viability or prognosis of it, that somehow people might use HCGs in that scenario. But actually, even if you don’t have access to an ultrasound, in which case you have to rely on an HCG, or let’s say you’ve got an ultrasound but you’re just too early to see the pregnancy definitively, you still have to take that HCG trend with a grain of salt because it’s just so variable. Not even every viable pregnancy has that doubling. That is so reassuring to us Because remember there was that case report I mentioned on a prior episode by Mr Kurt Barnhart himself, the know-it-all expert on ectopics that wrote the article that every intern in my program pasted on the wall, and in his case report someone who did ultimately have a viable pregnancy in the study had a very abnormal initial trend. It was one of those very early ones and it had increased by some tiny amount, like 7% in the early stages and it was well documented. And then that patient happened to decline the intervention for what was presumed to be a non-viable pregnancy and then it was viable. So he included that to say there’s really no good minimum cutoff to say this is definitely viable or definitely not viable, and the same idea holds true for all kinds of other scenarios. So definitely, including when the pregnancy is already visible on ultrasound.
Antonia: 3:55
But when I think about it, I might have actually checked an HCG myself too, and maybe I didn’t need to. I can think of maybe an example, like when there’s no fetal heartbeat yet, but the fetus is less than seven millimeters in size by crown rump length, and we’re thinking maybe we can at least look to see. Is it massively dropping? Is the HCG massively dropping? Even by that we wouldn’t necessarily diagnose that it is now a miscarriage. We would still wait for that ultrasound trend.
Antonia: 4:24
But sometimes people just want to know as much as they can as soon as possible. So those are some situations where we’ll talk about getting an HCG and sometimes get it too. But even then there is no specific diagnostic threshold of an HCG drop in the setting of an early intrauterine pregnancy to determine if you’re dealing with a miscarriage or not. So it can drop and you could still have a viable pregnancy. So really, ultrasound, and especially serial ultrasounds with an adequate time interval, are 100% superior to HCG trending, at least for diagnostic purposes. And of course, if the structure on the ultrasound doesn’t look normal, or maybe they have the signs of a molar pregnancy, with excessive nausea and vomiting or other symptoms. Then, yeah, maybe a very extreme HCG could help with that.
Howard: 5:16
Even that’s tricky. I’ve seen just miscarriages with normal tissue that had HCGs of 40 or 50,000 that I was convinced I was going to find a mole. And maybe you went with surgical management over medical management because I thought it was a mole. So even those can be confusing.
Antonia: 5:32
Yeah.
Howard: 5:33
But yeah, in general I think we do too much of this serial HCG following.
Antonia: 5:38
And a simple blood HCG test can cost about $70. Yeah, so let’s get into this series of famous articles. So first let’s talk about the WOMEN trial. That’s an all caps acronym. It was published in 2017 in the Lancet and it was a randomized control trial involving more than 20,000 women in almost 200 hospitals in over 20 countries to study the effects of tranexamic acid IV on women with postpartum hemorrhage.
Howard: 6:10
Yeah, this trial will be discussed anytime you hear a discussion of TXA. This trial is commonly summarized by stating that tranexamic acid reduced the rate of death from postpartum hemorrhage, which is technically true. That’s what they found. I think what’s not talked about so much was that the primary outcome was a composite endpoint of death from all causes or hysterectomy within 42 days, and for this outcome there was no difference discovered. So the death due to bleeding was a subset finding and the reduction of death in the subset was barely statistically significant, with a p-value of 0.045 and a number needed to treat of 267. But there was no increased causes of death in any other subsets from a statistically significant point of view. So they just concluded that death from hemorrhage was reduced.
Antonia: 7:01
Is this an example of where the other subsets were so small that you didn’t see a statistical significance, or is there something about the composite outcome of death plus hysterectomy that might make this difficult to interpret?
Howard: 7:14
Yeah, that’s a good point, and one of the reasons why I hate these sort of composite type outcomes is because this problem happens, but with rare outcomes like death. Sometimes breaking it up into subsets will produce this p-value of 0.045 that really doesn’t have the statistical validity you would expect it to. This is part of the multiple comparators problem and essentially you have a potential type 2 error with the other outcomes death due to other causes. In other words, there were more deaths in the TXA group from organ failure and sepsis, but these weren’t statistically significant and overall there were 4% more deaths from hemorrhage, with 1% more deaths from all other causes in the comparator group. So it does seem like that you could conclude, as people do, that less women died in the TXA group even when you aggregate them all together.
Antonia: 8:03
So were there more hysterectomies in the TXA group or more other causes of death? That then seemed to counterbalance the benefits in the composite outcome.
Howard: 8:15
Yeah, and that’s a good point too, and I don’t know that hysterectomy and death are sufficiently similar to be included in the composite outcome. But I see what they’re doing and, yeah, there were seven more hysterectomies in the tranexamic acid group, that is, 358 compared to 351, as well as the seven additional deaths in the causes of death that weren’t hemorrhage. So that’s 14 additional outcomes by the composite definition in the tranexamic acid group, whereas the placebo group had 36 additional deaths from hemorrhage. So, yeah, I do think that the combination of the hysterectomies might have diminished the mortality benefit, and we just have a large data pool where a few different outcomes here or there of these rare outcomes can really change the statistical view of it. So I think the bigger criticism of this paper is that it may not be relevant in first world countries where our approach to postpartum hemorrhage and the availability of resources for the patients is very substantially different. But I do think it’s fair to say that in low income and middle income countries where this study was conducted, the TXA likely leads to decreased deaths.
Antonia: 9:18
Well, so in the context of high income countries, there was a study published in the New England Journal of Medicine in 2018 in France where prophylactic TXA was found not to change the amount of blood loss or the need for transfusion with vaginal delivery. Specifically, if they were already getting postpartum Pitocin, then adding this did not improve anything. And then the same group in France also did this same study for cesarean delivery. They published in 2021 and essentially the same thing. They incidentally found, even though this was blinded, there was a lower estimated blood loss with the tranexamic acid group, but the quantitative blood loss was not less.
Howard: 10:04
Yeah, I think we all know we’re not good at estimating blood loss.
Antonia: 10:08
Yeah, yeah, exactly.
Howard: 10:09
And we should talk about quantitative blood loss mechanisms sometime and what impact they may have. It’s clearly true that when we do quantitative blood losses we’re getting a better result, but whether that translates to better patient outcomes yet is still a gray area. But well, so we’ve probably been overusing tranexamic acid since the woman trial, and its chief benefit is in helping to prevent the onset of DIC. So I think we need to remember that it doesn’t treat bleeding from uterine atony or vaginal tears or surgical oozing at a cesarean. But for a patient who’s having a significant hemorrhage and who’s on her way to receiving blood products or is receiving blood products, TXA might prevent some patients from starting down the cascade that leads to DIC. But it would be wrong to say that tranexamic acid has been shown to reduce the risk of hemorrhage or mortality. In settings like the United States. That just hasn’t been shown yet.
Antonia: 11:01
Yeah, of course it would have been nice if this something that you use later on when the hemorrhage gets worse, if you could use it earlier on and prevent it. But that’s not really the case. So it’s just something to keep in your back pocket once you’ve done three or four different things first, and not use it as the first thing necessarily. Because remember, as we’ve mentioned before, there have even been obstetric deaths related to TXA when it was inadvertently administered through the wrong line, like through an epidural catheter, and in the same way there could be deaths from inadvertent overdose of oxytocin, or probably giving that through the wrong line, or hypertensive crisis after methergen or a bronchospasm after hemobate. So with every medicine it just has to be used very judiciously and carefully. So let’s move on to the next trial. How about CLASP? It is hard to find just one single seminal study about aspirin and pregnancy, but this trial was published in the Lancet in 1994, and it was a randomized trial of low-dose aspirin for the prevention of preeclampsia in over 9,000 pregnant women.
Howard: 12:09
Yeah, the CLASP trial. I would love to get again the job naming trials or drugs, but this was the Collaborative Low-Dose Aspirin Study in Pregnancy CLASP and it was done because lots of small studies up to that point suggested that there might be a benefit of aspirin for a lot of different outcomes, actually including growth restriction. And then a theoretic mechanism from our understanding at the time of preeclampsia that there might be a mechanism by which aspirin might prevent preeclampsia. And while there were fewer cases of preeclampsia in the aspirin group, the difference was not statistically significant. But in a subset of women who entered the study prior to 20 weeks the difference was larger. So this trial spurred other trials and suggested that earlier treatment might be more effective at producing the desired outcome and generally showed that bleeding and other potential safety issues like fetal issues that we might have been worried about, weren’t significant. But they gave it to women well out of the first trimester in that trial.
Antonia: 13:08
There was an increased need for blood transfusion after deliveries, though some of us already now stop aspirin earlier to avoid that risk. So what were the trials that followed CLASP?
Howard: 13:20
Well, in 1998, there was a randomized controlled trial that was published in the New England Journal of Medicine where patients were enrolled between 13 and 26 weeks so still later than we like today and this trial again found no difference in the rates of preeclampsia. It also didn’t seem to matter if the patients in that study had had chronic hypertension or pre-existing diabetes or twins or even a history of preeclampsia in a previous pregnancy. These things we think of as high risk. There was also an Italian study published in the Lancet in 1993. That was a randomized controlled trial and again they found no difference in the rates of preeclampsia or birth weight or things like that.
Antonia: 13:58
So by 1998, still lots of large trials consistently finding no benefit of aspirin.
Howard: 14:04
Yeah, and that was the case up until 2017. And then there was a trial published in the New England Journal of Medicine from Europe, where patients were given a higher dose of aspirin 150 milligrams per day and were enrolled earlier, between 11 and 14 weeks, and in this trial of about 1,600 women, there was a statistically significant difference in the rate of development of preeclampsia. They also, though, included patients in that study by definition who were very high risk based upon things like first trimester ultrasound and uterine artery notching profiles and other serum analytes that were tested on these patients, and these are things that we don’t commonly do in the United States outside of clinical trials. So they were dealing with a predefined, very at-risk population with double the dose of aspirin.
Antonia: 14:50
So even now there’s still no single trial that shows that the 81 milligram aspirin in any way reduces preeclampsia.
Howard: 14:58
Yeah, and I think people are probably surprised to hear that. However, the practice comes from a meta-analysis of all these older trials, none of which found a difference that was statistically significant, but when their numbers were added together and people tried to educate that through meta-analysis, then they found this possible slight, modest effect.
Antonia: 15:18
Well, and I guess that can be tricky, because this is another example of where we don’t really see harm in a baby aspirin and then we see the potential benefit. So given the possibility that there could be about a 10% reduction in the rate of preeclampsia only from data pulled from lots of individual trials together that individually didn’t pass semester, then we use it because it might work. So this is maybe not exactly a thing we do without evidence, but it almost is.
Howard: 15:50
Right, and it’s a cautionary tale. There could be a publication bias for all these small trials that were published in the 90s and 2000s, and the ACOG bulletin about this, by the way, admits that many small negative trials aren’t likely to be published, and so this is a danger of trying to make something out of a bunch of small and perhaps individually inconsequential trials this file drawer bias, where the negative trials weren’t published. If there were more harms of aspirin that we knew about, we probably wouldn’t be recommending it. But this is also just a reminder to residents and others that you shouldn’t be so dogmatic about things even like baby aspirin, which is really a big part of our culture right now, or, frankly, a lot of other interventions, because we, to be honest, we don’t even really know for sure that it works or let alone how large the magnitude of effect is, and that’s very important.
Antonia: 16:39
Well, and that’s really nice for me because I’m allergic to aspirin, so good to know that it wouldn’t have really helped me that much anyway. But you might assume, from the USPSTF guidance at least, that aspirin is some kind of secret weapon against brachyclampsia. But, as we’ve just concluded, we don’t even know if it does anything, or at least not the 81 milligram dose that we frequently use. But we may potentially see future trials refine this. They might validate a higher dose aspirin or just its use in higher risk patients, or both. But yeah, I agree People shouldn’t be so adamant that baby aspirin is the key to preeclampsia prevention, especially in patients who don’t have a lot of risk factors or any high risk factors and a lot of these risk factors themselves are quite controversial.
Howard: 17:27
I didn’t put a link to that, but David Sachs in California wrote a piece in the New England Journal of Medicine sometime earlier this year I don’t remember the month, but it was commenting on the fact that ACOG still includes black race as a risk factor for one of the smaller risk factors for aspirin use.
Howard: 17:45
But he pointed out that in a lot of the studies, and even in a study that was designed in a different way, retrospectively to try to sort some of this out after use, black women who were put on aspirin didn’t have lower rates of preeclampsia. And the problem is and it’s definitely worth reading if anybody gets into the Journal of Medicine, find that article and read it. He makes a lot of good points, but the problem is that pretty much the same thing could be said with almost all the risk factors that we now include. So we don’t include the risk factors because a study has shown that use of aspirin reduces the rates of preeclampsia. We merely include them because those are things that increase your risk of preeclampsia but you can be at increased risk of preeclampsia and not benefit from aspirin.
Antonia: 18:25
There we go. Well, let’s move on to the next one, so this one. We’ll probably spend a little more time on the ARRIVE trial. We’ve talked about it before. Obviously. This was a multicenter trial published in 2018 that randomized low-risk nulliparous women at 38 weeks either to getting induced, at 39 weeks for delivery, and were completely low risk.
Howard: 18:49
Yeah, and we’ve discussed this before. The ARRIVE trial is definitely a controversial study still today. It was designed to see if routine induction at 39 weeks would prevent significant neonatal morbidity and mortality and we all know that it will prevent 100% of any stillbirths that would have happened after 39 weeks if you deliver the patient at the end and risks related to post-maturity syndromes. But those risks are already so low that the question is whether there would still be a net benefit or harm from that approach. And also, you have to keep in mind that we may just be shifting categories. There’s some evidence to suggest that preventing a fetal demise at 40 weeks doesn’t prevent a perinatal demise, because that newborn just dies in the neonatal period, so it’s an aggregate perinatal outcome that matters, not the fetal demise rate. So common sense isn’t always applicable here. Now maternal outcomes in this study were also followed to see if any potential improvements in total perinatal outcomes would come at the expense of maternal outcomes. And this is interesting because I recently heard a podcaster try to portray this the other way. She said that the trial was performed to see if induction at 39 weeks was associated with worsening neonatal outcomes. The assumption here would be that babies weren’t ready for birth yet and needed to cook longer which is a quote she used if the mother wasn’t already spontaneously laboring and that induction might be associated with worse outcomes for that reason Essentially a prematurity-like syndrome, even when they’re technically a term. Now that’s not actually true at all.
Howard: 20:20
I was surprised to hear that and I’m sure people are talking about it this way on the rounds and that’s not why the study was done. But introducing it in that way will make it sound like the primary finding was positive, since they didn’t find a difference in neonatal outcomes. The truth is they were looking for a benefit. They had hypothesized that there would be improved outcomes with induction of labor at 39 weeks by again reducing fetal demise and actually total perinatal demise. They did it the right way, and other conditions of course related to placental senescence as the pregnancy aged. So by finding no difference they actually failed to support their hypothesis. So this was a negative trial, not a positive trial. But it’s interesting how the tables turned by that reframe and I hadn’t heard it before. But in the study they did find a lower rate of cesarean delivery among the induction group and we’ve talked about this study before and some of the reasons why that lower rate of induction may not be clinically significant to your practice, or even a true finding.
Antonia: 21:19
Or at least not generalizable to practices with different baseline rates of primary cesarean than what they had in this study, especially not places that have a much lower baseline rate of cesarean. And yeah, that is interesting that reframing the intent of the study turns a negative into a positive. And I’m assuming that whoever this podcaster was that you’re referring to either had some connection to the study or some other personal reason for wanting their conclusion to be true that it is good to induce everyone as much as possible at 39 weeks.
Howard: 21:55
No, not at all. Just something, just a summative thing for students and residents, trying to teach them about the study and why they induce people at 39 weeks, but just a mistaken interpretation.
Antonia: 22:06
Okay, okay, well, maybe if they’re listening they’ll rethink that. It’s worth just rethinking that assumption. But we already understood that fetuses born between 39 and 41 weeks gestation have generally the same long-term and short-term outcomes, regardless of when in that interval they were delivered. And that’s one of the reasons why elective induction at 39 weeks is medically acceptable. So if you have the capacity at your facility, it’s totally fine to offer it, and that’s why repeat cesarean at 39 weeks is also acceptable. And these were both acceptable long before the ARRIVE trial. They were standard of care elective inductions any time after 39 weeks that it could be accommodated. So the reframe you mentioned falls flat on its face. That wasn’t a question before and nobody thought that 39 weeks was too early from the fetal maturity standpoint and we really did not need this massive trial to answer that question. That wasn’t even a question.
Howard: 23:10
Exactly Well. The authors were definitely hoping and trying to show that perinatal and neonatal outcomes would be improved. In fact, this is demonstrated by their willingness to let the expectant management group go all the way to 42 weeks or even just beyond, which is a bit more than a week later than most of us currently allow women to go, due to a known risk of worse perinatal outcomes after 41 weeks. Yet even with those truly post-term deliveries mixed in with the control group, there was no difference observed in the infants. As far as the different risk of cesarean delivery, yeah, they found a risk of cesarean of 18.6% in the induction group versus 22.2% in the expected management group, and this finding has not been replicated. Nor was it demonstrated in a similar trial in the United Kingdom published in 2016. That was also a randomized trial of labor induction where cesarean rates were similar in both groups.
Howard: 24:07
It also is contradictory to just a tons of other evidence about whether or not earlier induction increases the rate of cesarean, particularly in nulliparous patients.
Howard: 24:16
So the main criticism of the ARRIVE trial is that we don’t have consistency in the scientific literature, that the cesarean rate is lowered by these elective inductions and also that their rate of cesarean was so high to begin with. The baseline rate of 22.2% for the very healthiest population of patients that they take care of is high, and so the simple fact of inducing patients and putting them on a standard protocol may have been enough to lower that rate to something more acceptable. I actually spoke to one of the authors one time and asked this question and he said that, yeah, the ARRIVE trial should not be applied to people with a lower rate of cesarean frankly more appropriate NTSV rate than they had in their study. And note that that 22.2%, that’s not the NTSV rate, because the NTSV rate would include patients who might have preeclampsia or diabetes or oligohydramnios or growth restriction or other medical problems that they excluded. So they excluded every kind of high risk patient that might be included normally in your own NTSV rate.
Antonia: 25:20
Okay, and just to recap for our listeners NTSV means nulliparous term, singleton vertex, and it’s this acronym that we use to classify patients who should have no reason to have a C-section before their labor process starts Like they should expect to have a vaginal delivery, and that means any cesarean that happens for them is from something that happened during labor. And we use it as a professional quality measure because, with very few exceptions, it demonstrates what’s going on in your labor ward. Because, as we’ve discussed before, a lot of those cesareans in labor are done in kind of a subjective way and reflect how is the labor being managed by the care team and ultimately by the obstetrician. And by only looking at the nulliparous patients we eliminate lots of potential confounders about how prior deliveries would affect any multiparous patient’s current delivery.
Howard: 26:15
Yeah, On a later episode we should talk about the different ways of primary section versus NTSV, and I actually have an idea for a rate that measures the inverse. Instead of measuring your section rate, how about we measure your spontaneous vaginal delivery rate to exclude forceps and vacuums and everything else and also include those other patients, Because it could be that you’re doing more operative deliveries because you have a low section rate. You could be doing more operative deliveries just because you do too many, so maybe we should be measuring people’s spontaneous vaginal delivery rate. But we can debate that later.
Antonia: 26:51
Yeah, we’ll refine that idea and it sounds like something I want to come back to as well. So if you’re not already universally inducing every patient at 39 weeks and your primary cesarean rate is much lower than what it was for the folks in the ARRIVE trial, then their findings probably should not change your practice. But if your cesarean rate in those patients is closer to about 22%, then there probably are many other things to examine about your practice that might be more helpful than just simply adopting a universal 39-week induction policy.
Howard: 27:27
Yeah, If you’re trying to work on your section rate, it’s probably not this that’s going to make the biggest impact, but it might Well. The bottom line is it didn’t improve neonatal outcomes and only improved maternal outcomes. If you already are in an environment where you have too high of a cesarean rate and the findings not been replicated in our literature.
Antonia: 27:46
All right. Other landmark trials in obstetrics include the BEAM trial and others involving magnesium for prevention of cerebral palsy in neonates, and we discussed those more in detail in episode 10 of season six, so let’s not go back into those again here. We’ve also discussed the 17-hydroxyprogesterone caprate in detail way back in season one, episode seven, and the problems with the studies that led to its routine use and then eventually to its abandonment and et cetera. So we’ll skip over those again for now as well. But for those uninitiated to those topics, then they are some pretty important and controversial studies in our field. So if you’re listening to this in 2024, you probably still have a lot of patients, or maybe people you know, who used to do weekly progesterone shots in their pregnancies to prevent preterm delivery. So go listen to that episode to understand all of that if you don’t already have an intimate understanding. I don’t think we’ve really talked about the MAGPI trial as it relates to using magnesium to prevent eclampic seizures.
Howard: 28:55
Yeah, sure, yeah. So magnesium was originally used for something different than it came to be used for, and still is so. The MAGPI study was published in 2002 and was an international trial in 33 countries to see if prophylactic magnesium sulfate for preeclampsia would reduce eclampic seizures. They enrolled 10,110 women with both severe and non-severe disease and they found that magnesium reduced the risk of seizure by half, from 1.9% to 0.8%.
Antonia: 29:26
It is difficult sometimes to see how data from middle and lower income countries might apply to the US, but this study did encourage us to consider using magnesium even in patients who didn’t have severe features of preeclampsia, and that practice did become popular for some time after the MAGPIE trial. We were already using magnesium for seizure prophylaxis prior to magpie, so we also talk about magpie a lot because a lot of other outcomes related to the fetus which we discussed in a prior episode for preventing cerebral palsy they originated from this. But what was it about magpie that convinced people to use magnesium for the quote mild preeclampsia, but without severe features?
Howard: 30:14
Yeah, with moderate, or I should say, mild blood pressures, et cetera. Well, there were some smaller trials that looked at preeclampsia without severe features. We used to call that mild preeclampsia when this trial was published, including by Whitland in 1997, that found no differences and no seizures in the magnesium or the placebo groups, and then a similar small study by Livingston in 2003 that once again found no seizures in either group. So the incidence of seizure with preeclampsia without severe features is just very small. Magpie also used a higher blood pressure cutoff to define severe disease, and so it’s probable that many of the patients received antihypertensive medications before being randomized, so that patients with severe disease might be represented in the study as not having severe disease. And then, of course, the difference in the blood pressure criteria also probably misallocated some patients. So, yeah, there was this study ostensibly found a reduction in both groups, but it really wasn’t consistent with the evidence and it wasn’t done in a way that really could be shown to magpie and I think mostly we’ve moved away from that.
Howard: 31:32
ACOG doesn’t recommend it, but there was a few years after this where you saw a lot of all preeclamptics being magged.
Antonia: 31:39
Yeah, I’ve had some attendings say, well, back in the day we magged everybody, basically, but that’s not what any of us currently do. But yeah, 15 years ago or roughly thereabouts, there was a different understanding. So let’s move on to the next trial. This one was the seminal trial on fetal fibronectin. That was published in 1991 by Charles Lockwood in the New England journal of medicine. So this study claimed to find that the presence of a certain amount of fetal fibronectin, or FFN, in a cervicovaginal sample identified a subgroup of women at higher risk for preterm delivery. And of course then the FFN test was heavily marketed after this publication and is still used in various ways by obstetricians even today, but none of it can be said to have strong evidence backing it.
Howard: 32:30
Right, this study was looking at patients and trying to determine the value of a positive test again when they had an above a certain predefined threshold of fibronectin and, of course, later, as you said, it was marketed as having a value.
Howard: 32:42
But it was marketed for being valuable for a negative result in terms of having some predictive value that a patient was going to remain pregnant for the next week or continue to term or some metric.
Howard: 32:54
What it found was that fetal fibronectin had a sensitivity of 81.7% and a specificity of 82.5% for preterm delivery, at least in the population of patients that were studied, and it also found that it was almost always positive if the patient had ruptured membranes and that 83% of patients who had a positive fibronectin along with symptoms of preterm labor were likely to deliver, compared to 19% who didn’t have symptoms but still had a positive fibronectin. But this shows how even a marginal finding that may be clinically irrelevant can be turned into a very successful product. This gets back really to Bayes’ theorem and understanding that a sensitivity and a specificity of around 80% or so are actually meaningless, depending upon what your pretest probability is for that patient. So if you clinically see a patient who’s 32 weeks pregnant and she’s contracting every four, five, six minutes and her cervix is closed. And you recheck her in a couple of hours and her contractions are now a little spaced out. They’re every seven or eight minutes and her cervix is still closed. Well, what do we think? Her pretest probability is that she’s actually in preterm labor.
Antonia: 34:04
Well, it sounds low compared to the chance that this is just Braxton Hicks contractions that will go away without any cervical dilation. So I’m going to guess probably about 5% chance that she is actually in preterm labor, just from my experience.
Howard: 34:22
Yeah, Okay, so, and this is hard, this is what we do, right, we use our experience to develop these estimations, but we can also use data sets. But let’s call it about 3% pre-test probability that the patient is in preterm labor. Based upon that clinical presentation, she’s already shown she’s not changing or contractions are spacing out. So even with the sensitivity and specificity that this study found which the study has been criticized for lots of other reasons but even with those published numbers as they are, the positive predictive value of a positive fibronectin test on that patient is only 12%. So the question is does changing her risk of preterm delivery from 3% to 12%? Does that have any clinical value in how you would manage the patient?
Antonia: 35:06
Well, no, and that’s probably why we were mostly taught to ignore the test if it was positive and say, well, this doesn’t really inform us. But if the test was negative, then that we could be really assured. By that, even though that’s already not what it seemed like, we already thought she had a low chance of interim delivery Sure.
Howard: 35:27
Well, but a lot of people don’t ignore the test and what happened over the years is that physicians would act on that test and use a positive test as a reason to do some intervention, maybe give steroids or admit for prolonged observation or even for a while administer antibiotics like a Z-Pak or something like that. People were coming up with things to do and of course, any intervention you might do for that patient will seem very effective when the vast majority of patients like her are not in preterm labor and they won’t continue to dilate, and whatever you did seems to work.
Antonia: 36:01
Well, let’s agree that that positive test is meaningless and we should have ignored it, but what about the negative predictive value Like why is that also not valuable?
Howard: 36:12
Yeah. So if we take the same 3% risk, the same patient we just mentioned, then the negative predictive value for that same circumstance is 99.3%. So again the question is do you feel better about being 99% certain that the patient’s not going to deliver in the next week or so, compared to 97% certain?
Antonia: 36:33
Yeah, that doesn’t make any difference clinically and I can actually think of another reason where that could trip people up. So let’s say, conversely, someone comes in with a high risk of preterm delivery, so just for example, maybe someone that has twins and has had a prior preterm delivery, and then we get an FFN and it’s negative. Well, that’s actually not that helpful either because it doesn’t bring the risk down enough necessarily that you could be reassured. And I think that’s another common mix up because people will interpret it in the opposite way and think, wow, if someone has all those risk factors and still has a negative FFN, then we’re really doing good, Almost like that’s somehow a more dramatic change or somehow better than the low risk patient that had the negative FFN swab. But it’s completely the opposite If you actually do the calculations.
Antonia: 37:27
Let’s pretend, that patient I came up with has a 60% chance of preterm labor. Getting that negative FFN with those specificity and sensitivity numbers would bring that down to about 30%, which is still way too high to just send them home and do nothing. They’re still at least going to need observation. So I wonder, is there some kind of narrow risk range not quite super high risk but not quite super low risk, where maybe that FFN result actually would make the difference between sending home versus admitting or giving steroids, for example.
Howard: 38:03
Yeah, and that’s what tests are supposed to be for, right, those indeterminate probabilities. But the problem with any pre-test probability that you pick and throw into the calculator, the post-test probability isn’t moved that far. If she had a 30% chance before the test, now she has a 24% chance after a negative test. So are you going to do anything different with those two numbers? Does that change the degree of certainty that you have? Remove the needle enough for you that you’re going to do something different and act differently. You can try any number and the change will always be something underwhelming like that. So does it change what you do?
Antonia: 38:37
Yeah, you really need a test that really changes the risk into a different category, and it doesn’t sound like this. Does that ever?
Howard: 38:44
Right. So if you understand Bayes’ theorem and you understand and calculate positive and negative predictive values given this test’s performance, and run through these scenarios like this and think about them, you’ll see pretty quickly that there’s really no clinical utility in ever relying upon a fetal fibronectin and therefore in ever getting one. It shouldn’t ever change your management. But people do construct ways of making it change, like let me give them a Z-Pak so they try to integrate it, or maybe they integrate it with other information, like ultrasound, cervical links, for example, and that’s okay as an avenue of research to see how we can do with that. But right now it’s not evidence-based for prime time.
Antonia: 39:22
Yeah, and that is actually what ACOG says in practice bulletin 171 from 2016, that either FFN alone or in combination with cervical length measurement has not been shown by randomized controlled trials to provide any useful information for clinically managing patients.
Antonia: 39:41
And that feels very ironic to me because that very same year I think just before that came out I was an intern and I had put a ton of energy into doing a little quality improvement study on reducing patients’ triage wait times by having a preterm labor algorithm that brought in FFN for certain patients.
Antonia: 40:05
If I remember correctly, the first decision point was dilation and then, if they were not dilated, then they had their cervical length measured at the bedside and then, if that was in a certain indeterminate range, then we also sent the FFN that we had collected. Initially I was doing triage all the time I was an intern, so this was my life and of course, this was in a setting with 24 seven in-house OB coverage. There were multiple MD or DO OBs sitting right there 24 seven, including the little interns like me who would be the ones actually doing these exams and, in my case, collecting data and reporting it and everything. These exams and, in my case, collecting data and reporting it and everything. So it would be totally moot for somebody taking a home call, like I am now often driving up from 30 minutes away, to do this assessment, as opposed to them just getting a repeat cervical exam, as was the alternative to doing all of this.
Howard: 41:02
Well, the spirit of what of the intent I understand completely.
Howard: 41:05
If I recall, you also were just overwhelmed with triage people lining up to be assessed.
Howard: 41:11
So obviously finding a way to cut down wait times and get beds available and things like that is a very practical concern.
Howard: 41:18
So you keep it simple and let patients presenting with preterm contractions be monitored and rechecked after some period of time, let’s say two hours. Or you could use a more complicated decision tree and send the intern in with a speculum and a fetal fibronectin and an ultrasound, and we still have to get that test back and all those things and end up with the patient either discharged to home or admitted out of the triage room. And maybe you can cut that down really quickly to less than 30 minutes. And sometimes you just have to make decisions without complete certainty about what the outcome is and you obviously can’t admit everyone who’s contracting or who has more than a zero or some percent chance of something happening. And obviously, if you’d known that there was about to be new guidance published about how little the value of a fetal fibronectin combined with ultrasound is, you might not have committed to that in your project, but this is still happening all across the country for the same reasons.
Antonia: 42:11
Yeah, sometimes it might be helpful to get that cervical length, but even that isn’t necessary if you can just recheck them in most cases, and I think it probably was a good thing for me to experience something that I was practicing to be completely invalidated and canceled out by new guidance or evidence, because it certainly wasn’t the last time in my life that that was going to happen.
Howard: 42:34
Yeah, I’ve been through three or four generations of that so far, so that’s why we do the podcast. But at least there was subsequent literature to really help drill in that fetal fibronectin shouldn’t be used. There was a trial by Peaceman that showed that if you had a negative test, you could estimate a 99.5% chance that the patient wouldn’t deliver within seven days. But they also concluded that there was no clinical advantage in knowing that compared to using other clinical criteria, including just history and physical exam. A 2016 meta-analysis by Bergella concluded that testing among women with threatened preterm labor wasn’t associated with prevention of preterm birth, because we don’t have anything to do to prevent it or any other improvements in perinatal outcomes, and was just associated with higher cost.
Antonia: 43:17
Yeah, definitely not a good screen and also really not a good diagnostic test. So another seminal trial that maybe we should have discussed before discussing ARRIVE trial is the post term pregnancy trial and this was essentially like ARRIVE, except they used 41 weeks instead of 39 as their intervention cut off. This was published by Hannah and colleagues in 1992 in the New England Journal of Medicine and it was done in Canada in 22 hospitals over a seven-year period. So these women were randomized at 41 weeks either to getting induced right then or to continue expectant management until 42 weeks, and the patients in the expectant management group did get extra antenatal testing like how we do now. If someone is not getting induced at 41. We have them get NSTs or biophysical profiles or something.
Antonia: 44:11
So in this study they found that 24.5% of the women in the expected management group ended up with a cesarean, compared to only 21.2% in the 41-week induction group. 2% in the 41-week induction group, the excess cesareans, that’s about three percentage points. Those seemed to be mostly due to fetal distress that became present during that extra week. They did not find differences overall in perinatal morbidity or mortality but there were two still births in the expected management group. But they didn’t comment on this as being statistically significant, because they were not powered to find a true difference in the stillbirth rate.
Howard: 44:53
Yeah. So at the time this study was published it was definitely common for patients to go to 42 weeks and although we already knew that perinatal morbidity and mortality increased after 41 weeks that perinatal morbidity and mortality increased after 41 weeks, we were also concerned that earlier induction of patients, particularly those with an unfavorable cervix, would increase the cesarean rate and increase maternal morbidity and mortality. So this study was performed at a time when the total cesarean rate was around 25% or less and we were very conscientious of increasing maternal mortality by increasing the cesarean delivery rate. So we routinely allowed, and even recommended, that women should progress with monitoring to 42 weeks, because we believed it reduced maternal morbidity and mortality, even though the rates of perinatal complications might be a tad higher. But they used a prostaglandin gel for induction, and so this study was the first to look at earlier induction with rather novel cervical ripening agents.
Antonia: 45:49
Yeah, and now we have even better cervical ripening agents than what they did at that time. So if we were to repeat that trial now, we might expect an even lower cesarean rate in the 41-week group if everything else were to be held the same.
Howard: 46:05
Yeah, and this study is a lot like the ARRIVE trial in the sense that they were hoping to find a reduced rate of perinatal morbidity and mortality without excess negative maternal outcomes. But also like the ARRIVE trial, they found that maybe there was a lower rate of cesarean, but in this trial it was because the pregnancies that continued on, with significant placental senescence occurring from 41 to 42 weeks, did have more meconium and more distress and they did have fetal deaths despite any extra-anilatal testing. So this trial is important to know because the authors have arrived basically tried to do the same thing but pushed it from 41 to 39 weeks. But it took a while for 41-week induction to become more standard than 42-week induction and then later studies did show that the stillbirth rates declined in the US and Canada after the adoption of the 41-week induction rule. So this trial was the beginning of a lot of evidence that really did change management.
Antonia: 47:02
Yeah, and you can’t ethically say we’re going to make a study, that we’re going to power it to find the difference in stillbirths, so that you could truly say that. So I think this is one case where we look at those two extra stillbirths, even though they weren’t statistically significant, and we consider them.
Howard: 47:19
But more important, we now have population level data after the change. That shows that this helped.
Antonia: 47:24
Yeah, and let’s just review a little terminology about term and post-term and all of that. So just to give the whole spectrum, ACOG says early preterm is before 34 weeks. Late preterm is up, 34 to 37. Then 37 is early term, 39 is term, so that’s when we do our elective inductions et cetera. 41 is a late term and 42 plus is post-term. And each of those categories has a different descriptor and a different name because the prognosis and management are different at each of those time points. They’re not just randomly called different names for no reason.
Antonia: 48:06
So ideally we want to avoid all post-term pregnancies. We don’t want to get to 42 weeks and beyond for anyone, if we can help it. Basically we know that the stillbirth risk just keeps increasing. Every extra week that someone is pregnant is another week that the baby could just pass away in utero. That someone is pregnant is another week that the baby could just pass away in utero. So at 37 weeks it’s very low. It’s about 0.4 per thousand and it just keeps going up. Once you get to 43 weeks you are at least at 11 per thousand. So over 1%, which already is unacceptably high, and by some estimates it actually goes as high as 40 per thousand, so 4%, just for being at 43 weeks or really even as early as 41 weeks. The ranges are a broad range. By that logic, we will generally recommend some intervention at 41 weeks, whether it’s induction or cesarean or just increased fetal monitoring.
Howard: 49:08
Yeah, patients have to think of going past 41 in the same way that we think of preterm labor, like it happens. But it’s not supposed to. And unfortunately a lot of people think that going beyond 41 is like a really healthy I’m getting extra cooked or whatever. But it’s just as pathologic to go late as it is to go early.
Antonia: 49:25
Yeah, I think people don’t typically look at it that way. They just they’re like I want to wait till my body’s ready. Sometimes I’ve read of people going really far, like 44, 45 weeks, and not having good outcomes there.
Howard: 49:39
Yeah, or some get lucky and validate the rest.
Antonia: 49:41
True. So the ACOG bulletin on management of late-term and post-term pregnancies actually states that there’s no data from randomized controlled trials showing benefit to the babies from fetal surveillance in these 41-plus-week pregnancies, and so that’s an important thing to tell patients who decline induction because they may get a false sense of security from getting that antenatal testing and we can talk about that in another episode maybe, but most antenatal testing that’s recommended for any reason has not been shown to improve outcomes. I find that some patients think that, as long as the testing is okay, it’s completely safe to go up to 42 weeks or beyond, however long they want to go, that they would keep getting testing for. But that’s never been demonstrated. And of course in the study we just talked about there were still those two stillbirths, even though they were getting the testing and presumably the testing was normal. But we should be able to say that induction at 41 weeks will lower the patient’s risk of cesarean and ultimately will probably benefit the babies too.
Howard: 50:54
I also think people are confused by the way that the ACOG practice bulletin is written, because a lot of people interpret it as saying that patients should be recommended to have an induction after 42 weeks, but consider merely consider induction at 41 to 42.
Howard: 51:06
And that’s what it says, but that’s not necessarily what it means.
Howard: 51:10
I think a lot of ACOG bulletins are written in a way to protect obstetricians from lawsuits and things like that, and you have to understand that there are some allowances in there but also an expectation that you know what the right thing to do is.
Howard: 51:22
What it says is that induction of labor can be considered between 41 and 42 after it’s just spent several paragraphs telling you why induction at 41 weeks is safer for the mother and the baby. But the practical reality of scheduling is that a patient might be 41 and 0 on a Saturday and given your hospital and staffing situation it might not be feasible to do it until Monday or even Tuesday. She might get bumped because of another high-risk priority induction maybe a newly diagnosed preeclamptic or something with severe features sent from clinic and by the time you actually deliver she’s 41 and 2 or 41 and 3. So the bulletin is written in a way to protect the members of ACOG we physicians from unnecessary legal exposure. But the intent of it is for you to deliver these patients as close as possible to 41 weeks and to not let patients go beyond 42 weeks. If you can help it.
Antonia: 52:12
Yeah, ACOG. Of course it’s looking out for the patients, but it is looking out for us providers too. Their publications have to toe the line between being directive like this is the standard of care and this is what you should do but also still minimizing trouble for obstetricians if they find themselves in court after a bad outcome. Generally, despite doing everything right, Sometimes that can still happen, and ACOG understands pragmatic things like staffing and bed availability that affect our practice in the day-to-day sense, and so they want to allow for some flexibility there, just so that we can reasonably accommodate these recommendations. So you have to read between the lines of the ACOG practice bulletins. You read the lines and then read between the lines and then put them in context when you’re deciding with your patient what is the right thing to do for each one of them.
Howard: 53:05
Yeah, we should do an episode sometime about what the bulletins say versus what they actually mean, to demonstrate to us and understand the sometimes the dichotomous ways in which these bulletins can be read. To support a variety of practices, you can find underlying meanings that are not explicitly stated but easily deduced in almost everything that they put out.
Antonia: 53:25
Yeah, that would be a really interesting discussion. So back to the 41 week deliveries. We really should be inducing even low risk people, but especially people with any risk factors, by 41 weeks, unless they flat out refuse. But we should be giving them that recommendation and doing it if they agree. And in those cases when we do extra testing for someone that has declined to be induced, then they need to be made aware that the testing is not a guarantee that their baby will be okay, because it has not been shown to decrease the risk of late-term stillbirth and I think doing the testing sometimes can help make the case for induction if the testing is abnormal.
Antonia: 54:10
So if they’re not quite buying it that just because I’m 41 weeks you’re saying I need to be induced but my body isn’t laboring yet, Well, now let’s say you’ve done a biophysical profile and she has all ago. And if she’s been resistant to induction up until then, that might add a little bit more motivation or fuel to the fire, so to speak, as to why she really should be induced now. She already should have been, but especially now. But again, on the population level, that testing doesn’t actually help anything and we could talk about the original studies about biophysical profile and antenatal testing as well. Probably should save that for another episode, though, and maybe we can also do an episode like this later that focuses on landmark studies in gynecology, since all the ones we have just done now were just obstetrics related, so do you have any other bits for today, if we even have time for that?
Howard: 55:05
Well, at the end of last season we got out of order because we did two episodes in a row with listener questions, so we owe the listeners a historical segment to catch up.
Antonia: 55:12
All right. Well, let’s do the historical segment. So what do you have?
Howard: 55:16
Well, we talked about magpie today and the use of magnesium sulfate for the prevention of maternal seizures in eclamptics, and we referenced our lengthy discussion elsewhere about MAG for prevention of cerebral palsy. At one point magnesium was also used as a tocolytic, of course without scientific evidence, but its first use was for the treatment of eclampsia. Obstetricians became very comfortable with using magnesium for that indication and then started thinking it might be useful for arresting labor. Since it seemed to have an effect on reducing contractility in patients, they were using it on. So then, after it had been used as a tocolytic for a while, again without any clinical outcomes or studies supporting that use, people started to wonder about the neonatal outcomes and then created the current controversy about prevention of cerebral palsy. But I thought I would talk about how we came to use it for eclamptic seizure in the first place.
Antonia: 56:07
Okay, yeah, so I guess that was the beginning of the snowball. It does seem random that, of all the things in the world, someone thought to use magnesium to treat seizures.
Howard: 56:18
Yeah, definitely. Well, originally preeclampsia and eclampsia were treated with things like bloodletting to release toxin buildup. Oh yeah, like bloodletting to release toxin buildup.
Antonia: 56:25
Oh yeah, bloodletting was the treatment for everything at one point. So yeah, but they used to call preeclampsia toxemia.
Howard: 56:38
Right, yeah, and of course bloodletting didn’t work. But in 1916, joseph Dilley, the noted Chicago obstetrician, observed that maternal mortality with eclampsia at that year was about 45% and the treatment at that time was just sedation and trying to get them delivered.
Antonia: 56:50
Well, you’d think sedation and delivery would be easy enough. Except in 1916, they didn’t even have oxytocin and cesarean. Delivery wasn’t safe, and I doubt that their sedation methods were especially safe either.
Howard: 57:02
Right, yeah, so we’re in the pre-antibiotic era here, and there was some controversy about proceeding with rapid delivery because, I guess one there weren’t many good options, but also because some doctors thought that labor itself could stimulate seizures, which of course is a false concern. But the benefit of rapid delivery for eclampsia was shown by Durson and Brom.
Howard: 57:25
That’s the same Durhsen as the Durhsen incision of the cervix for an entrapped head, don’t get started with eponyms again, but yeah, so originally in 1906, a doc called Horn injected magnesium intrathecally to prevent eclamptic seizures. Intrathecal stuff had just come about in the last decade and people soon figured out that you could also just give it intramuscularly and this was used in regular epileptic seizures as well as to control seizures during this era. But in 1925, Lazard in Los Angeles, a very notable obstetrician of the era, started injecting it intravenously and gave it to over 500 women with preeclampsia and eclampsia over an eight-year time period. Now the doses they used then were smaller than the doses we use today, but folks started testing higher doses. And then Jack Pritchard at Parkland is generally credited with popularizing the intramuscular doses that we use today, and intramuscular magnesium sulfate was used for decades at Parkland, even when most everybody else had gone to intravenous use because of Pritchard’s legacy.
Antonia: 58:31
That’s interesting, but at this point they were still trying to proceed with the rapid delivery. Right If someone had a seizure? This wasn’t part of some push to delay delivery and just let these eclamptics sit there and get mag for days or weeks on end.
Howard: 58:45
Right. If anything, this helped reassure doctors who were afraid of stimulating seizures with an induction or something like that that they could go ahead and do it because they have this medication to reduce seizures. Even today you still hear people talk about stabilizing the patient on magnesium. This is a hangover from that, but undoubtedly there were enough people not delivering eclamptic mothers as quickly as they should because of those concerns and that did lead to still high mortality rates of eclampsia. In the 1940s Eastman, who was an editor of Williams, like Pritchard was later had started experimenting with the larger doses of magnesium and, along with Philip Steptoe in Baltimore at Johns Hopkins where they were, Williams moved from Hopkins to Parkland. So it started at Hopkins and then Eastman had it and then eventually it moved to Parkland where Pritchard was located, and then he was the editor with co-editors for several decades as well.
Howard: 59:39
Eastman and Steptoe published a paper together in 1945 in the Canadian Medical Association Journal and I’ll put a link to that and they reported their experience with 2,418 cases of preeclampsia and 3.8% of those patients developed eclampsia Of the eclamptics, 8% of the mothers died, 22% of the newborns died and most of the women had an average of 4.3 seizures. That compared to the other preeclamptics who didn’t seize of only having a 0.2% maternal mortality rate and a 7% infant mortality rate. He also noted that, although preeclampsia was equally common in white and black women, twice as many black women developed eclampsia, making the disease overall more fatal for black patients. They detailed a lot of the lessons learned about how to diagnose preeclampsia and understanding indications for delivery over their 20-year experience at Hopkins, but, importantly, they advocated for larger doses of magnesium and repeated doses over a lengthy induction. They also detailed when magnesium toxicity typically occurs with these higher doses, and this was really the forerunner of Jack Pritchard’s work.
Antonia: 1:00:47
And Jack Pritchard’s work magnesium for eclampsia. That was in the 1960s, so how did we get from that to tocolysis?
Howard: 1:00:56
Yeah, well, by the time Pritchard had standardized things around magnesium, he had no maternal deaths in his published series of patients. So this was a big contrast to the decades before. And Dr Zusman, another very famous name in obstetrics, gave us the four gram IV load followed by one gram per hour. That’s changed a little bit, obviously, but during this time people noted that when you had women on magnesium it seemed like their contractions were slower. Of course you’re trying to get these ladies induced and things like that, so you’re paying careful attention to that, and some of that even might be the concern of having to use this newfangled drug and I’m trying to get her delivered. You can hear people saying it.
Howard: 1:01:34
Well anyway, hall published this observation in 1959, as well as another doctor called Hutchison in 1964. And there were some in vitro experiments in the 1960s in Germany and in Japan and the US that showed that uterine contractility was slowed in the presence of magnesium. By 1965 in France Dr Lamont first experimented with it as a tocolytic and then in 1969, it was being used as a tocolytic at Columbia Presbyterian in New York without any controlled studies at that point. It also became used for tocolysis at the University of Virginia in 1970. And then we started to see these flawed studies that almost all of which compared it to the then standard tocolytic, which was alcohol starting to be published in the late 1970s. But that’s for a different episode.
Antonia: 1:02:21
Okay, and then, just to finish connecting those dots, when the safety of magnesium for fetuses and neonates was finally tracked in some of the magnesium tocolysis studies, particularly in the exposed preterm infants, then there appeared to be, at least in earlier studies, excessive perinatal mortality.
Howard: 1:02:57
And this led to larger studies to see was that actually a true finding? And then we ended up with the possible conclusion that exposure to magnesium might, if anything, reduce the, because it may come reduce CP at the risk of trading mortality for better survivorship. But you can listen to the other episode for your exploration and draw your own conclusions about that data.
Antonia: 1:03:13
All right, well, we should wrap up for today. The Thinking About OBGYN website will have links to studies that we discussed, and check out and follow our Instagram. There’s some nice little kind of visual summaries of things we talked about too, and then we’re going to continue on in a couple of weeks.
Announcer: 1:03:35
Thanks for listening. Find us online at thinkingaboutobgyn.com. Be sure to subscribe. Look for new episodes every two weeks.
