Clinical trials are the gold standard for understanding whether medical treatments work. They form the basis of whether drugs are approved by the FDA, and they’re used to demonstrate a drug’s effectiveness and to discover safety issues a drug may have. In these trials, patients are randomized to one treatment or another, and health outcomes are measured.
[time-brightcove not-tgx=”true”]
Trials are conducted on only a small swath of the population that may ultimately receive a treatment, however. The hope is that whatever is observed among patients enrolled in a trial will ultimately generalize to the broader population that receives treatment.
While ideally the subjects of a clinical trial should reflect the diversity of the population that’s likely to use the drug, in practice this doesn’t always happen. Trials often under-enroll women, the elderly, and minorities.
The under representation of these groups can have important consequences.
For example, in 1992, the FDA approved Ambien (zolpidem), a drug used to treat insomnia. Because women tend to metabolize the drug more slowly than men, they require lower doses in order to avoid excessive drowsiness. However, because women were underrepresented in the early clinical trials, this fact was overlooked, and it wasn’t until 2013 that the FDA officially required a lower dose for women. The consequences of this oversight were more than 700 reports to the FDA of traffic accidents potentially attributed to this dosing.
Examples like this highlight why it’s important for clinical trials to be representative. When it comes to women, one way to do this might be to have more women lead trials.
In a new study, we explored this idea by examining whether clinical trials led by women are more likely to enroll women as participants.
We assembled data on more than 10,000 clinical trials conducted over 15 years, linking data on the gender of the investigators who led those trials to data on the gender composition of patients enrolled in those trials.
Because women investigators may be more likely to study diseases that impact women (women may be more likely to study breast cancer, for example)—and similarly, men may be more likely to study diseases that impact men, like prostate cancer—we made sure to explicitly account for the disease being studied in a trial in order to address this potential source of bias.
We found that trials led by women investigators were more likely to recruit women as trial participants. On average, 54% of trial participants were women in trials where a woman was the lead investigator, compared to 47% for trials where the lead investigator was a man.
We also examined why trials led by women might enroll more women as trial subjects.
First, we found that trials led by women investigators were more likely to have women as staff. Research staff in clinical trials are often the front line, playing an important role in interacting with and enrolling patients. Having more women on staff could increase enrollment of women, especially if women trial participants are more comfortable interacting with other women.
Second, due to greater regulatory requirements and perceived legal risk, many clinical trials exclude pregnant women even in the absence of a specific reason to do so (e.g., some drugs are harmful to the fetus, and so excluding pregnant women would be normal in trials of these drugs). We found that women investigators were less likely to exclude pregnant women from clinical trials.
So, what do these results mean?
When women are underrepresented in trials, the overall findings of those scientific studies may not generalize well to them. The implication of this inequity in clinical trial representation is that treatments that are provided to women in the real world may lack the same level of evidence as for men, posing issues of treatment safety and efficacy.
Our study suggests that one way to increase the enrollment of women into clinical trials, and to ensure that equal standards of safety and efficacy apply to men and women, is to have more women scientists lead those trials. Organizations like the National Institutes of Health, universities, and companies have launched promising initiatives, but the leadership of clinical trials by women still lags behind men.
Increasing the proportion of clinical trial staff that are women may also help. Our study hints at the possibility that one reason trials led by women investigators enroll more women is because women investigators may employ more front line staff who are women. Women staff may try harder to recruit women into clinical trials or, alternatively, women may be more likely to enroll if the “face” of the trial to them is a woman. Training these staff may be an efficient way to increase the enrollment of women into trials.
The importance of recruiting diverse and representative clinical trial populations is driven by the intuitive fact that the effects of medical treatments can and do differ across groups of people. We should try to ensure that clinical trials are representative of the populations that treatments are intended for, and one way to do this may be to ensure that clinical trial teams are representative as well.