Literature on the management of term breech pregnancy
including the RCOG 2017 guideline update
North of England Breech Conference, Sheffield
Lawrence Impey, FRCOG, is a Consultant in Obstetrics & Fetal Medicine at John Radcliffe Hospital, Oxford. He is the head of the guideline group responsible for the new RCOG Breech guideline, published on March 16th 2017. He is widely published in the areas of labor, breech presentation and the origins of childhood handicap. He has written, edited and contributed to several leading text books in obstetrics and gynecology.
Mr Impey is Director of the sub-specialty training program in fetal and maternal medicine in Oxford. He also lectures regularly locally at Oxford University, nationally and internationally. Within the NHS, Mr Impey specializes in complex pregnancy, particularly for women with multiple pregnancies and those at risk of preterm delivery. He runs the breech and external cephalic version (ECV) service.
Disclaimer: As with all of the conference summaries that I write, I do my best to provide a detailed summary of each speaker. I am working from typed notes, not from recordings, so my summaries are more of an extended paraphrase and less of a word-for-word transcription. If something I have written is not accurate, please contact me so I can make the appropriate changes.
Lawrence Impey opened by nothing how he enjoyed the conference that took place in Sheffield two years ago. Today he talked about the recently updated 2017 RCOG Green-top Guideline for Mangement of Breech Presentation (PDF here) and about the evidence on breech birth. It’s very odd that we dispute evidence or facts when facts are really facts. [This elicited ironic laughter from the audience--who knows better than midwives and obstetricians how disputed facts can be?]
When he and his team put together the new guideline, it was very obvious what the evidence showed. He’s slightly puzzled about why there’s been any dispute about it at all. What is important is that we give the facts in a proper, fair, and unbiased way. He hopes that his "dry" talk today will do this. What women do with this information is what matters. As OBs, we forget the difference between relative versus absolute risk, and we focus too much on the immediate risks versus the later risks.
The 2017 RCOG guideline created more of a stir than he had anticipated. He shared several rather scathing critiques of the new guidelines [from people who obviously were not very comfortable with vaginal breech birth]. Here are some of the criticisms:
- "In my opinion a biased look at the literature"
- "This recommendation is based on poorly designed studies and is designed to scare women into wanting a vaginal breech delivery"
- "We know many (if not most) of these women will require some kind of operative intervention in order to deliver (CS, forceps, generous episiotomy, fetal manipulation or even cervical incisions)"
- "...and we believe that the recommendation...is detrimental to maternal safety"
Another person found fault for different reasons:
- "this section does not respect the right of the woman to make informed decisions about place of birth"
Lawrence presented several key questions in the debate over breech birth: Is cesarean “safer” for a baby? And safer in what way? Is planned VBB safer for a cephalic baby than for a breech one?
In debates over safety, many different outcomes matter. Some are short-term, others long-term. Lawrence's presentation covered six main outcomes, all of which need to be taken into account.
- This baby: death (stillbirth, infant mortality)
- This baby: morbidity
- The mother this time
- The next baby
- The mother next time
- Other people’s babies: this is something that we easily forget. What we do with breech affects other mothers and other babies both locally and globally.
This baby: Death
The longest section of Lawrence's presentation focused on neonatal & perinatal death. He noted that the longer you drive, the more likely you are to be killed on the motorway. The same is true for pregnancy: the longer you are pregnant, the more likely you will have a stillbirth. If you simply end the pregnancy at 39 weeks, you aren’t exposing any babies to the risk of stillbirth after 39 weeks. (Kind of obvious!) If you “fail” to deliver a baby at 39 weeks, you introduce that baby to a 1/1000 risk of stillbirth.
Lawrence noted that if the only outcome you’re interested in is the prevention of stillbirth—and nothing else—then it is slightly safer for all babies to have a CS at 39 weeks than to wait for labor to begin. The risk of perinatal death for planned cephalic births compared to elective CS is around 1.3/1000 (1/1000 from failing to deliver at 39 weeks + 0.3/1000 intrapartum risk).
Arguments about term breech should be set in this context. Any trial that randomizes babies to elective cesarean section versus labor will expose the baby in the labor group to a slightly higher excess perinatal mortality by nature of waiting for labor to begin and the process of labor itself. This matters because there are other outcomes to take into account, not just perinatal mortality.
Lawrence then addressed the 2000 Term Breech Trial by Hannah et al (TBT). This randomized controlled trial suggested that 1% of babies died because of planned vaginal breech birth. Since it's the only RCT on term breech, the TBT is seen as the definitive answer. However, there are both strengths and weaknesses--"angels and demons" in his words--in this trial. He did not spend too long on them since the literature has already played them out in great detail. But in his view, one of the TBT's major problems was that some of the babies were randomized incorrectly. And the findings of the TBT aren't applicable to a modern setting in many ways; 31% of the babies had no prenatal ultrasound, 13% had no obstetrician in attendance, fetal monitoring was rare, prolonged second stages were allowed, and 10.6% of the babies fell outside the weight range of 2500-4000 g. By Glezerman's analysis, only 16 of the 69 "morbid" babies had outcomes attributable to the mode of delivery.
Lawrence then addressed a number of other influential studies that have influenced the debate over perinatal mortality (PNM) and neonatal mortality (NNM) with regards to breech:
- Meta-analysis by Berhan & Haileamlak (2016). This study suggseted a PNMR of 3/1000 for pVBB vs 0.5/1000 for ECS. However, it was retrospective, heterogenous, and had unstated management and case selection.
- PREMODA study by Goffinet et al (2006): This multi-center French & Belgian study is a very important series comparing 2502 pVBB and 5573 pCS. It wasn't a RCT, and the two planned arms (VBB vs ECS) aren't directly comparable. The PNMR with planned VBB was 1.2/1000 compared to 1.4/1000 with planned cesarean. To him, this study describes one very important thing: what can be achieved in terms of the safety of VBB. Those breeches were "managed" in a very different way than most people would look after breech births in the UK.
- Vlemmix et al (2014): This population-based study comes from Dutch registry data and looked at 58,320 term breeches from 1997-2007. It is impressive data, and we need to look at it very carefully. All babies were alive at the start of labor, so it eliminated the antepartum stillbirth issue. The important point is the PNMR of 1.6/1000 for pVBB in this study vs the 1/100 quoted in the TBT.
Overall, there is a difference in PNMR between ECS and planned vaginal birth. However, the TBT has exaggerated the risk of PNM by a factor of 5-10. This influenced his Grade C recommendation in the 2017 RCOG guidelines that put the risk of perinatal morality at 0.5/1000 with CS at 39 weeks, 2/1000 with pVBB, and 1/1000 with planned cephalic birth.
This bears repeating: planned vaginal birth for a head-down baby is twice as dangerous as elective cesarean at 39 weeks--looking only at perinatal mortality--but there is no guideline telling women with head-down babies that vaginal birth is too dangerous! We have to remember this when looking at breech data.
Thus, the 2017 guidelines counseled that "Clinicians should counsel women in an unbiased way that ensures a proper understanding of the absolute as well as relative risks of their different options." Just saying that vaginal breech birth is "four times more dangerous" than an elective cesarean (relative risk) is misleading and doesn't put the numbers in context of their absolute risk.
This raises some important questions: Can the mortality rate be further reduced with VBB? Can we get it down to the cephalic mortality rate? One possible solution comes through better in-labor management; another comes from better case selection. When looking at small studies that are underpowered to detect mortality, it is reasonable to assume that if the morbidity rate is better, the mortality rate may also be better. This guided his level C recommendation that "Selection of appropriate pregnancies and skilled intrapartum care may allow planned vaginal breech birth to be nearly as safe as planned vaginal cephalic birth."
This baby: Morbidity
Most studies of vaginal breech births show a higher short-term morbidity (usually measured by Apgar scores, NICU stays, and trauma). However, Lawrence noted an "enormous amount of hypocrisy in obstetrics" in how morbidity is interpreted with breech versus VBAC or EFM. For example, VBAC has a 0.08% risk of HIE and 0.04% rate of PNM, yet the RCOG 2015 VBAC guidelines state that "Women considering a vaginal birth after a previous caesarean section can be assured that it is a clinically safe choice for the majority of women." Regarding EFM, neonatal seizures are twice as common with intermittent auscultation than with EFM for low-risk women (Alfirevicz et al 2017), yet intermittent monitoring is still commonplace and accepted for low-risk women.
In other words, obstetrics picks and chooses what risks to downplay and what risks to emphasize, depending on the issue at hand.
The mother this time
Maternal morbidity should also matter to us. It's highly dependent on the emergency cesarean rate. In the TBT, 36.1% of mothers had in-labor cesareans, compared with 29% of women in the PREMODA study and ~45% in the Vlemmix study. A mom with a 95% chance of needing a CS in labor would be best counseled to plan a CS than a mom with a very low chance of needing an in-labor CS. And the chance of an emergency CS is relatively high for a pVBB. Even taking that into account, though, it’s still slightly safer for a mother to plan a VBB than to plan a cesarean for a breech baby.
The next baby
The next baby is also really, really important, especially in other parts of the world where women have large families. There is some data that suggests a two-fold increase in stillbirth for women with a previous CS. That may or not be right; it’s probably more due to the indication for the original CS than the surgery itself.
In the UK, over 50% of women with a previous CS have a repeat CS. Repeated surgeries increase the risk to the mother and to the baby.
We have a really serious responsibility to set an example for other parts of the world. A colleague in Africa told him that for every 1 baby you save by doing a CS for breech, you will kill 5-10 mothers down the road. Just having had a CS makes women much more likely to have other obstetrical problems.
There are also long-term risks to the baby born by cesarean sections, including increased rates of obesity (Darmasseelane et al 2014, Huh et al 2012), type-I diabetes (Cardwell et al 2008), asthma (Huang et al 2015, Black et al 2016), and chronic immune disorders (Sevelsted et al 2015).
Larwrence warned that we are obsessed with the immediate over the long-term, both as parents and as health professionals. We have an obligation not to wreck things in the long-term and not to create disease. It’s very important to bear this in mind and to understand that cesarean section is not a benign procedure.
And there are confounding variables to sort out: how much of the risk comes from the indication versus the cesarean itself?
The mother next time
For a mother with a cesarean scar, there are several increased risks regardless of whether she has a VBAC or a repeat CS for her next baby: blood transfusion, endometriosis, hysterectomy, death, and placenta accreta/percreta.
Other people's babies
Lawrence sometimes jokes with total sincerity: “I want you to have a vaginal breech birth because I need the practice.” Other people's babies are tremendously important, and needing practice is a very reasonable reason to promote vaginal breech birth. In the TBT, low PNMR countries had twice the morbidity during planned VBB than high PNMR countries. Why is this? He theorizes it's because providers in low PNMR countries were less skilled; they had less practice because cesarean section for breech had already become common even before the TBT.
In the UK today, 1/3 of all term breeches are still undiagnosed when labor begins. Until relatively recently, Oxford had a high rate of undiagnosed breech birth. In fact, last year his unit lost 3 undiagnosed breeches who died on the highway en route to the hospital. Providers today don’t know what to do for undiagnosed breech babies. In Oxford, half of them are not diagnosed until full dilation. We need to be able to do a vaginal breech for those 1/3 of all breech babies. This argument is a very, very strong one, Lawrence noted.
Lawrence's presentation raised several important questions about how planned VBB compares to cephalic birth:
- Is it as safe?
- Can it be made as safe?
- How can it be made safer?
This conference should strive to answer those 3 questions.
When looking at breech birth, many outcomes matter, not just perinatal mortality. Short-and long term outcomes matter for both mother and baby. The next baby matters. Other people's babies matter.
One audience member mentioned how we calculate stillbirth at different weeks of gestation might affect how we counsel women. She mentioned a different method for calculating stillbirths, taking into account the babies already born, and it makes it more of a flat line vs a sharply increasing line after 40 weeks.
Another audience member asked a question about ECV.
A third person asked whether the cesarean rate for breech had changed after Oxford started doing routine 37 week anatomy scans (which has greatly reduced their undiagnosed breech rate).