TORONTO – A Quebec study has found an association between certain anti-inflammatory medications taken early in pregnancy and miscarriage.
The study published Tuesday in the Canadian Medical Association Journal finds the risk of miscarriage is 2.4 times greater for women who are prescribed non-aspirin, non-steroidal anti-inflammatory drugs, known as NSAIDs.
The class of drugs is used to relieve pain and includes naproxen, ibuprofen, diclofenac and celecoxib.
But Dr. Gideon Koren, director of the Motherisk program at the Hospital for Sick Children in Toronto, said he has “major problems” with the study.
The researchers used data from the Quebec Pregnancy Registry, which has tracked pregnancies since 1997, as well as the Quebec insurance plan known as RAMQ, which provides drug coverage to people on social assistance and those who don’t have access to a private drug insurance plan – about 36 per cent of pregnant women in the province.
Koren said that going to databases for information works well when it involves prescribed drugs only, but in this study, one of the drugs involved – ibuprofen – is also available without a prescription.
“The ibuprofen is a huge part and we don’t know who bought it or not, and who took it, if the group that did not have miscarriages bought it more than the other group,” he said Tuesday in an interview.
“You cannot use a prescription database for a drug which is given not just for prescription. It just doesn’t jibe.”
He also pointed to a new study from Sweden that interviewed subjects to see what drugs they took compared to what was prescribed for them, and it was discovered that many women didn’t end up taking the medications.
In the Quebec study, researchers looked at 4,705 cases of miscarriage up to the 20th week of gestation, and found 7.5 per cent had been prescribed a non-aspirin NSAID. A control group of 47,050 women who did not miscarry included 2.6 per cent who were exposed to non-aspirin NSAIDs.
Co-author Anick Berard said the increased risk of spontaneous abortion was found regardless of which NSAID the women had been prescribed, and it went from a doubling of the risk up to a tripling of risk.
She said some women may be suffering from chronic conditions that require medication, and the majority of inadvertent exposures occur in women who don’t plan their pregnancies.
“Plan your pregnancy, discuss it with your health-care provider in order to fully understand the benefits and the risk of using a medication during pregnancy,” recommended Berard, who’s with the University of Montreal and is director of the research unit on medications and pregnancy at CHU Ste-Justine.
“Women have to understand that there’s no risk of zero.”
“I think that NSAIDs – because the majority of the use is for pain and migraine headache, for example – I think that women can very well go … at the beginning of their pregnancy without these drugs.”
She said there are alternatives to NSAIDs that women with headaches can take for pain during the first trimester.
Koren agreed that for a mild headache, acetaminophen would be preferred to ibuprofen for a pregnant woman.
But he noted that half of all pregnant women do not plan pregnancy, and many of the women who turn to the Motherisk counselling program took a drug before they knew they were pregnant.
Whether someone should take a particular drug might depend on other considerations, he added.
“If a woman has very severe rheumatoid arthritis and without taking naproxen or any of the others, she cannot function, she’s very different from a woman who has once in a while a headache,” he said.
“And then they’re both very different from a woman with migraine headaches who may not be able to be in class as a teacher if she’s not controlling (the pain). So the answer clearly cannot be based on this study. This study is too flawed.”
Koren said some previous studies have shown an association between NSAIDs and spontaneous abortion, and others haven’t.
He said he’s aware of a very large study that’s not yet published that does not show this association.
Berard said her analysis tried to take away the risk of miscarriage that would be attributable to the underlying disease for which a woman is taking medication – for instance, lupus or rheumatoid arthritis.
“We took good care in trying to tease out the drug effect from the condition effect. However, I have to say, I have to be really honest, it is an observational study, it’s not a randomized controlled trial,” she said.
“So we cannot be sure, 100 per cent sure, that we completely took away the effect of the underlying disease.
“That being said, our risk estimate is so strong, you know, it’s 2.4, so it is very big. And it’s very unlikely that even if we didn’t completely take away the effect of the underlying condition, it is unlikely that that risk would go from 2.4 to (nil). So there would still be an increased risk.”