Study shows New Zealand maternity system needs improvement
Large Otago study shows New Zealand maternity system needs improvement
Lead author, Ellie Wernham
Researchers at the University of Otago, Wellington, have released the findings of a major five-year study of New Zealand’s maternity outcomes.
The research, published in PLOS Medicine (Wednesday 28 Sept, at 7am), compares birth outcomes for babies born to mothers registered with medical lead maternity carers, such as obstetricians or GPs, with those who had midwives as lead maternity carers.
The study found that adverse health outcomes were substantially lower in the medical-led births group compared with the midwife-led group.
New Zealand adopted an autonomous midwife-led model of maternity care in 1990. This is the first detailed study examining what effect, if any, midwife-led care has on specific health measures for newly born infants in New Zealand.
The authors say the findings need to be interpreted in the context of New Zealand’s good overall internationally-comparable birth outcomes, and in context of previous research that supports the many benefits of midwife-led care such as greater patient satisfaction and lower intervention rates.
“As a practising midwife I saw first-hand the many benefits of our midwife-led continuity of care model. However our study has identified that there may be aspects of our maternity system where improvements can be made that provide for better outcomes for babies,” says co-author Ellie Wernham, a former midwife and Master of Public Health graduate from Otago.
The study identified an unexplained excess of adverse events in midwife-led compared to medically-led deliveries in New Zealand, where midwives practise autonomously.
The team of researchers, including Ms Wernham and senior author Professor Diana Sarfati, an epidemiologist and Co-Head of Department of the Public Health Department at Otago, examined major adverse perinatal outcomes between babies whose mothers had midwife-led and those who had medical-led care during pregnancy in a population-based retrospective study of more than 240,000 births over a five year period between 2008 and 2012 in New Zealand.
“We studied data for babies where there was no major fetal, neonatal, chromosomal or metabolic abnormality identified and the mother was first registered with a midwife, obstetrician or general practitioner,” says Ms Wernham.
They found that mothers with medical-led care compared with midwife-led care had lower odds of some adverse outcomes for infants. These included 55% lower odds of oxygen deprivation during the delivery, 39% lower odds of neonatal encephalopathy, a condition that can result in brain injury, and 48% odds of a low Apgar score. The Apgar score is a measure of infant well-being immediately after delivery, and a low score indicates an un-well baby.
“These findings demonstrate a need for further research that investigates the reasons for the apparent excess of adverse outcomes in midwife-led care,” says co-author Professor Diana Sarfati,
She says that this kind of review of the safety of midwife-led maternity system is essential for New Zealand to have the best system and outcomes it possibly can, and is also of relevance to other countries that might also be considering adopting this model of care.
“This research has been done very carefully and has been extensively reviewed in New Zealand and internationally before it was submitted,” she says.
The research team believes the Government and health agencies will need to consider and address the results of the study for the benefit of maternity care in New Zealand and is gratified to see that the Ministry of Health is taking these results seriously.
“Our primary aim with this work is to improve the quality of the maternity system, which already works well. The Ministry is taking this research very seriously and has already set up an evaluation to address some of the issues raised in the paper,” says Professor Sarfati.
The study has the unequivocal backing from Otago’s Pro-Vice-Chancellor of Health Sciences, Professor Peter Crampton, who says that the Ministry of Health, midwifery groups and the medical profession now need to work together for improvements to what must remain a midwife-led maternity system.
Professor Crampton says the sole purpose of this research is to improve the quality of New Zealand’s maternity system.
“All parts of the health system are (or at least should be) under frequent scrutiny with the aim of driving the quality agenda. This research contributes to our understanding of the performance of our system.
“We (the researchers and me personally) support
the current model of midwife-led maternity care. This is
important research. It tells us that there is scope for
improvement in the system. Together, we – researchers, the
Ministry of Health, the midwifery profession, the medical
profession – have a joint responsibility to work together
to address the issues identified in the research,” he
He adds that: “The results cannot be explained away for methodological reasons. The key messages remain essentially unchanged however we slice and dice the data.”
Co-author and epidemiologist Professor Di Sarfati’s critique of the Plos One Dutch midwives’ commentary in the same journal:
In response to this accompanying commentary on the Otago study, Professor Sarfati, co-author and Public Health researcher from University of Otago Wellington, says:
This is a high quality study. Several of the assertions in the “Commentary” (editorial) paper are wrong (eg that a high quality study would include a whole lot of other outcomes and that observational epidemiology can’t inform causality).
1. Our study aimed to investigate adverse fetal and neonatal outcomes in the NZ maternity system, not to investigate the relative benefits and harms of midwife led care. We included all relevant outcomes for which data were available, and defined these prior to doing any analysis. There was no selective reporting whatsoever, and the idea that this information is ‘not helpful’ is “extraordinary”.
2. In relation to how we should have assigned midwife vs medical led care, we were interested in model of care. Good midwife-led care may well involve the midwife taking advice from a doctor if complications arise, for example. That care is still midwife-led. We are very clear about that in the paper. So the suggestion that we could have used a different measure of midwife-led that was only from the time of labour is not consistent with this idea and more importantly, does not, in any way, explain the results we found. We found that only 1.9% of women changed LMC from registration to onset of labour, and removing these women made no difference to the results.
3. The idea that you cannot draw causal relationships from observational studies is wrong, and a common misconception among non-epidemiologists. If that was the case we would still not be able to say that smoking causes lung cancer (for example).
4. We assessed the likelihood that distance to hospital could explain these results (or any other unmeasured confounding factor) and found that it was exceptionally unlikely.
5. We completely agree that the different elements of midwife led care (and maternity care in general) should be unpacked to assess what is working well and what is working less well. We make that point very clearly in the final paragraph of the paper, and that is the obvious next step. Routine data does not allow us to do that.
6. Our study meets all criteria for a high quality study (and we were required to provide a check list to the journal for their review to ascertain that is the case). Finally, of note, this paper was extensively reviewed nationally and internationally by peer reviewers with a range of technical expertise (epidemiology, obstetrics, midwifery), and no methodological flaws that could explain the results were identified. The epidemiological reviewers (including John Lynch, Professor Charlotte Paul, Barry Borman and Trish Priest) in particular, were happy with the paper from a methodological perspective.