We write in response to the research published by Werner et al in PLOS Medicine on 27 September 2016 (1). The authors have compared birth outcomes in women who started their pregnancy care with self-employed midwives (paid for by taxpayer contributions) compared with those who booked with doctors, predominantly private obstetricians who charge a fee for service. The findings were that women who booked for pregnancy care with a self-employed midwife had an increased risk of their babies being born in poor condition with asphyxia or neonatal encephalopathy compared with those in the medical model.
This was an observational study, excluding women cared for by public hospital medical and midwifery staff, linking maternity carer claims and hospital discharge datasets. Confounding due to socioeconomic differences between the compared groups was acknowledged (women booking with a self-employed midwife differ in many ways from women who pay for a private obstetrician) and an attempt made to adjust statistically using maternal BMI, smoking, age, parity, medical risk factors, trimester at booking for pregnancy care, and a population level variable for social deprivation.

The authors have interpreted these adjusted results to be evidence of a causal relationship between the model of care and the increased neonatal encephalopathy observed in women who booked for care with a self employed midwife. We do not agree that causality can be concluded from the findings of this study for a number of reasons.
In an observational study it is generally considered unwise to draw causal relationships, even after adjustment, as it is imposssible to control for all the factors that differ between such disparate groups and it is likely that residual confounding will still be present (2). Furthermore the odds ratios in this study are small and the confidence intervals almost include one. There are many examples of observational studies finding relatively small ORs (0.5-2) which have since been shown to be erroneous when randomised controlled trials are conducted. Grimes and Shultz write “In general, unless RRs in cohort studies exceed 2 to 3 or ORs in case control studies exceed 3 or 4, associations in observational research findings should not be considered causative. Specifically, RR between 0.5 and 2 is a zone of potential bias” (2).

This paper did not report information about the mode of birth and admission of babies to neonatal units. This information would assist with a fuller understanding of differences and possible trade-offs between the two models of care. This is a particular issue as many births conducted by private obstetricians occur by elective Caesarean section, prior to the onset of labour, and thus are not at risk of the outcome of interest here. These issues are highlighted by data from National Women’s Health, a large hospital that provides maternity services to the central Auckland area. National Women’s is the birthing facility for 42% of all New Zealand’s private obstetrician births. In this service, the elective Caesarean rate at term from 2006-2015 for doctor-led care was 32.8%, and for self employed midwives was 7.4%. It is likely that this large difference in elective Caesarean rate between maternity provider groups is not explained by clinical need but by the woman’s choice of provider (and by extension of mode of birth). Our view is that women with elective Caesarean births should be removed from the denominator when the outcome of interest is intrapartum hypoxia. This marked difference in elective Caesarean rates is another reason why it is important to report other outcomes such as neonatal unit admission between the two groups.
We welcome the call for further research. The negative publicity from this paper has been unfortunate with sensational headline statements leading to widespread criticism of New Zealand midwives resulting in loss of morale in the workforce and concern by pregnant women. In our view this is most undeserved.


Professor Cynthia Farquhar, Postgraduate Professor of Obstetrics and Gynaecology, University of Auckland, New Zealand.

Professor Lesley McCowan, Professor of Obstetrics and Gynaecology,Head of Department of Obstetrics and Gynaecology, University of Auckland, New Zealand.

Dr Sue Fleming, Director National Women’s Health, Auckland City Hospital

References
Ellie Wernham , Jason Gurney, James Stanley, Lis Ellison-Loschmann, Diana Sarfati.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 http://dx.doi.org/10.1371...

Grimes DA1, Schulz KF. False alarms and pseudo-epidemics: the limitations of observational epidemiology. Obstet Gynecol. 2012 Oct;120(4):920-7.