Why Randomized Controlled Trials (RCTs)?
Pt. 2: Hormone Replacement Therapy
This is Part 2 in a series called “Why RCTs?” which explores experiences with and without the benefit of randomized study across disciplines. In case you weren’t shocked enough by the robot babies account from last week, this blog post explores another catastrophic outcome, this time in the medical field. This time, the RCT produced results drastically different from those in prior observational studies, causing the medical community to reconsider long-held beliefs about the benefits of a treatment for cancer patients.
For many years hormone replacement therapy (HRT) was a common treatment for women with menopausal symptoms. Many of the same symptoms also happen to be common among premenopausal women who survive breast cancer. Based on large-scale observational studies, HRT was considered safe and to have many positive benefits such as preventing heart disease and osteoporosis. Sounds great, right?
Unfortunately, not always. The results of an RCT run by the Women’s Health Initiative (WHI) suggested the possibility that HRT may actually increase the risk of breast cancer and myocardial infarction. In fact the trial was terminated for safety reasons when women in the HRT treatment group started to experience another breast cancer event at higher rates than those in the control group.
How did these studies differ? Why weren’t the observational studies getting the full picture? The observational studies were plagued by many nonrandom differences between groups of women who had chosen to receive HRT and those who didn’t that may have skewed the results of the studies. For example, women who had chosen to take HRT had lower heart disease, were more likely to lead healthier lifestyles and be consistent with HRT treatment over time. For the RCT, the absence of menopausal symptoms was a pre-requisite for entering the study, unlike in the observational studies, where the women who had chosen HRT tended to be younger and lacking in estrogen.
RCTs can fill in data one might not even know is missing, and reveal critical issues with previous research methods. The WHI’s RCT “changed abruptly” the use of hormone replacement therapy and is now no longer as recommended for women as it used to be. We need RCTs to understand what interventions work and what do not, especially when the stakes are this high. It is understandable to be concerned at first with randomizing in the law and tinkering with the type of representation litigants receive, especially when this representation may have a profound impact on clients’ lives. But unless we conduct rigorous evaluation, we cannot know what legal services are helping people with their serious, life-changing issues. As long as research is conducted ethically, sensitive and important issues are where we need rigorous evaluation the most.