In Hans Christian Andersen’s tale of the Emperor’s new clothes no one dares to say they don’t see a suit of clothes on him for fear they will be seen as stupid and incompetent. It takes the cry from a small child, “but he isn’t wearing anything at all”, to identifying the farce being carried out.
Sometimes research papers are put out with misleading media releases and political agendas that go unquestioned by a media hungry for controversy and the next sensational headline. In this blog we will identify the naked Emperor in the form of the recent New Zealand paper (NZ) published by (2016), titled A Comparison of Midwife-Led and Medical-Led Models of Care and Their Relationship to Adverse Fetal and Neonatal Outcomes: A Retrospective Cohort Study in New Zealand. The Wernham paper caused consternation around the globe with doctors waving it in triumph pretending the Emperor had a magnificent outfit on while midwives scrambled to understand what was happening, crying amidst the crowd, “but he isn’t wearing anything at all.”
How did something that was fairly low level scientific evidence get more attention, and lead to such public questioning of the safety of midwifery care, than 15 randomised controlled trials and a (CSR) on this issue?
Just a reminder about the Level 1 evidence of continuity of midwifery from over 17,000 women randomised in 15 separate RCTs:
“This review suggests that women who received midwife-led continuity models of care were less likely to experience intervention and more likely to be satisfied with their care with at least comparable adverse outcomes for women or their infants than women who received other models of care. Further research is needed to explore findings of fewer preterm births and fewer fetal deaths less than 24 weeks, and all fetal loss/neonatal death associated with midwife-led continuity models of care.”
How did we ever think the Emperor had new clothes?
The first alert in this recent saga is the media release that came out from the first author’s university, strictly embargoed beforehand to excite the ‘crowd’ awaiting the emperors arrival. The media release revealed the first bias in the authors’ agenda and was the ultimate hook for the media:
“Mothers using autonomously practising midwives throughout their pregnancy and childbirth are more likely to have adverse outcomes for their newborns than those who use obstetricians, according to a retrospective study of nearly a quarter million babies born in New Zealand published in PLOS Medicine by Ellie Wernham of University of Otago, New Zealand, and colleagues.”
Firstly, this study was never about midwifery care during childbirth, or pregnancy for that matter. Midwives also look after women cared for by private obstetricians so this care is never just about medical care just as it is never just about midwifery care. Secondly, there was no statistical difference in perinatal mortality. You would have hardly known this from the media reports. Thirdly, the authors were clearly data dredging when they combined Intrauterine hypoxia, birth related asphyxia and neonatal encephalopathy in order to get a highly significant outcome. Rare adverse events and small numbers were sensationalised in the media release (“55 percent lower odds of birth related asphyxia, 39 percent lower odds of neonatal encephalopathy, and 48 percent lower odds of a low Apgar score at five minute after delivery”). Neonatal encephalopathy occurs 1-2 in 1000 births and is a rare event. Presented this way makes it sound so dramatic and it takes only one or two cases to change the outcome.
Why the Emperor is actually naked
The authors were unable to look at actual care during childbirth because they don’t appear to have this data, so they took model of care at booking and then misled the media and public that this was an indication of care at birth, when it was not. The problem with this is while all women who book with private obstetricians will remain under the care of private obstetricians from booking to birth, between 30-35% of women under midwifery care will be referred during pregnancy to a doctor. Despite this fact all outcomes (only adverse perinatal ones) in the paper are reported as due to midwifery care, when they are clearly not.
One could argue that the randomised controlled trials (RCTs) of continuity of midwifery care reported in the use a similar method – that is model of care on booking and intention to treat analysis. However, the difference is randomisation reduces selection bias and the study groups should be as similar as possible at the outset so the researchers can isolate and quantify the effect of the intervention they are studying (in this case midwife or medical care). In a RCT you can see what care women got and you would also know the mode of birth and maternal outcomes, which are not reported in this study. RCT’s can be used to change practice but lower level evidence should not; yet that has not stopped groups such as the calling for this in Australia.
The NZ study had several concerning limitations that were not adequately considered in the unfolding debate:
1. One of the most significant findings of the CSR of continuity of midwifery care was the 24% reduction in preterm birth under midwifery care. There was also a significant reduction in perinatal mortality. Only women over 37 weeks were included in the recent NZ study, so there was no chance to see whether this important effect was seen in this study.
2. Not only are of long term outcomes but there were a large number of missing Apgar scores and this was greater for women who booked with obstetricians.
3. The inclusion of women more than 42 weeks, which were seen in larger numbers in the midwife booked group and are more likely to have stillbirths associated with prolonged pregnancies, is concerning. If the authors took 37 weeks gestation as a cut-off to exclude preterm birth (higher risk), why not take 41+6 to exclude the higher risk post-term pregnancies. It would have been very interesting to know how many adverse events were seen in the post-term group. Women choosing midwifery care are more likely to not want to be induced and to go over 42 weeks, as is seen in this study.
4. The inability to separate antepartum stillbirth from intrapartum stillbirth is critical in trying to assess the impact of birth provider on outcomes and this could not be done, despite the study protocol suggesting it would be.
5. In the study protocol published with the paper neonatal nursery admissions were examined but not reported. When we look at the author’s Master’s thesis where this information is available, more neonatal admissions are reported for babies born to women who booked with private obstetricians. This was not reported in this paper. One has to ask, why?
6. In the first author’s Master’s thesis (where this study originally came from), substantially lower rates of caesarean section (22% vs 32.9%) and instrumental birth rates (9% vs 12.3%) are reported for women who booked with midwives, leading to significantly less maternal morbidity. Again this was not reported, giving a very one-sided view considering the authors are virtually questioning the entire NZ maternity system.
7. There appears to be quite a bit of missing data in this study and it is unclear how this was dealt with in the analysis.
8. Many socio demographic variables are not accounted for (e.g. alcohol and drug use), and others such as smoking are notoriously underreported. Midwives tend to look after women with greater socio demographic disadvantage and mental health issues. None of this is adjusted for.
9. Other medical complications that arise following booking, such as gestational diabetes, pre-eclampsia, etc are not accounted for and may be increased in women who book with midwives due to ethnicity factors, life style etc.
11. There is no difference in PMR between Australia and NZ despite the fact that 30% of care in Australia is by private obstetricians whilst in NZ around 90% of women have a midwife as a lead care provider.
12. A previous NZ paper that also hit the media headlines in recent times, purporting to show the risk of perinatal death was higher when midwives were in their first year following graduation, has recently been questioned by the who have been unable to replicate the study. This is worrying.
13. of low risk women in NSW who had a birth in a private hospital under private obstetric care with low risk women who had a birth in a public hospital with midwife/medical care we found greater morbidity for women giving birth in a private obstetric model of care.
The political fallout from this paper has been extraordinary, for it actually tells us very little. No practice changes could ever be made based on this study. The Emperor may have no clothes, but the delusion has been maintained by a misleading media release, politically motivated reporting of findings by the authors, a hungry unquestioning media sensing blood in the water and wanting sensational headlines, and obstetricians determined to drag the advances made by the profession of midwifery back to the ‘good old days’ when they were compliant handmaidens.