I’ve said it before and I’ll say it again: No discussion of reproductive justice is complete without attention to birthing and birthing care.
A blithe, shallow, naive focus on “reproductive choice” is not enough. We don’t have free choice in birthing care right now, and there’s no point in pretending that we do.
Truly free choices are almost impossible within a societal and medical patriarchy in which birthing is considered a stupendously dangerous, messy, primitive, terrifying process which must be timed and controlled and scrutineered and interfered with in the normal course of things. Woman don’t have free choice when they’re being coerced, threatened with court orders, restrained and railroaded into “active management” of birth and surgical procedures. Women don’t have free choice when the “dead baby card” is being simplistically played, over and over, whenever they raise questions about the necessity of interventions – while risk to the woman and fetus/infant from interventions is dismissed or ignored. Women will not have free choice so long as medical unions collude with government to marginalise and suppress midwifery care. Women don’t have free choice when a standard-issue hospital birth is the only birth they can access or afford. And women don’t have free choice when they know that any mention of medical rapes[link trigger warning] will only be met with shocked silences, hasty lectures about how they should be grateful for their baby, and referrals for ‘counselling’.
Now that I’ve got that off my chest, on to the actual topic, my journalwatch post du jour: “Neonatal Mortality for Primary Cesarean and Vaginal Births to Low-Risk Women: Application of an “Intention-to-Treat” Model“. Marian F. MacDorman, Eugene Declercq, Fay Menacker, Michael H. Malloy. Birth, Volume 35 Issue 1 Page 3-8, March 2008.
You might have noticed this headline in the Sydney Morning Herald last week: “Death twice as likely by caesarean”:
BABIES born by elective caesarean are almost 2½ times more likely to die within their first month than babies born vaginally, researchers have found, adding weight to the argument that caesareans should only be carried out in emergencies.
The study, which involved more than 8 million births in the US over four years, is the first of its kind to focus on full-term babies born to women with no medical reason for choosing a caesarean over a vaginal delivery, an increasingly common phenomenon in Australia.
The study, published in the international journal Birth, only included women who had not had a previous caesarean; were giving birth to a single baby which was head down in the cervix; were between 37 and 41 weeks gestation and had none of the 16 common risk factors, such as diabetes or hypertension, associated with birth complications, in a bid to ensure that only low-risk births were evaluated.
It found the mortality rate for babies born vaginally was less than one in 1000 births while the rate for elective caesareans was 1.73 per 1000.
The obstetrician consulted for ‘balance’ by the SMH reporter pooh-poohed the increased mortality as not thaaaaat high:
A professor in obstetrics and gynaecology at the Australian National University medical school, David Ellwood, said the risk was still very small. “We have enough evidence now to know that caesareans should only be done when there is a medical indication, but when you look at the overall risk here, it is not that high.”
Is the number really “not that high”?
There are 26 000 births yearly in Western Australia alone. Taking the most charitable view, the discreprancy between the maximum C section rate recommended by the World Health Organisation (15%) and the actual C section rate (34% and rising in 2005) accounts for nearly 5 000 non-medically-necessary C sections in WA each year. Going by the numbers here, there may be an excess mortality of around 4 babies every year in WA from unnecessary C sections – and we’re one of the smaller states, accounting for around one-tenth of the country’s population. So we could be talking about around forty dead babies, every year. Not that high? Is that worth playing the “dead baby card”?
Total perinatal deaths are around 10 per 1000, most of them non-preventable, so a number approaching one-tenth of the total perinatal mortality rate is very much non-trivial. Forty percent of those 10 per 1000 perinatal deaths are low birthweight and premature babies, and one-third of the perinatal deaths are associated with lethal birth defects. There aren’t many deaths left after subtracting those two categories. Which is exactly what the authors of this study did.
[The numbers for women dying unnecessarily may be the same or higher; studies are tending to find a doubling or tripling of maternal mortality with C section, but there is a fair bit of variation in the figures, and it’s not my major focus in this post, though it is a huge issue in itself.]
The study itself
The SMH’s reporting isn’t 100% accurate, unsurprisingly. This month’s issue of “Birth” is available full-text online, and the study is here:
The study possibly addresses selection bias possibly as well as can be done without a prospective randomised controlled trial of C section vs vaginal birth for normal pregnancies. [How any such trial would pass an ethics committee, given the mounting data on the risks of C section to mother, baby, and future pregnancies, is beyond me.] The methodology was recommended by a National Institutes of Health conference.
First, a note on methodology, for those who aren’t familiar with the “intention-to-treat” concept. It is not enough to take all the C sections, and all the vaginal births, and add up the deaths for each to get the number of deaths ’caused’ by C section. That would be falling hard into the ‘correlation does not equal causation’ trap. C sections are more likely to take place after complications have occurred with a pregnancy or birth, so you would expect to see a higher perinatal mortality associated with C section than with vaginal birth. An ‘intention to treat’ analysis bases the categories on the inferred initial intention: is there evidence that the mother and carer planned for a vaginal birth, or for a C section, in the first place? This, an absolutely key feature of the study design, went completely over the head of the SMH journalist.
The authors analysed data from over eight million births in the USA from 1999 to 2002. Only singleton, term, vertex (head-first), births with no reported medical risk factors, no placenta previa, and with no prior cesarean section were included.
The “planned vaginal delivery” category was formed by combining vaginal births and cesareans with labor complications or procedures since the original intention in both cases was presumably a vaginal delivery.
The “planned C section” category was a group who underwent C section with no labor complications or procedures.
The medical risk factors in the exclusion criteria included anemia, cardiac disease, acute or chronic lung disease, diabetes, genital herpes, hydramnios/oligohydramnios, hemoglobinopathy, chronic hypertension, pregnancy-associated hypertension, eclampsia, incompetent cervix, previous infant 4,000+ g, previous preterm or small-for-gestational-age infant, renal disease, Rh sensitization, uterine bleeding, and ‘other’. The results were controlled for sociodemographic factors (birthweight, race, parity, smoking, education) and the presence of congenital malformations.
It is particularly important to note here that the “planned vaginal birth” category did not only include babies who were ultimately born vaginally. The ‘planned vaginal birth” group included all C sections where a labour complication or procedure was recorded, indicating that vaginal birth had been planned but the plan had changed in labour for a medical reason. (This was 8% of the PVB group.) The idea is that the ‘planned C section’ group isn’t having its stats muddied with C sections after complicated labours, and the ‘planned vaginal birth’ group isn’t having its stats sanitised by only including those in which the birth happened to go swimmingly.
From the abstract:
Results: The unadjusted neonatal mortality rate for cesarean deliveries with no labor complications or procedures was 2.4 times that for planned vaginal deliveries. In the most conservative model, the adjusted odds ratio for neonatal mortality was 1.69 (95% CI 1.35–2.11) for cesareans with no labor complications or procedures, compared with planned vaginal deliveries.
Conclusions: The finding that cesarean deliveries with no labor complications or procedures remained at a 69 percent higher risk of neonatal mortality than planned vaginal deliveries is important, given the rapid increase in the number of primary cesarean deliveries without a reported medical indication.
For those not familiar with 95% confidence intervals, the null hypothesis (“planned C section is associated with no change in perinatal mortality”) would only be considered unfalsified if the confidence interval included 1.0. With a leetle bit of simplification, the statisticians are 95% confident that the ‘real’ odds ratio lies within that confidence interval, 1.35-2.11.
The “most conservative model”, as well as controlling for medical and sociodemographic factors and congenital malformations, also excluded all those infants with an Apgar of less than four, including excluding states who don’t record Apgars on birth data. This was because intrapartum hypoxia might also be both a reason for performing a cesarean section and a contributor to infant death – an attempt, perhaps to detect under-reporting.
The biggest potential flaw, as I see it? Under-reporting of labour complications. Though why they would be disproportionately under-reported in the group proceeding to C section and not under-reported in the group proceeding to vaginal birth, I can’t quite figure. Anyone? The researchers didn’t just depend on the absence of data entry on complications; they also looked for ticks in the “no complications” and “no medical risk factors” checkboxes.
Now, this data isn’t perfect. It is, however, suggestive; it is based on a large and comprehensive data set, and performed in a population and an environment fairly comparable to Australia’s (in demography, health risk factors, medical technology and C section rates). It is thoroughly physiologically plausible. The OB professor quoted in the article waved it away as being a trivial number (something I debunked above), not as being seriously flawed in design.
Now what are we going to do about it?
Background: If you haven’t been following my recent posts and comments about birthing, this will catch you up:
BB Bounce (see comments)
This subthread on ‘coached pushing‘.
Background on “Deliver big babies early!” eejitry: “macrosomia” data