An Odd Case Shows the Need to Register a Protocol
As Alice discovered in Wonderland, it helps to have a plan. For clinical research it’s especially important to know where one is going with research. This is precisely why medical journals require that clinical researchers register a protocol for their studies before they start. If a study lacks that, a reputable journal will not publish it.
Unfolding right now is a case that may illustrate just how important this point is.
Impressive Results, But No Registered Protocol
A week ago, we wrote about a study of targeting anti-obesity medicines based on clinical phenotypes. The results of this study seemed impressive. Patients receiving targeted therapy achieved 15.9 percent weight loss, compared to 9.0 in the control group. The study abstract said that patient assignment to these two groups was random.
But then questions arose. Assuming researchers had registered the trial protocol required, researcher Kevin Hall looked it up on ClinicalTrials.gov. He found a protocol that appeared to be the right one. It had the same investigator and the same purpose. But he also found discrepancies:
“These are encouraging results about personalized phenotype-guided obesity treatment, but I have some questions and concerns about discrepancies between what was reported in the manuscript and what was in the study registration…
“The more I look, the more questions I have.”
The first author on this study replied that the trial was not registered in the NCT registry.
This leaves us with even more questions. Did investigators register this study somewhere else? If not, why did the journal publish it? Obesity, the journal that published this study, has a clear policy: “All clinical trials, regardless of when they were completed, must be pre-registered.”
Suddenly, the findings of a study that looked impressive at first glance now seem questionable.
Avoiding Issues
In a recent paper, Brian Nosek and colleagues explain why pre-registering a study protocol is so important:
“Preregistration clarifies the distinction between planned and unplanned research by reducing unnoticed flexibility. This improves credibility of findings and calibration of uncertainty.”
The bottom line here is really quite simple. In clinical research – or really any scientific research – it’s not OK to figure out what you’re doing after the fact.
Click here for the paper by Nosek et al. For more on the importance of registering clinical trial protocols, click here.
Planners Happy Day, artwork by Wyndham Lewis / WikiArt
Subscribe by email to follow the accumulating evidence and observations that shape our view of health, obesity, and policy.
April 7, 2021
