I can still recall that crushing moment decades ago, as if it were today. I had just spent my budget and time running an experiment in the lab that yielded no statistical results, to only find out days later that a professor in the lab down that hall had the same exact result decades earlier. Why don't I have a place to look up these failed experiments, so we as scientists don't waste time and money? It was disheartening to realize that studies with "no statistical significance" are considered less important than those "with statistical significance", by scientific publications when the foundation of the scientific process involves valuing and using the information that can be found in both sets of results.
FDA reviewers have been wrestling with this very same issue for decades. They balance protecting applicants IP, and not wasting new applicant's time and resources with a study design they know will yield no results. We can see this manifest in FDA reviewers making data collection "recommendations" to new applicants. These recommendations can seem vauge and frustrating as the FDA's role is not to design the trials but to approve or reject them. Because they must protect all applicants IP they can not disclose the reason for their recommendations despite potentially having a wealth of knowledge on the issue.
In 2000 momentum to change this, starting with the FDA requiring clinical trial design to be registered in the NIH ClinicalTrial.gov database prior to any study commencing. This was only a partial step though, while studies are registered, not all studies publicize their results. In recent years, there has been tremendous critisim of the scientific methodology applied to clinical trial studies. The lack of publications yielding no results led to an online debate online debate erupting to ask if science broken , because of data being "cherry picked" to yield positive results.
Proponents of transparency have been calling for full disclosure of clinical trials. Yet, over the years there has not been enough momentum to publish all studies including those with no results or negative results.
It's now looking like this may actually change. A new study of 55 large clinical trials testing heart-disease treatments has shown that registered clinical trials make positive findings vanish. This is disturbing for healthcare outcomes and scientific integrity. The study in Nature found that 57% of the trials published prior to the requirement to register clinical trial design, reported positive effects from treatment while only 8% of studies published after after the FDA's requirement was instituted in 2000 had positive effects.
The simple act of transparency requiring the registration of clinical trial has drawn the non-positive results out of the shadows. It also shook up the existing paradigms of medical device design, placing clinical trial design at the absolute forefront of success.
Medical device companies, innovators, and consultants need to simultaneously strategize their clinical trial design and new device validation designs, or risk being left in the dust by competitors.
In my previous blog post Don't believe those that tell you tracking down FDA regulatory classification for your product requires time and perseverance: both PMAs and 510ks have clinical trial data and ClearRoadmap can help.
ClearRoadmap™ enables you to quickly and effectively find clinical trial information for biotech diagnostic, medical device and mobile medical devices. It is the only database with Next Generation search capability, allowing users to search in-context: Indication of Use, Study Outcome, Recall as well as 17 other different categories. We a global platform, which lets you find devices that have clinical trial data and compare them side-by-side, when and where you need it.
-- Vizma Carver, Founder and CEO, Carver Global Health Group LLC and ClearRoadmap™