Bringing ICAN’s 25-year-plus tradition of support and education in the mother-to-mother and sister-to-sister model into the internet age, we have invited passionate bloggers to join us around our virtual circle of women. We hope to introduce you to new voices that you have not heard before, and also to respected voices that will already be well-known to you.
Today we welcome Henci Goer is an award winning medical writer and speaker. She has published two books: The Thinking Woman’s Guide to a Better Birth and Obstetric Myths Versus Research Realities (a new edition of which is in press).
Krista Cornish Scott, ICAN’s education director, asked me take a look at a study she found on one of the pro-cesarean websites, and I was happy to oblige.
Dahlgren LS, von Dadelszen P, Christilaw J, et al. Caesarean section on maternal request: risks and benefits in healthy nulliparous women and their infants. J Obstet Gynaecol Can 2009;31(9):808-17.
Objective: To determine the risks and benefits of an elective Caesarean section (CS) at term in healthy nulliparous women. Methods: We conducted a population-based cohort study of deliveries between 1994 and 2002. Using bivariate and multivariable techniques, we compared maternal and neonatal outcomes in healthy nulliparous women who had undergone elective pre-labour CS (using breech presentation as a surrogate) with those in women who had undergone spontaneous labour with anticipated vaginal delivery (SL) at full term.
Results: There were 1046 deliveries in the pre-labour CS group and 38 021 in the SL group. Life-threatening maternal morbidity was similar in each group. Life-threatening neonatal morbidity was decreased in the CS group (RR 0.34; 99% CI 0.12 to 0.97). Subgroup analysis of the SL group by mode of delivery demonstrated the increased neonatal risk was associated with operative vaginal delivery and intrapartum CS but not spontaneous vaginal delivery.
In the discussion section the authors write of the gap in neonatal morbidity:
We do not suggest it is necessarily the mode of delivery itself causing the neonatal morbidity. The increased morbidity observed in these groups is likely associated with the indication for the operative vaginal delivery or intrapartum CS.
and they conclude:
Further research is needed to better identify women with an increased likelihood of an operative vaginal delivery or intrapartum Caesarean section, as this may assist pregnancy caregivers in decision-making about childbirth.
This was disconcerting. Several of the authors are researchers whose work I know and respect, and, until this study, their work supported physiologic care. Why is this study not like their others? What happened here?
What happened was an illustration of how medical model thinking can unconsciously bias even good researchers. These investigators assembled an ultra-low-risk group of nulliparous women (term, singleton, head down baby, no medical complications, spontaneous labor onset), found that only 63% had a spontaneous vaginal birth (22% had an instrumental vaginal delivery and 15% had a cesarean, of which 2% also had an attempt at instrumental delivery), and assumed this appalling statistic had to be due to problems originating in the women. I beg to differ; my first thought was, “What an indictment of medical model management!”
The authors may dismiss the idea, but typical obstetric management such as epidural analgesia, continuous electronic fetal monitoring, and preset time limits for making progress are known to inflate instrumental and cesarean delivery rates or both, and serious neonatal complications can result from labor management, for instance, as Amy Romano and I will document in the new edition of Obstetric Myths Versus Research Realities, I.V. fluid overload can cause overly rapid breathing (transient tachypnea), narcotics cause need for resuscitation, high-dose oxytocin (AKA “active management of labor”) can cause seizure, instrumental vaginal delivery can cause intracranial hemorrhage. In fact, with physiologic care, maternal outcomes might have looked better too. For example, 6% of the women suffered anal sphincter lacerations, 5% with spontaneous vaginal birth and 15% with instrumental delivery. With optimal care, that number could have been 2% . Instead of looking for hidden defects in healthy first-time mothers, I say let’s look at defects in their care providers. Numerous studies of practice variation, including some done by some of these very same researchers, document that the best way to “decrease likelihood of an operative vaginal delivery or intrapartum Caesarean section” is to choose a care provider with a high spontaneous vaginal birth rate. To quote Childbirth Connection’s critique of the 2006 NIH conference on elective first cesareans, “We should not be asking healthy women to choose between vaginal birth with avoidable harms and birth by major abdominal surgery.”
The study also has another major problem: while we have 38,000 women in the spontaneous labor onset group, we have only 1050 women in the cesarean group, far too few to show differences in occurrence of severe or life-threatening morbidity. The rare, surgery-related catastrophe is a crucial consideration when performing elective surgery on a healthy woman. For example, a U.S. study of millions of babies found that elective cesarean was associated with an excess neonatal mortality rate of about 1 in 1000. Comparing 1050 planned cesareans with 38,000 planned vaginal births is like comparing 1050 smokers to 38,000 nonsmokers and deciding smoking is OK because analysis didn’t find differences in lung cancer or emphysema, when moreover, as already noted, all of the nonsmokers were exposed to second-hand smoke. And this is without considering, as the study authors themselves acknowledge, that the first cesarean increases risks in future pregnancies and births. To repeat, the no-risk strategy is not to cherry pick candidates for vaginal birth, but to institute care that supports the natural process and intervene medically only when lesser measures have failed and the benefits of intervening clearly outweigh the risks.
To give them their due, investigators have done a fine job collecting and analyzing their data, but as Hall and Menticoglou write, “One of the most influential biases in the acquisition of evidence is choice of the question, and the best evidence in answer to the wrong question is useless” (p. 488). I contend that “What are the benefits of performing major surgery on healthy women?” is the wrong question.
Finally, OT, but another example of how the medical model distorts thinking, in the process of justifying why elective cesarean for breech is a good proxy for elective cesarean at maternal request, the authors write, “Since the results of [the Term Breech Trial] were published, 87% to 97% of women throughout the world have chosen CS for delivery of a breech-presenting infant,” which they immediately follow by, “In Canada, the percentage of physicians offering vaginal breech delivery decreased from 84% before the trial to 14% after the trial.” Newsflash: you can’t be said to make a choice when you only have one option.
Henci and Amy Romano are in the final stages of completing the manuscript for a top-to-bottom new edition of Obstetric Myths Versus Research Realities, to be published by University of Michigan Press. Look for it late in 2010. In the meantime, you can find Henci at her “Ask Henci” forum and Amy at her blog, Science and Sensibility.