Imagine if we did a study on triple bypass heart surgery and divided patients into three groups. The first group contains people who have no heart trouble and don't have surgery. The second group contains people who have no heart trouble but have surgery anyway. The third group contains people who have unstable angina and undergo surgery. Guess which group would have the best health outcomes. Not surprisingly it would be the group who have no heart trouble and don't undergo surgery. That's because they were healthiest to begin with and were not exposed to the risks of surgery.
That's pretty much the study that the World Health Organization did on C-section in Asia, Method of delivery and pregnancy outcomes in Asia: the WHO global survey on maternal and perinatal health 2007—08. They compared three groups: women who had no reason for C-section and underwent vaginal delivery, women who had no reason for C-section and had a C-section anyway (subdivided into antepartum and intrapartum C-sections), and women who had medical necessitating C-section and underwent a C-section (subdivided into antepartum and intrapartum C-sections). They also include a group for women who had operative vaginal delivery (forceps or vacuum) though they did not specify whether there were medical indication. It would hardly be surprising if the group that had no medical complications and underwent uneventful vaginal delivery would be the group with the best neonatal outcomes. And that would not be an indication that C-section was dangerous or inappropriate because we would expect that women experiencing complications would have higher rates of neonatal mortality.
What is surprising is what the WHO researchers actually found: the group of babies with the lowest neonatal mortality were born to women with NO reason to have a C-section but who had one anyway! In fact, the babies of women who had C-sections without a medical indication had the best results on every possible indicator.
Tthe World Health Organization is insisting that the interpretation of this study is simple:
To improve maternal and perinatal outcomes, caesarean section should be done only when there is a medical indication.
But that's certainly not what the data on perinatal outcomes shows. How about maternal outcomes?
Both blood transfusion and ICU admission were increased in the women who underwent C-section without medical indication, but the rate of hysterectomy was zero and the rate of maternal death was zero.
How the WHO researchers can justify their conclusions based on the data that they gathered? More to the point, what exactly did they think they were studying? If we want to find out the safety and efficacy of a procedure (like C-section) we identify specific conditions or risk factors and divide patients into the treatment group (C-section) and the control group (vaginal delivery). We certainly don't compare all women who had C-sections with all women who had vaginal deliveries because they are going to differ in very important ways.
This is a poorly designed study that can't possibly yield any valid results. The authors compounded their error by misinterpreting (basically ignoring) the results that they got and instead reached a conclusion decrying C-section that seems to have been determined before the study began.
TIP:
Enter the amount, and click "Tip" to submit!
MissingK8
aliquot
Roger Fallihee
madcelt
Norwonk
John Moore
Salon.com
Comments
The women who had C-sections with no medical indication had a lower risk of death and a lower risk of hysterectomy, the two most serious complications. The women who had C-sections with no medical indications also had the lowest risk of neonatal death and the lowest risk of fetal death, the two most important outcomes.
The authors tried to obscure these finding by burying them within a perinatal “index” and a maternal M&M “index”.
As you know, an index of outcomes can be useful if there are a great deal of possible complications, all of approximately the same severity. So an index might have been useful had the possible complication of surgery been for example: transfusion, ICU admission, wound infection, IV antibiotics, hospital readmission, etc.
An index is inappropriate if there are only a few outcomes (as in this paper) or if it combines outcomes of wildly different severity (as in this case). In such situations and index is often use to obscure outcomes. The WHO group desperately wanted to conclude that vaginal delivery was “safer” than C-section with no indications. The data, of course, show otherwise. There were fewer hysterectomies and maternal deaths in the C-section group, so they added in transfusions and ICU admissions to create an “index” that made vaginal delivery look better.
There were fewer neonatal and fetal deaths so they added in NICU stay to create an "index" that made vaginal delivery look better.
The fundamental question ought to be: what method of delivery ensures the safest outcomes for the baby while minimizing serious morbidity to the mother? If that's the question, the data in this study indicate that C-section without maternal indications is safest.
When you say "WE" I take it you are including yourself, as either a practicing physician, or one of the WHO study team.
I (as in ME) doubt that either of those is the case. BTW, the charts are illegible Amy.
Just thought I's toss that out there.
"I'm just tossing this out but I'd guess that women who elect to have a c-section (i.e. plan it that way in advance) are also likely to have had superior (or at least better than average) health care in general and certainly a high degree of pre-partum care, otherwise no C-section plans would have been made."
That's a possibility. There are many important ways in which they may differ, such as underlying medical illnesses, pregnancy complications, race, income; any or all of these might effect the outcome.
It's like doing a study on triple bypass and finding out that people who don't have heart disease live longer than people who have life threatening heart disease and require a coronary bypass. That wouldn't tell us what we really need to know: if you have life threatening heart disease, is triple coronary bypass safer than other possible treatments like balloon angioplasty or medication.
In order to get valid results, the WHO group would have been better served by comparing women who had various medical reasons for interventions and comparing those who had C-sections for the interventions with those who didn't. For example, they could have compared those women who had breech babies whose babies were delivered by C-section to women whose breech babies were delivered vaginally. That way, the two groups are likely to be similar and the only difference between them is whether or not they had a C-section.