Category Archives: Using statistics

Does the term “TOLAC” tweak you?

On the acronym TOLAC (trial of labor after cesarean)….

Some studies break out statistics in four ways.

1. ERCS/D (elective repeat cesarean section/ delivery)
2. VBAC (vaginal birth after cesarean)
3. CBAC (cesarean birth after cesarean aka cesarean after planned VBAC)
4. TOLAC (VBAC + CBAC stats)

Because we are unable to predict who will have a VBAC or CBAC, the TOLAC stat enables us to review outcomes from a variety of angles:

  • TOLAC vs. ERCS
  • VBAC vs. ERCS
  • CBAC vs. ERCS

Some women find the TOLAC acronym offensive, because it implies “trying,” so practitioners sensitive to this may way to use the phrase “planning a VBAC.”   Understanding that TOLAC isn’t a dig at moms, but just a straightforward, objective term that care providers use, can (hopefully) take the sting out of the word.

Remember, your care provider is not your girlfriend.  They use clinical terms because that is the language of their world. They speak like clinicians because they are clinicians. All that said, providers who are aware of how the term TOLAC is received by some women use the term “planned VBAC.”

So moms, you use the language that works for you! Just remember that TOLAC is really more of a clinical term and when your provider uses it, it doesn’t necessarily mean that they are a jerk.  They just may have forgotten to code switch from clinical to sensitive language.

Moms don’t typically say, “I’m so excited for my TOLAC!” However, if you do, you might make your provider laugh and connect with them on a human level.

Two points for the person who knows how this picture is relevant…

20120808-102648.jpg

False comparison: Fatal car accidents and VBAC

RETRACTION/ CORRECTION: I originally posted this article challenging the thought that you are more likely to die in a fatal car accident than during a VBAC.  I tried to crunch the numbers in the way that I felt most accurate.  However, it has been bugging me ever since because there is no accurate way to compare these two events and I should have emphasized that more. We can accurately and fairly compare the risks of VBAC to the risks of a repeat cesarean or the risks of a first time time mom.  However, it is a misleading to compare the risks of birth to non-birth events because they are to different.  While I did discuss this at great length at the end of this article, the title I originally chose (Myth: Mom more likely to die in car accident than VBAC) just continued to feed this false comparison.  I have since updated the article and title.  I apologize for any confusion I caused.

_____________________

On fatal car accident statistics: There are many, many variables that factor into an individual’s risk of dying in a car accident.  The most accurate way to calculate your risk is by miles driven.  To learn more, please refer to the National Motorists Association’s document “Understanding Highway Crash Data.” I use the figures below in order to get an average rate for the purpose of discussion.

On terminology: Read why I use the term TOLAC.

_______________________________

Prepare yourself for yet another installation to the Birth Myth series.  I’ve heard this sentiment many times over the years and I’m sure you have too.   The well-meaning people who share this “statistic” simply desire to give moms seeking information on VBAC some encouragement:

If your husband is worried about you dying during a VBAC, tell him you are four times more likely to die in a car accident on your way home from work today.  Sorry if that sounds morbid, but the odds of the mother dying in a VBAC are truly minuscule.

Another article (filled with inaccurate statements, contradictions, and oodles of statistics without sources) recently making the rounds on Facebook says one of the risks of hospital birth is the 1:10,000 risk of a fatal car accident on the way to the hospital.

While these statements are very comforting, as birth myths tend to be, they are false comparisons.  We can accurately and fairly compare the risks of a TOLAC to the risks of a repeat cesarean or the risks of a first time time.  However, it is a misleading to compare the risks of birth to non-birth events.

Comparing unlike risks

Many birth advocates try to weigh the event of uterine rupture against other life events in an attempt to give context, but this is a misleading and inaccurate comparison.  Andrew Pleasant in his article entitled, Communicating statistics and risk, explains:

Try not to compare unlike risks.  For instance, the all-too-often-used comparison ‘you’re more likely to be hit by a bus / have a road accident than to…’ will generally fail to inform people about the risks they are facing because the situations being compared are so different.  When people assess risks and make decisions, they usually consider how much control they have over the risk.  Driving is a voluntary risk that people feel (correctly or not) that they can control.  This is distinctly different from an invisible contamination of a food product or being bitten by a malaria-carrying mosquito.

Comparing the risk of a non-communicable disease, for example diabetes or heart disease, to a communicable disease like HIV/AIDS or leprosy, is similarly inappropriate.  The mechanisms of the diseases are different, and the varying social and cultural views of each makes the comparison a risky communication strategy.

Take away message: Compare different risks sparingly and with great caution because you cannot control how your audiences will interpret your use of metaphor.

Comparing lifetime/annual risk to your risk of something happening over a day (or two)

Your annual or lifetime risk of something happening will often be higher than your risk of a birth related complication.  But this is because the annual risk of something measures your risk for 365 days.  The lifetime risk of something is often based on 80 years.  You are likely to be in active labor for one day, maybe two.  To compare the risk of something that happens over 1-2 days to the aggregate risk of something that could happen any day over 365 days or 80 years is unfair and confusing.

Look at something like your lifetime risk of breast cancer which is often quoted as 1 in 8.  So one could easily say, “Hey, I have a greater risk of breast cancer over my lifetime than I do have a uterine rupture!”  But, let’s look at this a bit more:

Again, I refer to Andrew Pleasant’s article, Communicating statistics and risk:

An oft-reported estimate is the lifetime breast cancer rate among women. This rate varies around the world from roughly three per cent to over 14 per cent.

In the United States, 12.7 per cent of women will develop breast cancer at some point in their lives. This statistic is often reported as, “one in eight women will get breast cancer”. But many readers will not understand their actual risk from this. For example, over 80 per cent of American women mistakenly believe that one in eight women will be diagnosed with breast cancer each year.

Using the statistic ‘one in eight’ makes a strong headline but can dramatically misrepresent individual breast cancer risk.

Throughout her life, a woman’s actual risk of breast cancer varies for many reasons, and is rarely ever actually one in eight. For instance, in the United States 0.43 per cent of women aged 30–39 (1 in 233) are diagnosed with breast cancer. In women aged 60–69, the rate is 3.65 per cent (1 in 27).

Journalists may report only the aggregate lifetime risk of one in eight because they are short of space. But such reporting incorrectly assumes that readers are uninterested in, or can’t comprehend, the underlying statistics. It is critically important to find a way, through words or graphics, to report as complete a picture as possible.

Take away message: Be extra careful to ensure your readers understand that a general population estimate of risk, exposure or probability may not accurately describe individual situations. Also, provide the important information that explains variation in individual risk. This might include age, diet, literacy level, location, education level, income, race and ethnicity, and a host of other genetic and lifestyle factors.

To compare events that are so different like the risk of a fatal car accident and the risk of TOL maternal mortality is inaccurate and doesn’t help moms understand their options.  Your risk of a car accident depends on how much you drive, when you drive, if you are distracted or on medication, etc, etc, etc.  The variables that impact your risk of dying during a  TOLAC are very different.  However, one way these two events are similar:  Sometimes we can make all the “right” or “wrong” decisions and the element of luck will sway us towards a good or bad outcome.

The problems with birth myths and false comparisons

False comparisons and birth myths like this are shared with the best of intentions.  So often the risks of VBAC are exaggerated for reasons having nothing to do with the health of baby and mom.  Birth advocates share these myths (which they believe to be true) as a way of boosting the morale of moms seeking VBAC as these moms are constantly faced with a barrage of unsupportive comments from family, friends, and even care providers.

The problem is, women make plans to have (home) VBAC/VBAMC based on these myths.  They make these plans because birth myths make the risk of VBAC, uterine rupture, infant death, and maternal death look practically non-existent.  That is dangerous.

Perpetuating these myths impedes a mom’s ability to provide true informed consent.  If a mom thinks her risk of uterine rupture is similar to a unscarred mom or a unscarred, induced mom, or less than her risk of getting struck by lighting or bitten by a shark, she does not have accurate picture of the risk.  And if she doesn’t understand the risks and benefits of her options, she is unable to give informed consent or make an informed decision.

Birth advocates get all up in arms about the mom who plans an elective, primary cesarean section without “doing her research.”  Or the mom who consents to an induction at 38 weeks because her OB “said it was for the best.”  Or when an OB coerces a mom into a repeat cesarean by saying the risk of uterine rupture is 15%.  Shouldn’t we be just as frustrated when moms plan (home) VBACs based on misrepresentations of the truth?  Shouldn’t we hold ourselves to the same standard that we expect from others?

The second problem with perpetuating these false comparisons and myths is that once women learn the true risks, they seem gigantic in comparison to the minuscule risk they had once accepted.  Now VBAC seems excessively risky and some loose confidence in their birth plans.  Birth advocates do not support moms by knowingly perpetuating these myths.  The reality is, the risks of VBAC are low.  We don’t need to exaggerate or minimize the benefits or risks of VBAC.  If we just provided women with accurate information from the get go, they would be able to make a true, informed decision.

The third problem is that we really look dumb when we say stuff like this.  If we want to be taken seriously, we really need to double check what we pass on.  I encourage you to ask for a source when someone says something that sounds to good to be true or just plain fishy.  (And hold me to the same standard!)  I often ask people for a source for their assertions… with varying results.

Sometimes people have a credible source available and share it with me.  I learn more and it’s all good.  Other times, people get angry.  They think I’m challenging them or trying to argue with them.  But the truth is, I’m just trying to learn. What I have found is, when people get angry, it’s sometimes because they don’t have a source and they are insulted that I didn’t accept their statement at face value.  They have just accepted what a trusted person told them as the truth and expect me to do the same.

Doesn’t it strike you as odd that some people encourage the continual questioning OBs and the medical system, yet expect you to accept what they say as The Truth no questions asked?   “Question everyone but me.”  Why?  Why is it when we question an OB, that’s a good thing, yet when we hold our birthy friends and colleagues to the same standard, that is being argumentative?  I say, ask for the source.  From everyone.

Take away messages

It is inaccurate and misleading to compare two events that are as different as a fatal car accident and TOL maternal mortality.  Period.

Let’s stop this false comparison and bring us back to what we should be comparing TOLAC/VBAC to: the risks of a repeat cesarean.

When women plan a VBAC based on false information,  their confidence can be shattered when they learn that the risk of uterine rupture and maternal death are much higher than they were lead to believe.

When women plan a VBAC based on false information, they are deprived of their right to informed consent.

While the risk of scar rupture is very different than the risk of a fatal car accident, it is similar to other serious obstetrical emergencies such as placental abruption, cord prolapse, and postpartum hemorrhage.

_________________________________

Guise J-M, Eden K, Emeis C, Denman MA, Marshall N, Fu R, Janik R, Nygren P, Walker M, McDonagh M. Vaginal Birth After Cesarean: New Insights. Evidence Report/Technology Assessment No.191. (Prepared by the Oregon Health & Science University Evidence-based Practice Center under Contract No. 290-2007-10057-I). AHRQ Publication No. 10-E003. Rockville, MD: Agency for Healthcare Research and Quality. March 2010.   http://www.ahrq.gov/downloads/pub/evidence/pdf/vbacup/vbacup.pdf

_________________________________

For those who are interested in the reasoning and mathematics from the original article:

But, if we were going to compare the unlike risks of a fatal car accident and TOLAC, this is how I would do it: compare the daily risks of the events.

Maternal death and TOL

Per the report presented at the 2010 NIH VBAC conference entitled Vaginal Birth After Cesarean: New Insights (Guise, 2010):

Overall rates of maternal harms were low for both TOL [trial of labor] and ERCD [elective repeat cesarean delivery]. While rare for both TOL and ERCD, maternal mortality was significantly increased for ERCD at 13.4 per 100,000 versus 3.8 per 100,000 for TOL . . . The rate of uterine rupture for all women with prior cesarean is 3 per 1,000 and the risk was significantly increased with TOL (4.7/1,000 versus 0.3/1,000 ERCD).

Put another way, there is a 0.0038% (1 in 26,316) risk of maternal death during a trial of labor.  For a mom to die is very rare.

Risk of a fatal car accident

Of the 311,000,000 people living in the US (US Census, 2012), about 40,000 die annually (Beck, 2006) from car accidents in the United States which gives us a annual rate of 0.0129% (1 in 7,752).  (But remember, this is a very rough representation of the risk due to all the factors I previously mentioned.)

Many women look at this number and say, “See, you are more likely to die in a car accident than during a TOL.”

But remember, 0.0129% (1 in 7,752) is the annual rate of Americans dying due to car accidents.

To compare something like your annual risk of a fatal car accident to your risk of dying during a TOL is an unfair and inaccurate comparison.   It would be more accurate (though still a false comparison) to compare your daily risk of a fatal car accident (because most people travel in a car every day) to the risk of maternal death during a TOL because you are not in labor every day for a year.  Let me explain.

Comparing TOL maternal mortality to fatal car accidents

Often this false comparison is expressed as, “You are more likely to have a fatal car accident on the way to the hospital than have a uterine rupture or die during a VBAC.”  But the risk of a fatal car accident on the day you drive to the hospital is not 1 in 7,752.  That is your risk over a year.  We have to estimate your risk on that day you drive to the hospital by dividing 0.0129% by 365 days which equals 0.00003534% or 1 in 2,829,458.

No matter what stat we use from any study, the risk of maternal mortality during a TOL is much greater.  (But remember, this is a false comparison anyways!)

Guise’s data pegs the risk at 0.0038% or 1 in 26,316 which is 107.5 times greater than the risk of a fatal car accident as you drive to the hospital in labor.  This does not mean that the risk of dying during a TOL is so large, but rather our daily risk of a fatal car accident is so small that it’s literally theoretical.  (Read Kim James’ “Understanding obstetrical risk” for more.)

What about the risk of uterine rupture?

Using the 0.47% (1 in 213 TOLs) risk of scar rupture (Guise, 2010), the risk of a fatal car accident is 13,283 times smaller.

Why don’t we spread the risk of rupture/maternal mortality across the entire pregnancy?

After I initially published this article, someone left this great comment on Facebook:

I get this, but I also get why using annual stats of car accidents would be accurate when you are looking at uterine rupture rates themselves and not just during TOL, since a risk of rupture exists throughout pregnancy and not just during labor and mom would be pregnant for approximately 10 months or more.

I wondered about the best way to crunch the numbers because these events are so different and thus so difficult to compare.  In the end, it is a false comparison, but here was my original thinking….

Most Americans are in a car everyday, so they have that risk – no matter how small – every day unless they are not in a car in which case their risk is zero.  The risk is primarily associated with being in a car.

The risk of uterine rupture and maternal mortality is primarily associated with being in labor, so we can’t spread the risk of rupture/maternal mortality across the whole pregnancy because the risk of rupture/maternal mortality is not the same from conception to delivery.

One study examined 97% of births that occurred in the The Netherlands from 1st August 2004 until 1st August 2006 and found that 9% (1 in 11) of scar ruptures happened before the onset of labor. When we take 9% of the overall rate of scar rupture 0.64% (1 in 156) (including non-induced/augmented, induced, and augmented labors), we get a 0.0576% (1 in 1736) risk of pre-labor scar rupture and a 0.5824% (1 in 172) risk of rupture during labor (Zwart 1009). Since the risk of rupture is not the same over the entire pregnancy and labor, we cannot accurately calculate a daily risk of rupture.

In other words, the risk of rupture is rare before labor (0.0576% or 1 in 1736) and then becomes uncommon when labor begins (0.5824% or 1 in 172).  Even though we could go into labor anytime during pregnancy, the risk before we go into labor is so small in comparison to the risk when we actually go into labor.

Myth: Two numbers less than 1% are similar

I have often heard, “If two numbers are less than 1%, they are similar.”  Typically
this is expressed while comparing the risks of rupture in an unscarred versus scarred uterus.   But is this true?  How different can two numbers less than 1% be?calculator-983900_1920

Two numbers less than 1% are no more similar than two numbers greater than 1%

Just because two numbers are less than 1%, that doesn’t make them any more similar than two numbers greater than 1%.  A 2% risk of something happening is very different than an 89% risk.  While they are both greater than 1%, they represent drastically different levels of risk.

2% = 1 in 50 risk

89% = 1 in 1.12 risk

89% represents a 44 times greater risk than 2%.

What about numbers less than 1%?

It might seem rational that since numbers less than 1% are so small, that there wouldn’t be as much of a difference between them.  But numbers less than 1% work in the same way as those greater than 1%.   Let’s run a few and measure the difference.

1 in 100 represents 1%.

1 in 1,000, is the same as 0.1%, and is 10 times smaller than 1%.

1 in 10,000, is the same as 0.01%, and is 100 times smaller than 1%.

1 in 100,000, is the same as 0.001%, and is 1,000 times smaller than 1%.

1 in 1,000,000, is the same as 0.0001%, and is 10,000 times smaller than 1%.

Comparing small risks

According to Zwart* (2009), the risk of uterine rupture in:

– an unscarred mom is 1 in 14,286 (0.007% or 0.7 in 10,000) and

– a scarred mom is 1 in 156 (0.64% or 64 in 10,000).

(Both statistics include non-induced/augmented, induced, and augmented labors.)  Even though both numbers are less than zero, they represent very different levels of risk.  In fact, the risk of rupture in an unscarred mom is 91 times smaller than a scarred mom.  It’s not that the risk of rupture is excessively high in a scarred mom, but that it is so very, very, very low in an unscarred mom.

Using the language from Kim James’ handout Understanding Obstetrical Risk, the risk of rupture in an unscarred mom would be described as “very rare” whereas the risk of rupture in a scarred mom would be described as “uncommon.”

Take away messages

Just because two numbers are less than 1% does not mean that they are similar.  Numbers below 1% represent just as much of a range as numbers greater than 1%.

While the risk of scar rupture is very different than the risk of unscarred rupture, it is similar to other serious obstetrical emergencies such as placental abruption, cord prolapse, and postpartum hemorrhage.
________________

* Zwart (2009) differentiated between uterine rupture and dehiscence, featured 358,874 total deliveries, 25,989 of which were trials of labor after a cesarean.  Zwart included 97% of births in The Netherlands between August 1, 2004 and August 1, 2006, making it “the largest prospective report of uterine rupture in women without a previous cesarean in a Western country.”

Zwart, J. J., Richters, J. M., Ory, F., de Vries, J., Bloemenkamp, K., & van Roosmalen, J. (2009, July). Uterine rupture in the Netherlands: a nationwide population-based cohort study. BJOG: An International Journal of Obstetrics and Gynaecology, 116(8), pp. 1069-1080. Retrieved January 15, 2012, from http://onlinelibrary.wiley.com/doi/10.1111/j.1471-0528.2009.02136.x/full

The best compilation of VBAC/ERCS research to date

“There is a major misperception that TOLAC [trial of labor after cesarean] is extremely risky” – Mona Lydon-Rochelle PhD, MPH, MS, CNM, March 2010

In terms of VBAC, “your risk is really, really quite low” – George Macones MD, MSCE, March 2010

Both Drs. Macones and Lyndon-Rochelle are medical professionals and researchers who made these statements at the 2010 NIH [National Institutes of Health] VBAC Conference. Now you may think, “Wait a sec. Everything I’ve heard from my family, friends, and medical provider is how risky VBAC is and how cesareans are the conservative, prudent, and safest choice.” Why the discrepancy between the statements of these two prominent care provider researchers and the conventional wisdom prevalent in America?

It’s likely that your family, friends, and even your medical provider are not familiar with the latest and best compilation of VBAC research that was released in March 2010. It’s also possible that they are not familiar with the latest VBAC recommendations published in July 2010 by the American Congress of Obstetricians and Gynecologists (ACOG). Additionally, there are often legal and non-medical factors at play that influence how care providers counsel women on VBAC, including pressure from hospital administrators.

When I come across any VBAC study, I always wonder if it made the cut to be included in the 400 page Guise 2010 Evidence Report that was the basis for the 2010 NIH VBAC Conference. Guise 2010 reviewed each published VBAC study, performed a quality assessment, and assembled an excellent review of the VBAC literature to date:

Quality assessment is an assessment of a study’s internal validity (the study’s ability to measure what it intends to measure). If a study is not conducted properly, the results that they produce are unlikely to represent the truth and thus are worthless (the old adage garbage in garbage out). If however, a study is structurally and analytically sound, then the results are valuable. A systematic review, is intended to evaluate the entire literature and distill those studies which are of the highest possible quality and therefore likely to be sound and defensible to affect practice.

Guise focused on these key questions: “1) a chain of evidence about factors that may influence VBAC, 2) maternal and infant benefits and harms of attempting a VBAC versus an elective repeat cesarean delivery (ERCD), and 3) factors that may influence maternal and infant outcomes.” Ultimately, this 400 page document was distilled into the 48 page VBAC Final Statement produced by the NIH VBAC Conference.

This is wonderful because people who want the big picture, can read the VBAC Final Statement whereas those who want to know the exact figures, how studies were included/excluded, and the strength of the data available, can read the Guise 2010 Evidence Report.

You can get a feel for the topics presented at the NIH VBAC Conference by reading the Programs & Abstracts document. If you want more detail, you can watch the individual presentations. I was there for the three day conference and it was eye opening. I wish more medical professionals and moms were aware of this information as they are excellent resources for anyone looking to learn more about VBAC.

Everyone wants to know the bottom line: what is the risk of death or major injury to mom and baby. Here is an overview of maternal and infant mortality and morbidity per Guise (2010). It’s important to remember that the quality of data relating to perinatal mortality was low to moderate due to the high range of rates reported by the strongest studies conducted thus far. Guise reports the high end of the range when they discuss perinatal mortality which was 6% for all gestational ages and 2.8% when limited to term studies. This is a long way of saying, we still don’t have a good picture of how many babies die due to uterine rupture.

It’s also important to remember that the statistics shared in Guise (2010) are for all VBACs. They include all scar types, women who have had multiple prior cesareans, induced/augmented labors, etc. It would have been helpful if they had broke out the data in these ways as we know we can reduce the risk of rupture (and thus perinatal mortality) through spontaneous labor.

While rare for both TOL [trial of labor after cesarean] and ERCD [elective repeat cesarean delivery], maternal mortality was significantly increased for ERCD at 13.4 per 100,000 versus 3.8 per 100,000 for TOL. The rates of maternal hysterectomy, hemorrhage, and transfusions did not differ significantly between TOL and ERCD. The rate of uterine rupture for all women with prior cesarean is 3 per 1,000 and the risk was significantly increased with TOL (4.7 1,000 versus 0.3 1,000 ERCD). Six percent of uterine ruptures were associated with perinatal death. Perinatal mortality was significantly increased for TOL at 1.3 per 1,000 versus 0.5 per 1,000 for ERCD… VBAC is a reasonable and safe choice for the majority of women with prior cesarean. Moreover, there is emerging evidence of serious harms relating to multiple cesareans… The occurrence of maternal and infant mortality for women with prior cesarean is not significantly elevated when compared with national rates overall of mortality in childbirth. The majority of women who have TOL will have a VBAC, and they and their infants will be healthy. However, there is a minority of women who will suffer serious adverse consequences of both TOL and ERCD. While TOL rates have decreased over the last decade, VBAC rates and adverse outcomes have not changed suggesting that the reduction is not reflecting improved patient selection.

Women are entitled to accurate, honest, and high quality data. They don’t deserve to have the risks exaggerated by an OB who wishes to coerce them into a repeat cesarean nor do they deserve to have risks sugar-coated or minimized by a midwife or birth advocate who may not understand the risk or whose zealous desire for everyone to VBAC clouds their judgement. Sometimes it can be hard to find good data on VBAC which is why I’m so thankful for the 2010 NIH VBAC Conference and all the excellent data that became available to the public as a result. There are real risks and benefits to VBAC and repeat cesarean and once women have access to good data, they can individually choose which set of risks and benefits they want. I think the links I have provided above represents the best data we have to date.

Confusing fact: Only 6% of uterine ruptures are catastrophic

It is important to note that the information shared in Guise (2010), the 400 page Evidence Report on which the 2010 NIH VBAC Conference was based, collected the best data we have available on trial of labor after cesarean.  That said, they reported, “Overall, the strength of evidence on perinatal mortality was low to moderate” due to the wide range of perinatal mortality rates reported by the studies included in the report.  Bottom line: We still don’t have an accurate idea of how deadly uterine rupture is to babies.  This is a topic on which Guise recommended future researchers focus.  I highly recommend that anyone interested in TOLAC (trial of labor after cesarean), especially those who blog or share information on social networking sites, review this very important document as it is a fascinating analysis of the best research we have to date on TOLAC.


How many times have you heard, “Only 6% of uterine ruptures are catastrophic” or “Uterine rupture not only happens less that one percent of the time, but the vast majority of ruptures are non-catastrophic?” But what does that mean? Does that mean only 6% of uterine ruptures are “complete” ruptures? Result in maternal death? Infant death? Serious injury to mom or baby? This article will explain to you the difference between uterine rupture and uterine dehiscence as well as explain the source and meaning of the 6% statistic.

Distinguishing between uterine rupture and uterine dehiscence

First, it’s important to understand what a uterine rupture is and how that differs from a uterine dehiscence. Uterine rupture, also called true, complete, or even (to further add to the confusion) catastrophic rupture, is a opening through all the layers of the uterus. Per a Medscape article on Uterine Rupture in Pregnancy:

Uterine rupture is defined as a full-thickness separation of the uterine wall and the overlying serosa. Uterine rupture is associated with (1) clinically significant uterine bleeding; (2) fetal distress; (3) expulsion or protrusion of the fetus, placenta, or both into the abdominal cavity; and (4) the need for prompt cesarean delivery and uterine repair or hysterectomy.

Whereas a uterine dehiscence, also called a incomplete rupture or a uterine window, is not a full-thickness separation. It’s often asymptomatic, does not pose any risk to mom or baby, and does not require repair. Again, I refer to Medscape:

Uterine scar dehiscence is a more common event that seldom results in major maternal or fetal complications. By definition, uterine scar dehiscence constitutes separation of a preexisting scar that does not disrupt the overlying visceral peritoneum (uterine serosa) and that does not significantly bleed from its edges. In addition, the fetus, placenta, and umbilical cord must be contained within the uterine cavity, without a need for cesarean delivery due to fetal distress.

When reading medical studies, look for how they define uterine rupture in the “Methods” section. While some medical studies combine the statistics for rupture and dehiscence, ultimately reporting an inflated rate of rupture, other studies distinguish between the two events.

So, what does the 6% statistic mean and where did it come from?

The statistic “Only 6% of uterine ruptures are catastrophic” is from the Evidence Report (Guise 2010) which was the basis of the 2010 NIH VBAC Conference and it refers to the rate of infant death due to uterine rupture. Here is the exact quote:

The overall risk of perinatal death due to uterine rupture was 6.2 percent. The two studies of women delivering at term that reported perinatal death rates report that 0 to 2.8 percent of all uterine ruptures resulted in a perinatal death (Guise 2010).

In other words, of the women who had uterine ruptures, 6.2% (1 in 16) resulted in infant deaths. When we limited the data to women delivering at term, as opposed to babies of all gestational ages, the risk was as high as 2.8 (1 in 36)%.

When we look at the overall risk of an infant death during a trial of labor after cesarean, the NIH reported the rate of 0.13%, which works out to be one infant death per 769 trials of labor.

The source of the confusion

The problem with this statistic is that some people have misinterpreted it to mean that only 6% of ruptures are true, complete uterine ruptures. In other words, if we take the 0.4% (1 in 240) uterine rupture rate (Landon, 2004), they believe that only 6% of those ruptures or 0.024% (1 in 4166) are true, complete ruptures. This is false. The 0.4% uterine rupture statistic measured true, complete, uterine ruptures in spontaneous labors after one prior low, transverse (“bikini cut”) cesarean.

So how many dehiscences did Landon (2004) detect? Landon reported a 0.7% uterine rupture rate and a 0.7% dehiscence rate. (Note that these statistics include a variety of scar types as well spontaneous, augmented, and induced labors.) So Landon found that dehiscence occurs at the same rate as uterine rupture.

I think the best way to avoid confusion is to use very clear language: 6.2% (1 in 16) of uterine ruptures result in an infant death. Put another way, for every 16 uterine ruptures, there will be one baby that dies.

Elapsed time and infant death

What determines if a baby dies or has brain damage? Some research on infant cord blood gases has suggested that if the baby isn’t delivered (almost always by CS) within 16 – 17 minutes of a uterine rupture, there can be serious brain damage or death to baby. You can watch a presentation from the 2010 NIH VBAC Conference entitled “The Immediately Available Physician Standard” by Howard Minkoff, M.D. for more information or read his presentation abstract.

Now you know the difference between uterine rupture, uterine dehiscence and the meaning of the 6% statistic. It’s helpful to understand the terminology used in relation to uterine rupture otherwise it can be very confusing as you wade your way through the statistics! It’s also very important for people to use specific words whose definitions are clear instead of words such as “catastrophic” that could mean multiple things.

Read more about uterine rupture, scare tactics, birth myths, cesarean section, and the steps for planning a VBAC.

Afterward – The big picture

The following are excerpts from the Evidence Report (Guise 2010) , the 400 page evidence report assembled for the 2010 NIH VBAC Conference. The limitation of Guise (2010) is that these stats are for all VBACs – all scar types, multiple prior cesareans, induced/augmented labors, etc. It would have been helpful if they had broke out the data in these ways.

While rare for both TOL [trial of labor] and ERCD [elective repeat cesarean delivery], maternal mortality was significantly increased for ERCD at 13.4 per 100,000 versus 3.8 per 100,000 for TOL. The rates of maternal hysterectomy, hemorrhage, and transfusions did not differ significantly between TOL and ERCD. The rate of uterine rupture for all women with prior cesarean is 3 per 1,000 and the risk was significantly increased with TOL (4.7 1,000 versus 0.3 1,000 ERCD). Six percent of uterine ruptures were associated with perinatal death.” Perinatal death due to UR from term studies was 2.8%. “Perinatal mortality was significantly increased for TOL at 1.3 per 1,000 versus 0.5 per 1,000 for ERCD… VBAC is a reasonable and safe choice for the majority of women with prior cesarean. Moreover, there is emerging evidence of serious harms relating to multiple cesareans… The occurrence of maternal and infant mortality for women with prior cesarean is not significantly elevated when compared with national rates overall of mortality in childbirth. The majority of women who have TOL will have a VBAC, and they and their infants will be healthy. However, there is a minority of women who will suffer serious adverse consequences of both TOL and ERCD. While TOL rates have decreased over the last decade, VBAC rates and adverse outcomes have not changed suggesting that the reduction is not reflecting improved patient selection.

A systematic review strives to be patient-centered and to provide both patients and clinicians with meaningful numbers or estimates so they can make informed decisions. Often, however, the data do not allow a direct estimate to calculate the numbers that people desire such as the number of cesareans needed to avoid one uterine rupture related death. The assumptions that are required to make such estimates from the available data introduce additional uncertainty that cannot be quantified. If we make a simplistic assumption that 6 percent of all uterine ruptures result in perinatal death (as found from the summary estimate), the range of estimated numbers of cesareans needed to be performed to prevent one uterine rupture related perinatal death would be 2,400 from the largest study,204 and 3,900-6,100 from the other three studies of uterine rupture for TOL and ERCD.10, 97, 205 Taken in aggregate, the evidence suggests that the approximate risks and benefits that would be expected for a hypothetical group of 100,000 women at term gestational age (GA) who plan VBAC rather than ERCD include: 10 fewer maternal deaths, 650 additional uterine ruptures, and 50 additional neonatal deaths. Additionally, it is important to consider the morbidity in future pregnancies that would be averted from multiple cesareans particularly in association with placental abnormalities.

Myth: Risk of uterine rupture doesn’t change much after a cesarean

myth versus reality

1/18/12 – The difference in uterine rupture (UR) rates between unscarred and scarred uteri is significant: 1 in 14,286 in an unscarred uterus and 1 in 156 in a scarred uterus.  Another way to express this is: 0.7 in 10,000 (0.007%) in an unscarred uterus and 64 in 10,000 (0.64%) in a scarred uterus.  This 91 times greater risk does not mean that the risk of UR is so large in a scarred mom, it’s that it’s so very, very small in an unscarred mom.

________________________________________

I came across a couple different bits of (mis)information the past day that have really concerned me. In both situations, people, one of whom is a certified professional midwife (CPM), give false information regarding how a cesarean affects one’s risk of uterine rupture in future pregnancies.

First, a women with a prior cesarean asks for uterine rupture rates after a cesarean, “preferable one with stats” on Facebook. One woman gives this reply:

… almost all cases the risk of rupture is less than one percent, even after multiple sections, or special scars such as an inverted T. The risk is roughly double what it is for an unscarred uterus, but considering the tiny numbers it doesn’t really make a difference, especially since the vast majority of ruptures are not catastrophic in nature, something that is not differentiated in study results.

(There are several things that are false in this statement, but I’ll save those for another post.) Then later in the day, I came across this comment from a CPM’s website:

Will you do a vaginal birth after cesarean?
Yes. Studies have shown that there isn’t much of a difference in uterine rupture rates in someone that has had a previous cesarean and someone who has never had one. A lot of my clients are VBAC’s or attempted VBAC’s. I am completely comfortable with this.

Both of these representations of uterine rupture after a cesarean are erroneous. It’s especially disturbing that a midwife who is counseling VBAC moms and attending their births at home, is giving her clients grossly incorrect information. The risk of a uterine rupture does much more than double after a cesarean as the risk in an unscarred uterus is infinitesimal in comparison to a scarred uterus.

Comparing the risk of uterine rupture: Prior cesarean vs. no prior cesarean

I started looking around and quickly found Uterine rupture in the Netherlands: a nationwide population-based cohort study (Zwart, 2009) which contains the data I needed to compare the rates of rupture in unscarred vs. scarred uteri. You can read the study in its entirety here.

This study included 358,874 total deliveries, making it “the largest prospective report of uterine rupture in women without a previous cesarean in a Western country.” It also differentiates between uterine rupture and dehiscence which is really important because we want to measure the rate of complete rupture. (Remember how the lady from Facebook made the statement, ” the vast majority of ruptures are not catastrophic in nature, something that is not differentiated in study results.” That portion of her statement was also false.)

Zwart (2009) looked at 25,989 deliveries after a cesarean and found 183 ruptures giving us a 0.64% uterine rupture rate or 64 per 10,000 deliveries. 72% of those ruptures occurred in spontaneous labors. Of the 183 ruptures, 7.7% resulted in infant deaths representing 14 babies dying. This gives us a rate of infant mortality due to uterine rupture after a cesarean of 0.05% or 5 in 10,000 deliveries.

Zwart also looked at 332,885 deliveries with no prior cesarean resulting in 25 ruptures giving us a 0.007% uterine rupture rate or .7 per 10,000 deliveries. 56% of ruptures occurred in spontaneous labors. Of the 25 ruptures, 24% resulted in infant deaths representing 6 babies dying. This gives us a rate of infant mortality due to uterine rupture in an unscarred uterus of 0.0018% or 0.18 in 10,000 deliveries.

This study found that the risk of uterine rupture is 91 times greater in a woman with a prior cesarean vs. a woman without a prior cesarean. Not double, not similar, but 91 times greater.

It is important to note that, “severe maternal and neonatal morbidity and mortality were clearly more often observed among women with an unscarred uterine rupture as compared to uterine scar rupture.” Meaning, if an unscarred mom ruptures, her baby is more likely to die than a scarred mom. We see this when we compare the 24% of unscarred ruptures that resulted in an infant death vs. the 7.7% of scarred ruptures that resulted in an infant death which represents a 3 fold greater risk.

However, due to the fact that uterine rupture occurs more frequently in a scarred uterus, the risk of infant mortality due to uterine rupture after a previous cesarean was 27.8 times greater than the risk of infant mortality after a rupture in an unscarred uterus.

In other words, while ruptures in unscarred uteri are more deadly to infants, more infants die due to ruptures in scarred uteri because they occur more frequently.

OBs are often vilified (rightfully so) for giving women inflated rates of uterine rupture and I’ve documented several examples here: Another VBAC Consult Misinforms, Scare tactics vs. informed consent, Hospital VBAC turned CS due to constant scare tactics, and A father says, Why invite the risk of VBAC?.

As a result, women seek out midwives thinking that they will be a source of accurate information and judicious support. But what happens when your midwife tells you that your risk of uterine rupture has not increased as a result of your prior cesarean section? If you have done your homework, hopefully you find another midwife fast. I would really question the skills and knowledge of a midwife who is so unknowledgeable on the risks of VBAC and yet attends VBAC births in an out-of-hospital setting.

But suppose your haven’t done your homework, you trust your midwife, and you move forward with your plan to have a VBAC at home based on the incorrect statistics she supplies. I can’t begin to imagine the rage I would feel if I decided to have a home VBAC based on false information provided by my care provider, and then the unimaginable happened, and I ruptured, and then I learned the truth: that my risk of uterine rupture increased 91 times as a result of my prior cesarean. I would be beyond angry. I would feel so betrayed.

It’s unfortunate when a woman chooses a mode of delivery based on false information. Whether it’s a a woman deciding to have a repeat cesarean due to the exaggerated risk of uterine rupture provided by her OB or a woman deciding to have a (home) VBAC due to her midwife playing down and underestimating the risk of uterine rupture. It is just as bad to minimize the risk of uterine rupture as it is to inflate the risk.

While the risk of rupture in a spontaneous labor after one prior low transverse cesarean is comparable to other obstetrical emergencies, it is important for women weighting their post-cesarean birth options to know that their risk increased substantially due to their prior cesarean. It is important for them to understand the risks and benefits of VBAC vs. repeat cesarean. It is important for them to have access to accurate information and be able to differentiate between a midwife’s/blogger’s/doula’s/birth advocate’s/person on Facebook’s hopeful opinion vs. documented statistics.

I implore those who interact with, and have impact on, women weighing their birth options: do not pass along information, no matter how great it sounds, if you don’t have a well-designed scientific study supporting it. If you hear a statistic you would love to use and share, just ask the person who gave you this information,”What is the source?” and use the citation anytime you quote the statistic. But if the person doesn’t have a well-designed scientific study, be wary and don’t use the stat. This way, we can reduce the rumor and increase the amount of good information on the Internet. I know, a lofty goal.

Read more birth myths debunked including Lightning strikes, shark bites, and uterine rupture and Myth: Unscarred mom induced (with Pit) as likely as VBAC mom to rupture.

__________________________________

Zwart, J. J., Richters, J. M., Ory, F., de Vries, J., Bloemenkamp, K., & van Roosmalen, J. (2009, July). Uterine rupture in the Netherlands: a nationwide population-based cohort study. BJOG: An International Journal of Obstetrics and Gynaecology, 116(8), pp. 1069-1080. Retrieved January 15, 2012, from http://onlinelibrary.wiley.com/doi/10.1111/j.1471-0528.2009.02136.x/full

rutpures in scarred uteri

Myth: Induced unscarred mom as likely as VBAC mom to rupture

Fact or MythUpdate 1/20/12 – Someone who believed this birth myth to be true, told me that the source of this information was an OB from St. Louis who presented at the 2011 ICAN conference. I contacted ICAN and they said that the person must be referring to Dr. George Macones. Yet, no one on the ICAN Board, who were seated at the front table during his presentation, remembers him saying that induced, unscarred women have the same risk of uterine rupture as a VBAC mom. And I would think that if he gave a stat like, everyone would have remembered because it is quite a remarkable statement as you will see shortly. While many women repeat, believe, and defend this statement, no one has supplied one study to me to support it.

Update 1/21/12 – Ruth S Beattie Dicken, the Speaker Chair of the 2011 ICAN conference contacted me via Facebook and said, ” Dr Macones did not say that. Nor did any other OB. I sat in on every session with OB speakers.”

Update 1/21/12: The difference in uterine rupture (UR) rates between unscarred, induced uteri and scarred uteri is significant: 2.2 per 10,000 in an unscarred, induced uterus and 64 in 10,000 in a scarred uterus. But it’s not that the risk of UR is so large in a scarred mom, it’s that it’s so very, very small in an unscarred mom, even when she is induced.

_________________________________________________

OBs are often vilified (rightfully so) for giving women inflated rates of uterine rupture and I’ve documented several examples over the years: Another VBAC Consult Misinforms, Scare tactics vs. informed consent, Hospital VBAC turned CS due to constant scare tactics, and A father says, Why invite the risk of VBAC?. But the midwife (or OB, but it’s generally a midwife) who gives false information that minimizes the risk of rupture is just as harmful to the VBAC mom. Since I wrote Lightning strikes, shark bites, and uterine rupture, I’ve been making mental notes of other birth myths that seem to be forwarded from woman to woman, without anyone asking, “That’s a great statistic! What’s the source?”

There is one that I hear quite often:

A woman without a prior cesarean whose labor is induced is just as likely as a VBAC mom to experience an uterine rupture.

Recently, I heard it again and I really wanted to know if there was some study that demonstrated this. It’s a logical conclusion that inducing an uscarred woman would increase her risk of rupture as uterine rupture is listed as a risk for Pitocin and prostaglandins (such as Cytotec and Cervidil) but how much does induction increase the risk of uterine rupture in an unscarred uterus? And does the rate of rupture increase so much that it is the same as the risk of rupture in a VBAC mom? I had unsuccessfully looked for that information in the past, so I went to my Facebook page and asked if anyone had a source.

Several women responded who had heard this information, two of which from their midwives which is really frightening. Unfortunately, no one who responded could cite where they heard this information. So I started looking and found Uterine rupture in the Netherlands: a nationwide population-based cohort study (Zwart, 2009).

This study included 358,874 total deliveries, making it “the largest prospective report of uterine rupture in women without a previous cesarean in a Western country.” It also differentiates between uterine rupture and dehiscence which is really important because we want to measure the rate of complete rupture. You can read the study in its entirety here.

The role of induction in scarred and unscarred uterine rupture

Zwart utilized multiple methods of induction: cervical prostaglandins (sulproston, dinoproston, and misoprostol aka Cytotec), oxytocin (Pitocin) and mechanical dilatation. Prostaglandin “dosages ranged from 0.5 to 2.0 mg with a minimal interval of 4 h in between,” but they do not provide the dosages of the women who ruptured.

Of the 208 scarred and unscarred uterine ruptures, 130 (62.5%) occurred during spontaneous labor reflecting 72% of scarred ruptures and 56% of unscarred ruptures. 28 (13.5%) ruptures occurred during cervical prostaglandin induction. 22 (10.6%) ruptures occurred during oxytocin (Pitocin) induction.

It seems that there were women who were induced with prostaglandins and Pitocin as measured in Table 5. But there is no measure for women who ruptured and were induced with both prostaglandins and Pitocin in any of the uterine rupture tables.

There is no mention of Bishop’s score, but they did provide the “reasons for induction with prostaglandins [in scarred women which] included (nearly) post-term pregnancy (n = 10), intra uterine fetal death/ multiple congenital abnormalities (n = 5), elective (n = 3), pregnancy induced hypertension (n = 2), intra uterine growth restriction (n = 1) and prelabour rupture of membranes (n = 1).”

Interestingly, this Netherlands-based study found “there was a trend towards more liberal use of prostaglandins for induction of labour in low-volume hospitals as compared to middle- and high-volume hospitals (24.4% versus 13.0% of cases, P = 0.29).”

It’s also interesting that there were no maternal deaths even though “18 [unscarred] women (72%), rupture occurred outside office hours.”

The risk of uterine rupture in an induced labor without a prior cesarean

The study found, ” In 11 women [without a prior cesarean who experienced a uterine rupture], labour was induced, in all but one with prostaglandins.” Said in another way, 40% of the unscarred women who ruptured were induced with prostaglandins versus only 12.1% of scarred moms who ruptured.

So Zwart found that it’s not the Pitocin that causes the ruptures in unscarred moms, it’s the prostaglandins. This is logical because prostaglandins are harder to control. If the uterus is hyper-stimulating due to prostaglandins, they continue to work on the uterus even after they have been removed from the cervix. Pitocin, on the other hand, has a short half-life so the body responds quicker to the drip being turned off in the event of uterine hyper-stimulation.

While we know that there are 332,885 unscarred women included in this study, we don’t know the number or percentage of unscarred women who were induced. We need this information in order to calculate the rate of uterine rupture in induced, unscarred women.

So I did a little looking and I found Verhoeven (2009) which states ” In The Netherlands induction rates have remained stable over the last decades at approximately 15%.” Since the induction rate has been stable, and this study included 97% of births in The Netherlands between August 1, 2004 and August 1, 2006, I feel comfortable using this 15% rate of induction to calculate the rate of uterine rupture in induced, unscarred women.

So when we take 15% of the 332,885 unscarred women in the study, we get 49,933 induced, unscarred women.

Dividing the 11 ruptures that occurred in induced, unscarred women by 49,933 total induced, unscarred women, we get the following uterine rupture rate in induced, unscarred women: 0.022% or 2.2 per 10,000 deliveries.

Now let’s look at the rate of uterine rupture in women with a prior cesarean: “25,989 trials of labor were attempted in the Netherlands during the study [resulting in 183 ruptures.] The risk of uterine rupture would then be 0.64%” or 64 in 10,000 deliveries. This rate includes ruptures in induced and spontaneous labors, but we do know that 72% of those ruptures occurred during spontaneous labors.

In other words, a woman with a prior cesarean section has a uterine rupture risk 29 times greater than the risk of uterine rupture due to induction in a woman without a prior cesarean, 0.64% vs. 0.022%.

Another way to look at the data is: you would need to induce 4,546 women without a prior cesarean in order to get one uterine rupture due to induction.

While I hadn’t seen the numbers until now, I was always very skeptical when I heard this rumor. I’m glad to finally have hard numbers to share.

How does induction affect the rate of uterine rupture in an unscarred woman?

Next, since I had all the data available, I wanted to calculate how induction affects the rate of uterine rupture in an unscarred woman. Remember that 10 of the 11 ruptures in induced, unscarred women occurred during the use of prostaglandins and we don’t have information on the dosage in those labors.

We already established that the rate of rupture in an induced, unscarred labor was 0.022% or 2.2 per 10,000 deliveries.

The remaining 14 ruptures of the 25 total ruptures occurred during spontaneous labor.

14 spontaneous ruptures among 282,952 spontaneous labors in unscarred women, gives us a 0.0049% uterine rupture rate or .49 per 10,000 deliveries.

As I suspected, an unscarred woman induced with prostaglandins has a greater risk of uterine rupture than an unscarred woman in a spontaneous labor, but now we have exact figures: 0.022% vs. 0.0049%. Prostaglandin induction in an unscarred woman increases her risk of uterine rupture almost 5 times, but the overall risk is still extremely low.

Moving forward

It was interesting to note that among women with a prior cesarean, 72% of ruptures occurred during spontaneous labor. The scar itself, that prior cesarean surgery, is what increases the risk of uterine rupture the most. With this in mind, the researchers state:

With 29% of all previous caesareans being performed for breech presentation, we clearly show the negative side effects and long-term adverse consequences of routinely performing elective caesarean for breech delivery . . . the only means of reducing the incidence of uterine rupture is to minimise the number of inductions of labor and to closely monitor women with a uterine scar. . . Ultimately, the best prevention [of uterine rupture] is primary preventions, i.e. reducing the primary cesarean delivery rate. The obstetrician who decides to perform a caesarean has a joint responsibility for the late consequences of that decision, including uterine rupture.

This is why more hospitals offering breech vaginal birth and VBAC, such as Portland, OR based Oregon Health & Science University (OHSU), is so important. Read more about OHSU’s mission to reduce the cesarean rate.

As I say in Myth: Risk of uterine rupture doesn’t change much after a cesarean:

While the risk of rupture in a spontaneous labor after one prior low transverse cesarean is comparable to other obstetrical emergencies, it is important for women weighting their post-cesarean birth options to know that their risk increased substantially due to their prior cesarean. It is important for them to understand the risks and benefits of VBAC vs. repeat cesarean. It is important for them to have access to accurate information and be able to differentiate between a midwife’s/blogger’s/doula’s/birth advocate’s/person on Facebook’s hopeful opinion vs. documented statistics.

I implore those who interact with, and have impact on, women weighing their birth options: do not pass along information, no matter how great it sounds, if you don’t have a well-designed scientific study supporting it. If you hear a statistic you would love to use and share, just ask the person who gave you this information,”What is the source?” and use the citation anytime you quote the statistic. But if the person doesn’t have a well-designed scientific study, be wary and don’t use the stat. This way, we can reduce the rumor and increase the amount of good information on the Internet. I know, a lofty goal.

I use the data in this same study to debunk the myth: the risk of uterine rupture is roughly double, or not much different, from an unscarred uterus. . . more dangerous information from what should be trusted sources.

Read more birth myths debunked including Lightning strikes, shark bites, and uterine rupture and Myth: Risk of uterine rupture doesn’t change much after a cesarean.

____________________________________________________

Notes: This study found that there were 183 ruptures after a prior cesarean and states in the abstract that this reflects a rate of 0.051% or 5.1 per 10,000 deliveries. But the problem is, they divided the number of uterine ruptures after a cesarean by the total number of women (with a prior cesarean and without.) It’s only towards the end of the study do they state the risk of uterine rupture in a woman after a prior cesarean is 0.64%. So, this is a little confusing and is another example of why reading the entire study, rather than just the abstract, is so important.

____________________________________________________

Verhoeven, C., Oudenaarden, A., Hermus, M., Porath, M. M., Oei, S. G., & Mol, B. (2009). Validation of models that predict Cesarean section after induction of labor. Ultrasound in Obstetrics & Gynecology, 34, pp. 316-321. Retrieved January 15, 2012, from http://onlinelibrary.wiley.com/doi/10.1002/uog.7315/pdf

Zwart, J. J., Richters, J. M., Ory, F., de Vries, J., Bloemenkamp, K., & van Roosmalen, J. (2009, July). Uterine rupture in the Netherlands: a nationwide population-based cohort study. BJOG: An International Journal of Obstetrics and Gynaecology, 116(8), pp. 1069-1080. Retrieved January 15, 2012, from http://onlinelibrary.wiley.com/doi/10.1111/j.1471-0528.2009.02136.x/full

entirety

Lightning strikes, shark bites & uterine rupture

When someone understates the risk of UR, I think it’s just as important the clarify as it is when someone overstates the risk. How else are women to make an informed decision? Just as it’s plain wrong for an OB to tell a woman with one prior low transverse cesarean that she has a 20% risk of rupture, it’s equally wrong when VBAC advocates say the risk is virtually non-existent.

Over the years, I have heard the statement: “You are more likely to be struck by lightning or bitten by a shark than experience uterine rupture!”

Today I’m going to get the statistics and run the numbers so you can see for yourself how the risk of these events compare.

Uterine Rupture

For this exercise, we will use the uterine rupture (UR) rate based on one prior low transverse (bikini) cut cesarean in a spontaneous labor determined by Maternal and Perinatal Outcomes Associated with a Trial of Labor after Prior Cesarean Delivery (Landon 2004):

Risk of uterine rupture: 1/240 or 0.4%
Risk of infant death or oxygen deprivation: 1/2000 or 0.05%

Lightning Strikes

Using the faulty theory I’m going to calculate the number of Floridians, since it is the “lightning strike state,” who would be struck by lightning.

Let’s assume that the risk of getting struck by lightning in Florida is the same as uterine rupture (even though the saying goes the risk is greater): 1 in 240 or 0.4%.

With 18,328,340 people living in Florida, that would mean that 76,368 people are struck by lightning every year in Florida. According to the CDC, that is more than the number of Americans who die annually from diabetes (72,449), Alzheimer’s disease (72,432), and influenza and pneumonia (56,326).

Using the National Weather Service stat that 10% of people struck by lightning die, we would have 7,636 people dying in Florida every year from lightning strikes. At that rate, you would have 209 people struck by lightning and 20 of those people dying every day in the state of Florida.

Now, I don’t live in Florida and I’m not an expert in lightning strikes, but that sounds like a lot of people dying.

Now let’s switch our assumptions and use the National Weather Service’s stats.

Odds of being struck by lightning in a given year (reported deaths + injuries) 1/700,000
Odds of being struck by lightning in a given year (estimated total deaths + injuries) 1/400,000

When we turn that fraction into a percentage, we get the following risk of being struck by lightening: 0.00025% – 0.00014%.

Using the National Weather Service’s statistics, we get 26 – 46 annual lightning strike related deaths or injuries in Florida.

Which sounds more reasonable to you? 26-46 Floridians struck annually by lightning or 76,368?

And that is assuming that the rate is the SAME as uterine rupture, but the rumor is that the rate of lightning strikes is HIGHER which means MORE than 76,368 Floridians are struck by lightning every year and more than 20 Floridians are dying daily from lightning strikes.

Now, does that pass the smell test? Does it seem reasonable in the least? It doesn’t to me.

Some would argue that in order to make the comparison, we need to eliminate the number of non-birthing people in Florida, but you really don’t because the lightning strike doesn’t know whether you are a man, woman, child, or menopausal. A Floridian women with one prior cesarean in spontaneous labor has the same risk as everyone else to be struck by lightning: 0.00025% – 0.00014%.

Shark Bites

From the Florida Museum of Natural History:

What are the chances of being attacked by a shark?

The chances of being attacked by a shark are very small compared to other animal attacks, natural disasters, and ocean-side dangers. Many more people drown in the ocean every year than are bitten by sharks. The few attacks that occur every year are an excellent indication that sharks do not feed on humans and that most attacks are simply due to mistaken identity. For more information on the relative risk of shark attacks to humans click HERE.

How many people are attacked each year by sharks?

Worldwide there is an average of 50-70 shark attacks every year. The number of attacks has been increasing over the decades as a result of increased human populations and the use of the oceans for recreational activity. As long as humans continue to enter the sharks’ environment, there will be shark attacks. For more information on shark attack statistics click HERE.

We have about 6.5 billion people on the world and 50-70 get bit by a shark annually which works out to 0.00000077% – 0.00000108%.

But this whole discussion is moot because it’s poor statistics to even compare these events (UR & lightning strikes or shark bites) because they are totally different types of occurrences.

The Actual Figures

This is a great chart from the Floria Museum of Natural History website entitled “A Comparison of Unprovoked Shark Attacks with the Number of Lightning Fatalities in Coastal United States: 1959-2008” where they show even in the state of Florida, over the past 49 years, there have been a mere 453 lightning fatalities and 585 shark bites. Remember that over 7,600 Floridans would be dying annually if the rate of uterine rupture was the same as the rate of lightning strikes.

Comparing Risks

There are some major problems when one is trying to compare risks of differing events.

One problem is when one uses a lifetime risk statistic as a means for comparison. You simply cannot take a statistic, like your lifetime risk of being struck by lightning (1 in 5000 which is significantly lower than one’s annual risk,) and compare that to your one-time risk of uterine rupture. If anything, using the annual risk of lightning strikes would be more accurate, but it still would be a false comparison.

An article by Andrew Pleasant entitled, Communicating statistics and risk, elaborates:

An oft-reported estimate is the lifetime breast cancer rate among women. This rate varies around the world from roughly three per cent to over 14 per cent.

In the United States, 12.7 per cent of women will develop breast cancer at some point in their lives. This statistic is often reported as, “one in eight women will get breast cancer”. But many readers will not understand their actual risk from this. For example, over 80 per cent of American women mistakenly believe that one in eight women will be diagnosed with breast cancer each year.

Using the statistic ‘one in eight’ makes a strong headline but can dramatically misrepresent individual breast cancer risk.

Throughout her life, a woman’s actual risk of breast cancer varies for many reasons, and is rarely ever actually one in eight. For instance, in the United States 0.43 per cent of women aged 30–39 (1 in 233) are diagnosed with breast cancer. In women aged 60–69, the rate is 3.65 per cent (1 in 27).

Journalists may report only the aggregate lifetime risk of one in eight because they are short of space. But such reporting incorrectly assumes that readers are uninterested in, or can’t comprehend, the underlying statistics. It is critically important to find a way, through words or graphics, to report as complete a picture as possible.

Take away message: Be extra careful to ensure your readers understand that a general population estimate of risk, exposure or probability may not accurately describe individual situations. Also, provide the important information that explains variation in individual risk. This might include age, diet, literacy level, location, education level, income, race and ethnicity, and a host of other genetic and lifestyle factors.

The second major problem is often the two things you are comparing are so different that the comparison is worthless. Again, I defer to Mr. Pleasant:

Try not to compare unlike risks. For instance, the all-too-often-used comparison ‘you’re more likely to be hit by a bus / have a road accident than to…’ will generally fail to inform people about the risks they are facing because the situations being compared are so different. When people assess risks and make decisions, they usually consider how much control they have over the risk. Driving is a voluntary risk that people feel (correctly or not) that they can control. This is distinctly different from an invisible contamination of a food product or being bitten by a malaria-carrying mosquito.

Comparing the risk of a non-communicable disease, for example diabetes or heart disease, to a communicable disease like HIV/AIDS or leprosy, is similarly inappropriate. The mechanisms of the diseases are different, and the varying social and cultural views of each makes the comparison a risky communication strategy.

Take away message: Compare different risks sparingly and with great caution because you cannot control how your audiences will interpret your use of metaphor.

Going Forward

It can be hard when wading through the (mis)information available on the internet about VBAC, but here are some tips to help you out.

1. Always find the source – If you find some great statistic, but there is no source referenced, be wary.

2. Verify the statistic – If there is a source listed, read through it. If there is no source listed, do a quick Google search. It didn’t take me long at all to find all the statistics in this article and run the math.

3. Leave a comment – If you find something on the internet that doesn’t pass the smell test, leave a comment on the blog or email the author asking for the source.

4. Be careful about forwarding things – There is so much misinformation on the internet, so do your friends a favor and don’t forward them emails or articles unless you have verified the information to be true. That is one way to quickly nip falsehoods in the bud!

For further reading on using statistics, check out, Correlation and Causation:Misuse and Misconception of Statistical Facts and Risk Communication, Risk Statistics, and Risk Comparisons: A Manual for Plant Managers