Some pregnant women can begin having contractions weeks before their due date — but whether they will soon give birth to a premature baby or are experiencing false labour can be difficult for doctors to ascertain.
So researchers in Toronto, along with collaborators in Australia and Singapore, have been working to develop a test to help tease out which women having pre-term labour will actually soon give birth and those who won't and can go home to continue their pregnancy.
Currently, the decision that a woman is in premature labour and soon to deliver is based on the contractions and other physical symptoms, such as back pain, pelvic pressure and a shortened cervix.
A test that measures the amount of a protein in vaginal fluid, called fetal fibronectin, or fFN, is also used when possible.
"The problem with that information is it's not very good at identifying women who are actually in pre-term labour," said
principal researcher Dr. Stephen Lye, noting that the fFN test can only predict which women are going to give birth imminently with about 30 to 40 per cent accuracy.
Most of the women won't deliver in the next 48 hours, but will settle down and probably the majority will go full-term of
near-term, said Lye, associate director of the Lunenfeld-Tanenbaum Research Institute at Mount Sinai Hospital in Toronto.
"The problem is that once the diagnosis of pre-term labour is made, there's a whole set of interventions that start taking
The woman would be admitted to hospital and probably given a drug to try to stop the contractions or at least to slow down the birth process in order to give the fetus more developing time in the womb.
Likely the mother-to-be will also be given steroids, with the aim of helping the baby's not-fully-developed lungs to mature as much as possible. (One of the biggest problems for preemies is oxygenation due to immature lungs, he said.)
"First of all, you don't want the hospital antenatal wards full of women that are not really in true labour because they use up hospital resources and there's lots of expensive tests being done," said Lye. "And the mothers are obviously stressed and away from their families. And they will be given drugs to try to stop the labour, which have some side-effects, and these steroids, which may or may not be good for the baby.
"We need to have a better test that has a higher ability to say: yes, this woman is going to deliver and we should do all those things, or no, this woman isn't going to deliver and we can watch her for a short while but then allow her to go home.
"That's what we've been looking at here," he said, noting that such a test could save tens of millions of dollars a year spent on treating women in hospital who are not about to give birth pre-term but are experiencing false labour.
In a paper published in the journal PLOS ONE on Wednesday, the researchers describe how they took blood samples from 154 women in Perth, Australia, who had gone into what appeared to be premature labour.
Using DNA sequencing, the researchers isolated a signature of nine genes that were able to identify which of the women would give birth, with about a 70 per cent success rate.
In 62 of the women who were able to have fFN testing — to do that test, the vaginal fluid cannot contain substances such as blood — the results turned out to have a predictive value in the 90 percentile range as to which women were not about to deliver.
Lye said about nine per cent of pregnant women experience pre-term labour, but most give birth at about 35 or 36 weeks
gestation; about two or three per cent will have babies born very prematurely at 25 to 27 weeks, and it is those infants who end up in neonatal intensive care units and are at risk for long-term health effects.
The University of Toronto professor of obstetrics and gynecology said he has been in contact with companies about developing a test, but more research needs to be done to replicate the results in more women.
"It's important not to have the public think there's a test tomorrow," Lye said. "It's promising, it's exciting and it's
evolving …[but] this has to be validated."