I would not classify most of my experiments as clinical trials, but now, under NIH rules that become effective in 2018, much of my work along with a significant portion of basic human behavioral research will be classified as clinical trials. This change is motivated by worthy goals but, in its current form, it has significant consequences. Let me tell you how we got here and what we need to do about it.
Many years ago when I was studying binocular vision, I stumbled on an interesting phenomenon in binocular rivalry. I will spare you the details, but I thought it could be significant so I started to discuss the observation with my more senior colleagues. Many of them told me that they had noticed the same thing years ago but had never quite gotten around to publishing the result. Indeed, one of them went over to his file cabinet and pulled out the file—a completely literal illustration of the “file drawer effect”. I published a couple of papers on the topic and that was that.
This is a somewhat atypical example of the file drawer effect. The effect is more traditionally defined as the selective publication of only positive (or statistically “significant” results), with other research results taking up residence in the proverbial file drawer, never to be seen again. One downside of the file drawer effect is that if I don’t tell you that my experiment failed, you may waste your time on the same experiment. Moreover, if there are, say, 20 failures of that experiment languishing in the files, the one published study with its p < 0.05 positive result is quite possibly just statistical noise…but it will be published, citable noise. There are strong reasons for us as a field to be concerned about the publication of such statistical noise.
On the other hand, wearing my journal editor hat, I am not actually that interested in publishing a badly designed, underpowered exploration of an uninteresting issue, simply in the name of transparency and openness.
I do not propose to resolve this issue here. I raise it as preamble to the controversy surrounding the NIH’s 2014 definition and regulations surrounding “clinical trials”. Specifically, that experiment I did years ago on binocular rivalry might be defined as a clinical trial next year and therein lies what could be a big problem. Our problem begins with a real problem faced by the NIH. It funds something like three billion dollars a year in research on humans and, apparently, quite a substantial fraction of the resulting work does not end up published in a timely manner or reported in any public database. Suppose you get a grant to do a clinical trial of a drug. You randomize half your patients into a group that gets the drug and half into the group that gets a placebo. It seems entirely reasonable to say that the tax paying public and the scientific community should learn about the results, even if they are negative or, perhaps, some unfortunate statistical mess.
Starting back, I think, in 2012, the NIH began a large effort to address the problem. They wrote a sweeping new definition of “clinical trial” (found here), and they announced that those trials were subject to a set of policies. There are various training and grant application requirements. Moreover, all trials needed to be registered in ClinicalTrials.gov and results needed be reported there. You can look into the details starting here. Lots of details needed to be worked out so the rules really kick in with the January, 2018, grant cycle.
Those of us doing basic research in experimental psychology, cognitive science, cognitive neuroscience, brain science and related fields heard about this push for more transparency in clinical trials but we did not pay much attention because we did not think we were doing clinical trials. We might have read some of the news pieces but we read them the way we might read about the ethical implications of CRISPR gene manipulation; interesting, but not what I do.
However, we were wrong. Under the 2014 definition, large swaths of basic research would be defined as “clinical trials” meaning that we could come under and this could pose a host of problems. To offer a sample:
- We will be required to submit future human subjects applications to a clinical trial-specific funding opportunity announcement (FOA). This has implications for review, though we have been told that NIH will work to keep the review process for basic science essentially unchanged..
- There are some current rules prevent training grants from being used to fund “clinical trials” though, again, we have been told that NIH will work to make F- , T-, and K-series grants a continuing option for human behavioral research.
- Many of our universities have substantial restrictions on “clinical trials”. Ironically, at a time when changes in the “Common Rule” looked likely to simplify IRB human subject regulations for some of our work, this change would complicate matters.
- Exploratory research with its little pilot studies, its dead ends, and its methodological fine-tuning is a square peg that just does not fit into the round hole of the NIH clinical trial process.
- Perhaps most fundamentally, labeling much of our work as “clinical trials” runs counter to the commonsense definition of the term. This risks confusing the public, including patients looking to register in “real” trials. It risks confusing the legislators who fund this work. My old binocular rivalry work might be held up to ridicule if described as a “clinical trial”. It was basic research that other researchers turned into fundamental research on conscious awareness that has a direct line to truly clinical work on the mental life of patients in so-called vegetative states. It was part of the great sweep of NIH-funded research, but not, in any real sense, a clinical trial.
I could go on, but what should we be doing now? We as a community have recognized for years that the basic behavioral research community faces many of the same issues around openness, transparency, and responsibility to the funding public and the scientific community. Indeed, many in our field have been working on projects such as registered reports in our journals, public posting of data, etc. We should be—and are—saying clearly to the NIH that we are eager to work with them on a national system of registration and results reporting that meets the interests of all stakeholders, including scientists engaged in basic discovery research in humans.
That said, it is late in the day. Until recently, we didn’t understand that the clinical trial policy applied to us but unless we take action, it will. We need to do a bit of advocacy because we want to make sure that we can continue to do our work with minimal impediments and we want to have a say in how the process will work over the long term. So, here is what you should do:
- Sign our open letter to NIH director, Francis Collins here.
- Comment on the Mike Lauer’s “Open Mike” blog here.
- Be sure that your university administration / grants office / etc. are aware of this issue as it applies to basic behavioral research.
- Keep your eyes and ears open for more news.
The Federation of Associations of Behavioral and Brain Sciences (FABBS) is acting as a clearinghouse and coordinator for efforts on this issue. I currently serve as FABBS president so please feel free to keep me posted on your actions on this front (firstname.lastname@example.org).