The Ministry of Health is investigating after a study revealed higher rates of damage to babies born under midwife care than those whose mothers were looked after by doctors.
The study has been criticised by the College of Midwives, who say the findings simply show a rich-poor divide.
The five-year University of Otago study of more than 244,000 babies born in New Zealand from 2008 to 2012 found mothers with medical-led care, such as by obstetricians or GPs, had babies with lower odds of adverse health outcomes at birth.
These included 55 per cent lower odds of oxygen deprivation during delivery, 39 per cent lower odds of neonatal encephalopathy, which can cause brain injury, and 48 per cent lower odds of poor health, as measured by a low Apgar score.
An Apgar score is a measure of infant well-being immediately after birth and a low score indicates an unwell baby.
Ministry of Health chief adviser child and youth health Dr Pat Tuohy and principal maternity adviser Bronwen Pelvin said the study's findings were unexpected, as international evidence shows a midwifery-led maternity system leads to similar outcomes for women and their babies.
"The Ministry of Health has referred the study to the National Maternity Monitoring Group for advice on whether further research needs to be undertaken to help us better understand whether the findings reflect something about the way the study was done, differences in the maternity care provided by midwives and doctors, and whether there are things we can change to get better outcomes for women and their babies," they said.
Issues that needed further exploration included looking at what other care was provided to women in the midwifery-led group, looking at differences between the types of women being compared in the study, and investigating how quickly specialist services respond to complications in pregnancy, labour and birth.
On Friday, the Ministry of Health sent out another statement about the study to reassure New Zealanders the maternity system was safe.
"The study itself backs up what we already know - that New Zealand has low rates of adverse outcomes for both mothers and babies, and which are similar to countries such as Australia and the United Kingdom," the ministry's chief medical officer Dr Andrew Simpson said.
"While our maternity care is good, there is room for improvement in any system. We are all working towards the same goal which is the best outcomes for mothers and babies."
A report on the study said it was not possible to definitively determine whether one model of care was associated with fewer infant deaths during birth.
Co-authored by a former midwife, the research is the first major safety review of New Zealand's autonomous midwife-led maternity system adopted in 1990.
Published in Plos Medicine this week, it found that adverse health outcomes were substantially lower for babies born to mothers registered with medical-led maternity carers than those who had midwives as their lead carer.
The babies studied had no major foetal, neonatal, chromosomal or metabolic abnormalities.
The report concluded the findings were a concern.
"These findings demonstrate a need for further research that investigates the reasons for the apparent excess of adverse outcomes in midwife-led care," co-author Professor Diana Sarfati said.
Co-author Ellie Wernham said as a former midwife she saw first-hand the benefits of a midwife-led model but the research showed aspects of New Zealand's maternity system could be improved.
Action to Improve Maternity (Aim) founder Jenn Hooper, whose 11-year-old daughter Charley was left severely brain damaged following a botched birth by midwives, welcomed the research and called for urgent investigation into the findings.
Hooper said the 700-plus families whose babies were damaged at birth that Aim has helped since it began eight years ago had "paid the price" for the midwife-led model.
"We need to get this sorted. I would like to think that further research would delve into the education, the training and the lack of internship [of midwives]. I wouldn't be surprised at all if that's a big part of the cause."
Newly graduated midwives do not require nursing training and can be self-employed in the community.
Previous health minister Tony Ryall sought advice on hospital internships but current minister Dr Jonathan Coleman said last year he would not re-look at the mandatory training.
New Zealand College of Midwives chief executive Karen Guilliland said women under the care of midwives were more likely to be rural and remote, Maori, Pasifika, younger, smoke, obese, sicker and book late with a carer.
"These are all well-known risk factors that cause the adverse outcomes described. What we are likely to be comparing is the difference between women who live in big cities (mainly Auckland) and who can pay thousands of dollars and afford to have a private obstetrician."
Guilliland said in Auckland's DHB about 25 per cent of maternity carers were obstetricians; something she called "unusual" as 93 per cent of carers in the whole country were midwives.
Tuohy said the "unexpected" differences could be the result of the way the study was done.
"We know the study looked at the LMC [leading maternity carer] at the time of registration for maternity care, but it didn't look at who else was involved in providing care, particularly at the birth."
Sarfati said she and Wernham had accounted for the differences Guilliland raised.
"In all cases those differences combined didn't account for the difference in outcomes.
"It didn't even come close."
She said there were also problems in concentrating on the person who delivered the baby, as suggested by Tuohy.
"For example if a problem occurs and it's missed and it's missed and it's missed, and then someone delivers the baby who is in a bad way, it wouldn't be fair to say that was the fault of the person who delivered the baby."
She said Tuohy's criticisms didn't explain the study's findings, which still showed an "excess in adverse outcomes" where midwives were the lead carer.
"The next step, having found there's an apparent higher rate of adverse outcomes in one group is to look at exactly what happened and why."
The New Zealand Nurses Organisation welcomed the news and said it could trigger a "wider investigation of the whole maternity system".
"Midwives are part of a much bigger service. We have a world class system and if we are concerned about outcomes then we need to turn our gaze to the wider health system, funding and access," NZNO chief executive Memo Musa said.
NZNO noted the research would now be out-dated and much has happened since the data used was received.
"The 2016 Cochrane 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," he said.
Mr Musa pointed out recent Ministry of Health referral guidelines setting out when a midwife must refer to a specialist have been put in place and the midwifery education programme has been updated since this research was done.