Please welcome a guest post by Kaitlin Bell Barnett, who writes about health and medicine for various print and online outlets. She’s at work on a book about medications and medical care in pregnancy. Her first child is due in July.
New FDA guidelines say that clinical trial participants should know not just the risks and benefits of the study they’re enrolling in, but how those risks and benefits compare to the typical care they’d get at the doctor’s office. These recommendations, which are expected to be finalized later this year, seem reasonable enough at first glance. But, they’re likely to further hamper research in a population that’s already badly underrepresented in clinical studies: pregnant women.
Researchers already avoid conducting studies on pregnant women because of concerns about risks to the fetus. In clinical practice, this lack of research often translates into women foregoing care, since doctors know very little about how certain drugs could affect the developing fetus. Better safe than sorry, they reason.
Pregnant women’s lack of inclusion in formal studies doesn’t mean they’re not being experimented on. They are – just not in a systematic way that would help clarify how helpful or harmful medications and other interventions actually are. The average expectant woman takes three to five prescription drugs during her pregnancy, not to mention many over-the-counter herbs and supplements.
“It’s really problematic that we don’t know even the basic pharmacokinetics of how certain medications work in pregnant women,” says Miriam Kuppermann, head of the clinical research program for the University of California, San Francisco’s maternal-fetal medicine program. “There’s good reason to believe that those will be different in pregnant women because of all the pregnancy hormones.”
When it comes to disclosing the risks and benefits of the typical “standard of care”, as the proposed FDA guidelines recommend, researchers face a catch-22: How can they inform pregnant participants – or participants who may become pregnant during the course of the trial – about risks and benefits that simply aren’t known?
The key lies in how they communicate risks, and how they present the “standard of care” for pregnant women.
Pregnancy and Depression
Take the case of depression, which affects 12 to 15 percent of pregnant women.
Some recent studies suggest that SSRI antidepressants may increase the risk of first-trimester miscarriage, preterm birth, minor complications after birth, and that one drug, Paxil, increases the risk of cardiac defects. Other studies, though, have found no increased risk. And a substantial body of evidence shows that mothers who are depressed while pregnant have a higher risk of preterm labor, low birthweight babies and other pregnancy complications that can hurt the baby, especially if the mothers are too depressed to eat properly, exercise, go to prenatal checkups, and otherwise adequately care for themselves.
Clearly, more research is needed on the most effective ways to treat prenatal depression. Informed consent materials and procedures for a study of a drug or therapy should explain the reasonably foreseeable risks and benefits of that intervention – both to mother and fetus – based on any previous data.
Because there is no single, accepted standard of care for prenatal depression, informing women about their alternatives if they were to seek treatment at the doctors should cover all the major options: the effectiveness and risks of antidepressants, different modes of therapy – and of untreated prenatal depression.
Too Much Information
It’s easy to see how all this talk of risks and benefits, most of them still ill-defined, could overwhelm prospective subjects.
“If you don’t know the risks and benefits, then telling pregnant women that they need to weigh them is just a way of making them feel stressed out,” says Georgetown bioethicist Rebecca Kukla, who has sat on IRBs, the governing bodies for universities and hospitals that handle informed consent procedures. “It’s literally an impossible burden to follow.”
Doctors, researchers and bioethicists have already voiced concerns that the requirements to disclose yet more risks will trip up research, not just involving pregnant subjects (a public comment period on the guidelines ended last month).That situation is likely to be even worse when it comes to pregnant women, since the quality of data for commonly-used medications that constitute the standard of care are is of such poor quality, and because the risks and benefits for both the mother and fetus must be weighed.
“If it were to have a general chilling effect I could definitely imagine that it would it would have an even greater chilling effect on pregnancy, because everything does,” says Margaret Little, who runs Georgetown’s Kennedy Institute of Ethics and has written extensively about pregnant women’s participation in clinical research. “There’s such an aversion to doing [this research] anyway that it adds one more stone to the big pile of rocks that are keeping people from doing this.”
Indeed, research shows more information is not always more helpful when it comes to informed consent, says Kuppermann, who studies methods of presenting risks and benefits to trial participants. Sometimes, simpler is better.
The number needed to treat and number needed to harm
To avoid drowning subjects in a sea of statistics, researchers should more widely employ two easy-to-understand measures of risk and benefit: the Number Needed to Treat (NNT) and the Number Needed to Harm (NNH). The NNT captures how many people would have to take a drug or receive the treatment in order for at least one person to benefit. Its twin, the NNH, calculates how many people must be treated in order for one additional person to be harmed
One study showed, for example, one additional miscarriage for every 26 women treated with antidepressants (the NNH), while another review of multiple studies showed 1 additional person saw improvement in depression symptoms for every 5 people receiving antidepressants, compared to one person in seven for placebo. By providing both statistics, researchers can more clearly communicate how the benefits of treatment stack up against the risks.
Including the NNT and NNH stats along with succinct summaries of research findings regarding the effects on both mother and fetus would help pregnant women evaluate the risks and benefits of enrolling in the trial – and of the treatments they would likely receive (or not receive) in ordinary medical care.
Then and only then can pregnant women truly know what they’re getting themselves into – whether they choose to enroll in the study or go to the doctor.