Several people have contacted me regarding a VBAC study recently published out of Canada. Many of the headlines about this study focused on two findings: “Absolute rates of severe maternal morbidity and mortality were low but significantly higher after attempted vaginal birth after cesarean delivery compared with elective repeat cesarean delivery” and “Failed vaginal birth after cesarean delivery was associated with a threefold higher rate of neonatal death.” What does this mean? Is VBAC unsafe? Should birthing families plan elective repeat cesarean sections?

I dove into the study and quickly saw how it could be misinterpreted by many. This is an example of how the findings of a single study can be communicated to the public in a few ways, including to stoke the existing fear surrounding VBAC. Studies like this can also be used by anti-VBAC providers to coerce people into repeat cesareans. After all, the study concluded that the risks of planned VBAC were “significantly higher.” So what I’d like to do is walk you through my process when evaluating medical studies. I’ll share the questions I typically ask when reading medical research as well as answer those questions for this specific study. Let’s get started.

Let’s ditch the judgement laden language

As I really dislike the terms “successful” versus “failed” VBAC, I’m going to use my own terminology here and I hope you adopt it:

• Rather than “attempted VBAC,” I’m going to use “labor after cesarean” (LAC) or planned VBAC.
• Rather than “failed VBAC,” I’m going to use “cesarean birth after cesarean” (CBAC).
• Rather than “successful VBAC,” I’m going to use just VBAC.

So, to recap, someone can plan a VBAC or labor after a cesarean and they will either have a CBAC or a VBAC. Doesn’t that language feel better?

This slight pivot in how we describe birth after cesarean honors the journey of the birthing person, regardless of how their labor plays out.

There is power and agency in planning a VBAC. Let’s embrace that and own it and then see how labor plays out.

This language is also especially sensitive to those who did plan a VBAC and then had a repeat cesarean. Using the term “failed VBAC” is so hurtful and it stings my heart to hear parents use this language to refer to themselves!

Judgement has no place in health care. Let’s change the language we use to reflect a more empathetic approach to childbirth.

Questions to ask of any VBAC study

Whenever I read VBAC medical studies, there are 10 primary questions that I ask:

1. Who was in the study?
2. How was the study conducted?
3. Did they control for scar type?
4. Did they control for the number of prior cesareans?
5. How many had spontaneous labors?
6. Did they report outcomes among spontaneous labors?
7. Did they report outcomes by intended or actual mode of delivery?
8. Did they distinguish uterine rupture from uterine dehiscence?
9. What were the maternal and neonatal outcomes?
10. What was the absolute versus relative risk of those outcomes?

These questions enable us to be very clear on what was measured, using what definitions, among what population and then how did they group various outcomes for reporting purposes? Finally, how did the relative and absolute risk compare between VBAC and repeat cesarean for individual outcomes as well as composite measures (where they group individual outcomes together.)

Who was in the study?

This is a very large study including 197,540 people with one prior cesarean and no prior vaginal deliveries, who were pregnant with one baby and delivered at term. Data was pulled from all hospital deliveries in Canada (excluding Quebec) between April 2003 and March 2015.

How was the study conducted?

There are many different ways medical studies can be conducted. Randomized controlled trials are the gold standard of medical research. We don’t see those a lot when it comes to VBAC primarily because people typically have strong feelings about how their babies are born. They don’t want to be randomly assigned to either planned VBAC or elective repeat cesarean. So this leaves us with studies examining how people want to birth or did birth.

Research that enrolls people and measure their outcomes, called prospective studies, are some of the strongest. There are also retrospective studies where the births already occurred and we look back in time to collect that data and then report the findings.

This study was a retrospective study examining the data from the Discharge Abstract Database of the Canadian Institute for Health Information. Huge databases like this have their limitations including what information is collected and coding errors. Specific data we would want to evaluate in VBAC studies might not be included in a national database.

Did they control for scar type?

Controlling for scar type means that they tracked what type of scar the participants had and then reported outcomes, like uterine rupture, for each scar type.

One of the variables this national database did not capture is what type of uterine scar each participant had. It’s unusual for a VBAC study to not mention scar type at all, but this is an example of the type of nuanced information we need in VBAC studies in order to make their findings useful. So it’s unclear if they just included low transverse scars or they included anyone with a prior cesarean. From how the study reads, I suspect it’s the latter.

That could impact outcomes as conventional wisdom dictates that classical uterine scars – which occur within the upper part of the uterus – have higher uterine rupture rates. Although as our reporting as shown, the risk of uterine rupture during a planned vaginal birth after classical cesarean is likely lower than many believe. So, that is an unknown that we need to pack away in our minds. Given that most people have low transverse scars, in the US, it’s about 95%, it’s likely that the overwhelming majority of the people included in this Canadian study had low transverse scars as well.

How many had spontaneous labors?

As we know that as induction increases the risk of uterine rupture, this is a good factor to parse out. Young reports that 3.7% of their study population was induced. That is very low in comparison to other VBAC studies where typically 30 – 60% of the study population are induced or augmented. Which brings up a good question: How many labors were augmented? They don’t report that, either because this national database didn’t capture that information, or perhaps it did, and the study authors choose not to include it.

Either way, this study does not break out any outcomes by induction or augmentation which means we don’t know how outcomes might compare between those who were induced, augmented, or labored spontaneously (without drugs to start or speed up their labor). This is unfortunate because that comparison enables us identify the “baseline” rate of complications like uterine rupture, which induction and augmentation impacts.

We also don’t know how induction or augmented might have impacted other outcomes like postpartum hemorrhage or hysterectomy rates.

Did they report outcomes by intended or actual mode of delivery?

Studies have to decide how they report their findings: by intended or actual mode of delivery. Intended means that outcomes are grouped by how the individual participants planned to birth. So if someone plans an elective repeat cesarean section (ERCS) and then goes into labor and has a VBAC, their birth outcomes go into the “Intended ERCS” category.

Reporting outcomes by actual mode of delivery means that for person who planned an elective repeat cesarean section and had a VBAC, their outcomes would be reported in the VBAC category because that is how they actually delivered.

The benefit of using intended mode of delivery is that it reflects real life, where we make plans, but are unable to control the outcome. As the National Institutes of Health said after its 2010 VBAC Conference, “There is still no evidence to inform patients, clinicians, or policymakers about the outcomes of intended route of delivery because the evidence is based largely on the actual route of delivery.”

But, reporting the outcomes for planned VBACs overall and then breaking out outcomes for those who had a cesarean birth after cesarean versus a VBAC, does give us a little whiff of intended mode of delivery.

This study reports outcomes by actual mode of delivery which is what many studies do. I wish studies reported outcomes by intended and actual mode of delivery because that enables us to look at the data from a couple difference perspectives.

Did they distinguish uterine rupture from uterine dehiscence?

Studies have a choice of how they report uterine rupture rates. Do they want to just focus on uterine rupture or do they want to report uterine dehiscence rates either as its own category or include it under the uterine rupture category? Including uterine dehiscences under the uterine rupture category is problematic because these are two distinct events carrying different levels of risks.

Uterine rupture is fundamentally a full thickness opening through the uterine wall. During a uterine rupture, the fetus is at risk of oxygen deprivation and the birthing parent is at risk for hemorrhage. Whereas uterine dehiscence is a typically asymptomatic separation of the inner layer of the uterus (myometrium) while the outer layer of the uterus (serosa) stays intact that poses no risk to the fetus and birthing parent. Typically uterine dehiscences are not even seen unless someone has a cesarean. Those in vaginal births are rarely noted.

Typically these two events are reported separately, as their symptoms and outcomes are quite different. This is why it’s always important to understand how an individual study defines and reports uterine rupture.

In this study, they combined uterine dehiscence under the label uterine rupture, resulting in an inflated uterine rupture rate. However, as is often the case, that truth is not clear to the reader unless they dive into the belly of the study and read the full text.

What were the maternal and neonatal outcomes?

Uterine rupture

As I read this study, I was initially shocked by their findings on uterine rupture. The overall uterine rupture rate for planned VBAC was 0.99%. This is relatively high for a population with one prior cesarean where only 3.7% were induced.

Note that the rate of uterine rupture rate among elective repeat cesareans is actually higher than the rate among VBACs. It’s important to remember that while elective repeat cesareans decrease the incidence of uterine rupture, they do not eliminate it.

Uterine rupture rates

ERCS: 1.75 per 1000 (0.17%)
LAC: 9.93 per 1000 (0.99%)
VBAC: 1.30 per 1000 (0.13%)
CBAC: 18.5 per 1000 (1.85%)

It was only when I saw the next section labeled, “Rupture not including dehiscence” that what the study authors did become clear.

They included full thickness uterine ruptures and typically asymptomatic uterine dehiscences under the label “uterine rupture.” This is how they arrived at such a high overall “uterine rupture” rate.

It is these two events combined that they are comparing throughout the entire study anytime they use the term “uterine rupture.” This is why asking the questions “what did they measure,” “what definitions did they use,” and “how did they group events” is so important.

So when we look at just full thickness uterine ruptures, we can see that the rate among VBACs (0.051%) is just slightly higher than those who planned an elective repeat cesarean (0.05%). Also note the rate among those who labored after cesarean was only 0.33%.

Rupture not including dehiscence

ERCD: 0.50 per 1000 (0.05%)
LAC: 3.3 per 1000 (0.33%)
VBAC: 0.51 per 1000 (0.051%)
CBAC: 6.08 per 1000 (0.608%)

This uterine rupture rate among planned VBAC is unusually low. Typically the rate is about 0.4% among those laboring spontaneously after one prior cesarean.

Additionally, Young reported that 372 repeat cesareans would need to be performed to prevent one uterine rupture. This is an important number to keep in mind as we consider all the risks and benefits of VBAC and repeat cesarean to the birthing parent and infant as well as the birthing parent’s future fertility, pregnancies, deliveries, and children.

Maternal mortality & morbidity

Given how the abstract references “absolute rates of severe maternal morbidity and mortality,” you would think that the rate of maternal mortality (deaths) would be reported as a separate figure than morbidity (complications). But it’s not. Which makes it impossible to determine the actual number of maternal deaths associated with labor after cesarean versus elective repeat cesarean section.

Past studies have reported the risk of planned VBAC maternal mortality as 1 in 26,000 with a 3.5 increased risk of maternal mortality associated with elective repeat cesarean section. (Guise, 2010) Given this study’s ample sample size, it would have been great to add their data to the larger story.

Again, under the category of “what are we measuring,” let’s look at what this study included under maternal morbidity and mortality:

“death, acute myocardial infarction, heart failure, pulmonary edema, disseminated intravascular coagulation, cardiac arrest, assisted ventilation, cardiac complications from anesthesia, cardiopulmonary resuscitation, adult respiratory distress syndrome, acute/unspecified renal failure, blood transfusion given PPH, shock procedures to control bleeding given PPH, cesarean hysterectomy given PPH, total hysterectomy open approach given PPH, subtotal hysterectomy open approach given PPH and repair of injury to bladder and urethra.”

Here are the numbers they reported:

Maternal morbidity and mortality

ERCS: 5.65 per 1000 (0.56% or 1 in 177)
LAC: 10.7 per 1000 (1.07% or 1 in 93)
VBAC: 7.14 per 1000 (0.71% or 1 in 135)
CBAC: 14.3 per 1000 (1.43% or 1 in 699)

But once we remove blood transfusions due to postpartum hemorrhages, we get what they call “Restricted severe morbidity:”

Restricted severe morbidity

ERCS: 4.65 per 1000 (0.46% or 1 in 215)
LAC: 7.19 per 1000 (0.71% or 1 in 139)
VBAC: 2.49 per 1000 (0.24% or 1 in 402)
CBAC: 11.9 per 1000 (1.19% or 1 in 84)

The pattern we have seen here has been replicated many times. We see that VBAC has the lowest complication rate followed by ERCS, LAC, and CBAC. This study affirms what others have found.

Neonatal death and morbidity

Young concluded, “Failed vaginal birth after cesarean delivery was associated with a threefold higher rate of neonatal death.” That is a statement that gets your attention quick. So this is when we ask the question of relative versus absolute risk. That sentence gives us the relative risk: how the risks of cesarean birth after cesarean compares to elective repeat cesarean section. But let’s actually look at the absolute risk: what the numbers are.

Neonatal death

ERCS: 0.08 per 1000 (0.008% or 1 in 12,500)
LAC: 0.22 per 1000 (0.022% or 1 in 4,545)
VBAC: 0.18 per 1000 (0.018% or 1 in 5,556)
CBAC: 0.29 per 1000 (0.029% or 1 in 3,448)

Overall, the risk is very low across the board. Yes, it’s lowest with an elective repeat cesarean section. But a 1 in 3,488 risk of neonatal death with CBAC is still low. Can you image if that was the headline? It wouldn’t be nearly as shocking as “3 times the risk.”

This is another example of how focusing exclusively on the relative risk (“3 times higher risk!), and not sharing the absolute risk (1 in 3,488), can be used by providers to coerce a pregnant person or unintentionally misinform the public. And while accurate, yes, the risk of neonatal death is three times higher during a CBAC than an elective repeat cesarean section, it doesn’t give the full story. It doesn’t communicate the low absolute risk.

While these small numbers reflect real babies and real parents experiencing real loss, it is still a pretty rare occurrence. This is what parents pregnant after a cesarean need to know: Yes, the risk is there. But it is low.

In fact, Young found that 9,056 repeat cesareans need to be performed to prevent one neonatal death.

This is another area where there are no right or wrong answers, but we do need to counter that fact with the increasing risks associated with multiple cesareans.

It’s because the absolute risk is low that many medical organizations including the American College of OB/GYNs and the National Institutes of Health encourage women to have access to VBAC.

The Bottom Line

So, what does this all mean? This study found what many others have before. Yes, there is a higher relative risk with VBAC, but the absolute risks are low.

This is why medical organizations around the world have supported increasing access to VBAC. The challenge is that hospitals and providers can be very slow to follow the evidence and pervasive fear surrounding VBAC persists.

This is why it is so important for all of us to know the facts so we can confidently debunk myths one by one.

I hope this quick overview gave you a better understanding of what this study measures, what it all means, and how to evaluate VBAC medical studies in the future.

As you can see, reading and interpreting medical studies is a skill set of its own. This is why I have developed trainings for parents and professionals where they can easily and quickly get up to speed on the VBAC evidence. Reviewing the evidence is very time consuming, but getting the facts to the public, so we can increase VBAC access, is worth it.