I’m collaborating on a study of traumatic childbearing experiences, so I’ve been thinking a lot about the types of injuries that can occur as a result of pregnancies, labors, and different types of deliveries.  My research partner forwarded me a recent blog post on a board-certified urogynecologist’s website titled “Cesarean on Demand Does Not Eliminate Risk of Prolapse.”  This post highlights the findings of a 2009 research article published in the International Urogynecology Journal. In this research, three groups of women were studied: “vaginal delivery with sphincter tear (n = 106), vaginal delivery without sphincter tear (n = 108), and cesarean without labor (n = 39).” [The numbers reflect how many women were in each of the three groups.]

Now, I’m no urogynecologist, but I found it hard to believe that c-sections “on demand” (a.k.a. without labor) would not at least reduce the risk of pelvic floor damage, including pelvic organ prolapse (pelvic organs “slipping out of place” when the supportive muscles and ligaments are weakened or torn).  Prolapse can greatly impact the health and quality of life: for example, women with prolapse may suffer one or both types of incontinence and/or painful sexual intercourse.

I recognize that many medical practitioners, authors and laypeople have come to believe we have too high a rate of c-sections here in the U.S.  Research studies, such as the one celebrated in the recent blog post, call into question whether there are any health benefits of c-sections without labor.  As a medical sociologist who teaches research methods, I consider it to be of utmost importance to discuss research findings with the highest degree of accuracy.  No study is perfect: no study is without bias and no study is without limitations.  So, I read the complete research article to find out if it truly supported the blog author’s contention that these researchers “found NO DIFFERENCE in moderate prolapse between the three groups.” 

I was struck by significant methodological flaws and limitations which, while acknowledged by the authors of the original article, were glossed over or flat out ignored by the author of the blog post.  I found myself asking several questions:

Question #1: how healthy were the women before this childbearing experience?  No one knows: the researchers admit, “our findings cannot be attributed with certainty to delivery method, since some women may have developed prolapse before delivery or pregnancy and prolapse was not assessed prior to delivery in this population.” 

Question #2: who were the women who participated in this study?  The women for this study were recruited from prior studies performed through the Pelvic Floor Disorders Network, specifically from the follow-up study to their CAPS Study (which focused on “fecal and urinary incontinence after childbirth”).  How can we rule out a self-selection bias of those women who said “yes” when they were recruited to this initial study?  Could it be certain women who had C-sections, perhaps those feeling some pelvic/vaginal discomfort immediately following delivery were more likely to say “yes” because they saw value in being interviewed about incontinence?

Question #3: did the researchers recruit enough women for each of the three groups to be able to answer their main question?  No.  The authors wrap up their article by noting that “further research would be required to determine whether cesarean delivery before labor reduces the incidence of pelvic organ prolapse.”  So, this research doesn’t actually determine anything about what they claim as their primary research question.  Why not?  The short answer is that they never got enough women to participate.  The authors claim that they would have needed 132 women per group in order to test the statistical significance of the difference in rates of stage II prolapse between those women who had C-sections without labor and those women who had vaginal deliveries.  While they got reasonably close to their sample size goal of 132 for the two vaginal delivery groups (106 and 108), they only got 39 women to participate in the C-section group.  Is this acceptable?  Statistical significance is key to evaluating any study because it means that the results are “probably true (not due to chance).”  The researchers finally own up to the likely irrelevance of their study towards the end of the published journal article: “Furthermore, our sample size was not sufficiently large to exclude a significant difference between groups.”  In plain language: they didn’t study enough women to know whether or not there are not real differences between the health outcomes for women who have c-sections without labor and those experience other types of labor and delivery.

Question #4: can the researchers say anything definitive that might help improve women’s health?  Hmmm.  The only factor they definitively connect with less pelvic floor damage is lower birth weight: I’m betting that it won’t surprise many to find out that smaller babies causes less damage.  But, what are we supposed to do with this finding? 

The author of that recent blog post dares to call it a “beautifully executed study,” and that’s why I had to wrote this post: to help those of us who are not medical researchers better understand what we should value and what we should question when it comes to research studies that can impact women’s health.