Well-designed clinical trials provide essential evidence that help define clinical practice. Trials assessing clinical interventions give guidance on whom to treat, how to treat, and clarify the effects of an intervention on mortality and morbidity. Despite the importance of clinical trials, several studies have shown that there is frequent underreporting of results from clinical trials.1–3 Because clinical trial results must be reported to advance clinical practice and health research, the National Institutes of Health has an important obligation to ensure that these investments ($3 billion annually) are producing results and that those results are shared widely and applied to clinical practice.
The National Institutes of Health recently proposed to expand the interpretation of the clinical trial definition to include clinical research activities that were previously exempt. This revised interpretation would now classify most clinical research activities as clinical trials, and this has the potential to disrupt the funding, conduct, reporting, and interpretation of clinical trials along with all other clinical research. This rather large swing of the regulatory pendulum could have far-reaching effects on patients, providers, investigators, journal editors, research funding agencies, the biomedical research industry sector, and others. The language of the proposed expanded definition reads:
“A research study in which one or more human subjects are prospectively assigned to one or more interventions (which may include placebo or other control) to evaluate the effects of those interventions on health-related biomedical or behavioral outcomes.”
REDEFINING CLINICAL RESEARCH AND CLINICAL TRIALS
Breaking this definition down, there are four main components required to meet the expanded National Institutes of Health definition of a clinical trial:
1. The study must involve one or more human subjects.
2. Participants must be prospectively assigned to an intervention.
3. The study must evaluate the effects of an intervention on the participants.
4. The effect under evaluation must be a health-related biomedical or behavioral outcome.
If the study in question meets all of these criteria, then according to the NIH, it is a clinical trial. If a study does not meet all of these criteria, then it is not a clinical trial.
There are several ways that this new definition encroaches on what was previously defined as clinical research, but not a clinical trial. Much of this expanded definition hinges on the interpretation of the term intervention, which is further defined by the National Institutes of Health as
“…a manipulation of the subject or subject's environment for the purpose of modifying one or more health-related biomedical or behavioral processes and/or endpoints. Examples include drugs/small molecules/compounds, biologics, devices, procedures (e.g., surgical techniques), delivery systems (e.g., telemedicine, face-to-face interviews), strategies to change health-related behavior (e.g., diet, cognitive therapy, exercise, development of new habits), treatment strategies, prevention strategies, and diagnostic strategies.”
This inclusive definition of intervention will rightly capture the many clinical trials that are mechanistic or exploratory that fall outside the traditional realm of efficacy trials. This will help capture these less-traditional clinical trials and ensure that they receive the same measure of oversight and transparency needed to fully vet new therapies, diagnostic procedures, and other potential advances. However, it could also include virtually any other investigator manipulation of the environment or interaction with a research participant, for example, changing light levels, presenting psychophysical stimuli, comparing interpretations of images, and so on.
By further categorizing any study of “one or more human subjects” as a clinical trial, the definition also expands to smaller groups and even single individuals. If qualified, these small studies would, in theory, be subjected to the same regulatory scrutiny as randomized clinical trials involving hundreds of participants (i.e., data safety monitoring boards, protocol registration, etc.). This is not likely the intention and would be a burden for all involved.
STEP RIGHT UP!
The National Institutes of Health have prepared a collection of hypothetical scenarios to help communicate how to interpret these new guidelines, and these scenarios are available online at https://grants.nih.gov/policy/clinical-trials/case-studies.htm. I have incorporated these scenarios into a 10-question quiz. If you want to test your skill at determining which studies qualify for the revised definition of a clinical trial, you can use this anonymous link to take the quiz: https://uab.co1.qualtrics.com/jfe/form/SV_9uxGTRhgmRZ5QlD.
As with any new regulatory guideline, implementation will determine its success or failure. Given the very broad new language of this definition, virtually any clinical research study could qualify as a clinical trial. Investigators have been vocal in their opposition to these new rules,4 fearing that an overzealous interpretation will result in a nightmare of new and burdensome requirements that do not apply and are inappropriate for most clinical research. This new regulatory policy has the potential to create mass confusion and administrative paralysis for investigators and program officials alike during the transition to a new interpretation of clinical research. Hopefully, reason will prevail, and projects that are designed to provide unbiased evidence for the purpose of guiding the evolution of clinical practice will be correctly classified as clinical trials, and those that do not serve this purpose will escape from unnecessary burden and confusion.
THE IMPACT ON JOURNALS AND PUBLISHING CLINICAL TRIAL RESULTS
Investigators and the agencies that fund clinical trials have an obligation to ensure that the knowledge gained from those who bear the risk by participating in clinical trials is shared for the benefit of others. Fulfilling this obligation by publishing study results maintains public trust in the institutions that conduct health research. The existence of a negative publication bias is well accepted; that is, for various reasons, studies that do not show a positive treatment effect are less likely to be published. For example, authors may submit negative study results for publication less frequently, or negative results may be less favorably received during peer review, and so on. Conflicts of interest are also a barrier to publishing negative results from industry studies where there is no incentive to aid competitors. Likewise, investigators seeking career advancement have a similar bias away from publishing negative results and toward publishing favorable results. The fact that negative-effect studies go unpublished dishonors the clinical research participants, distorts the truth, and wastes resources on misguided and unnecessary duplicate research. Journal editors have a responsibility to see that results from well-designed, well-executed studies are published. Nevertheless, journals cannot publish what is never submitted.
It is difficult to see how this revised clinical trial definition will hit the intended mark—a better publication record for clinical trial results. A requirement to register clinical trials prior to subject enrollment was initiated by the International Committee of Medical Journal Editors in 2005, expanded in 2008, and is now the standard of practice adopted by most top-tier journals. This practice and the CONSORT statement on publishing clinical trial results have elevated the level of transparency and thoroughness with which clinical trials are reported. These contributions are evidence that editors can and have improved the reporting of clinical trial results. However, changing the standards for what is called a clinical trial will cause confusion about how to equate very different studies when both are labeled as clinical trials. This will also devalue the current benchmark reserved for high-quality clinical trials. It is unclear how these revised definitions will improve author compliance with requirements to publish results from clinical trials. For journal editors, these new expanded definitions will likely breed confusion among editorial boards, reviewers, and authors as the community debates how to apply these new standards.
THE IMPACT ON PATIENTS AND THE PUBLIC
There is another significant risk to consider that could come from an expansion of the clinical trial definition. Aside from the potential administrative disruption that this new expanded definition could cause, this change could also erode public trust in scientific research. By expanding the definition of a clinical trial, its value will be diminished for patients and participants. Moreover, if the investigators and staff conducting clinical trials find it difficult to articulate the value and importance of clinical trials, participants will likely share in this confusion. Labeling most clinical research as a clinical trial will erode the value of clinical trials, potentially reducing the likelihood that patients will choose to participate in them.
If the target to address is a lack of published clinical trial results, then one should consider whether there is an existing remedy, or perhaps an even better one. An existing solution that requires no change is to deny future funding to investigators who do not publish clinical trial results, and this idea was stated recently: “The National Institutes of Health will withhold clinical trial funding if the agency is unable to verify adequate registration and results reporting….”2 Conversely, providing a portion of funding explicitly for publishing results or a publication bonus, that is, specific incentives for publication of results, would likely change behavior quickly.
Clinical trials emerged in the early 20th century as a way to fairly assess the impact of medical interventions.5 Fair comparisons of clinical interventions are the crux of modern clinical trials and include specific design elements that minimize potential bias such as random allocation of interventions for treatment and control groups; masking participants, investigators, and analysts from group assignments; and more. These core values and design constructs must remain the focus of the proposed clinical trial definition, or we risk devaluation of clinical trials, erosion of public trust in clinical research, and confusion among investigators, participants, editors, and others. For now, Optometry and Vision Science will continue to apply a traditional definition of clinical trials.
Michael D. Twa
Editor in Chief