Statistics in Decision Making

by Bronwyn Fackrell, M.S.

Here at Special Scars, we are frequently asked about the safety of pregnancy and vaginal birth with various types of uterine scars. Often times, care providers give one set of numbers and another set can be found in medical research literature, not to mention the countless anecdotes, both positive and negative. What’s a Special Scar woman to do? Having a little knowledge about statistics, their uses and limitations, can go a long way in helping each woman assess her risk and make the best decisions for herself and her family.

What are my odds of uterine rupture in a VBAC attempt?

This is probably the most frequently asked question. Everyone wants to know what the odds are in her own case. I hate to say it, but the short answer is, we don’t know. In the overall population of women attempting VBAC, we can expect a rupture rate of less than 1%, but we cannot predict which individual women will fall into that category. You may have factors that are either associated with higher rupture rates (a Special Scar, multiple scars, underlying general health issues, etc) or lower rates (previous vaginal birth, good overall health, longer interpartum interval, etc.) Each of those can be weighed, but we cannot create tailor-made odds. Researcher Caroline de Costa said, “Women having had a previous classical Cesarean section are a relatively small group, but amongst them are some who would wish to have a vaginal birth if they could safely do so. It might be that, contrary to current opinion, some of them could. At the moment, we just don’t have the evidence.”

My OB told me my odds of rupture are 10%.

Be suspicious of easily round numbers like 10, 25, 50, etc. That’s a sign it may be a guesstimate rather than based on research. Natural numbers aren’t usually so neat and tidy. Many care providers know very little about Special Scars. If they know anything at all, it’s that ACOG has quoted a rupture rate of 4-9% for classical, inverted T, and J scars. Those rates are based on research which does not differentiate between frank ruptures and non-traumatic scar separations noted on repeat c-section. (See Chauhan et all 2002) If you are quoted a rupture rate that sounds off, don’t be afraid to ask for a source. You may also consider countering with your own sources.

My OB didn’t believe the studies and said they were too small.

A limited sample size can skew the results one direction or the other. For instance, if a study had only two women and neither ruptured, can we conclude the rupture rate is 0%? Of course not. But how large should the sample size be? For events that happen around 1% of the time, we need sample sizes in the thousands. We currently do not have studies that large for Special Scars. The largest study, Shipp et al, 1999, included 377 women, for one specific special scar type (low vertical). The study we quote most often is Landon et al, 2004. It included 105 VBAC attempts among women with classical, inverted T, and J scars. Of the 105 women in that group, 2 ruptured, for a rate of 1.9%. In a group of 105, each and every woman has an impact of nearly 1%. That means each woman can significantly affect the end percentage. We would need a sample size more than 10 times larger to be confident we have an accurate picture. However, while it’s true that the available studies are small, it’s also true that they are the best we have to work with. You may want to point out to your care provider that none of the existing studies bear out the oft-quoted 4-9%.

Are the studies any good at all?

I don’t mean to imply that we should give up because we don’t have ideal data. However, we can’t have the same confidence that is now given to low transverse scars which are an order of magnitude more common. A further problem with many studies, including ACOG’s 4-9%, is that they lump scars together. Landon et al 2004, for instance, put classical, inverted T, and J scars all in the same category. There is no detail about how many of each scar were included, and how each individual scar fared. Furthermore, we don’t have solid evidence that these scars are functionally equivalent, as suggested by Patterson et al, 2002.

Someone has to rupture, right?

Wrong. In any given small group of women, it’s possible not to see a single rupture. A string of uneventful VBACs does not designate you the next rupture. Think about it. Why would the condition of another woman’s uterine scar affect yours? Now, it’s true that the more women there are, the more likely we are to see a rupture, but the simple counting is not what determines who will rupture next.

I used an online calculator to find my odds of a successful VBAC.

That’s a fun tool to play with, but don’t take it too seriously. It’s a simplistic way to look at your risk factors based on what the current research and practice suggest. The “exact” number it generates is only exact as far as the math goes. It’s far less so when applied to real human beings who have less measurable factors and random variables.

I’m more likely to die in a car crash than have a uterine rupture.

Strictly speaking this is true, but it is not really a fair comparison. Many of us get in a car daily. We have far more opportunities for fatal car crashes than we do for uterine ruptures. If we gave birth as often as we drove a car, we would be more likely to experience a rupture than a wreck.

Someone I know ruptured, so she made a poor decision.

You cannot judge the merit of a decision based on the outcome alone, because that was the one piece of information you did not have when you made the decision. This is an example of what’s called hindsight bias. If you had known ahead of time how it would turn out, it wouldn’t be much of a decision, would it? The outcome by itself does not determine whether you can accept either credit or blame. In his book Fooled by Randomness, statistician Nassim Nicholas Taleb said, “I will repeat this point until I get hoarse: A mistake is not something to be determined after the fact, but in the light of the information until that point.” (p. 58)

Other women with my scar type plan _______ (VBAC/repeat c-section/no more pregnancies). Should I do that too?

While it’s useful to know how others in a similar situation came to the conclusions they did, remember that this is your decision. You and your family are the ones who will live with the outcome. Remember too that we can’t know exactly how you compare and contrast with those other women.

How do I weigh my options?

Each woman needs to do this for herself. It’s very easy to get hung up on the numbers, but remember that they don’t tell the whole story. Rupture rates are only one factor to consider. Others include the impact of a rupture, (the current best evidence says 6.2% of uterine ruptures result in fetal death), desired family size and the cumulative effects of multiple c-sections, (the likelihood of potentially fatal placental complications increases with each c-section), the gestational age for repeat c-section and the possibility of iatrogenic prematurity, proximity to emergency facilities, and simply your own comfort level with every option. When Dr Mark Landon, a prominent VBAC researcher, was asked how to evaluate a woman with a J scar, he replied “It's very simple... what's the success rate, what factors if any are potentially influencing the increased risk for rupture and is that a level of risk acceptable to that woman undergoing child labor."

At the end of the day, you and your family are the ones most impacted by any choices made. Reading all the studies, talking to multiple health care providers, and listening to the stories of other mothers can be valuable, but ultimately, the choice is yours.