How NOT to Conduct a Randomized Clinical Trial

Dr. Levin will discuss best practices on planning and conducting a clinical trial with illustrative examples drawn from a real-life randomized controlled trial designed to examine the efficacy of graded exercise therapy (GET) and cognitive behavioral therapy (CBT) for chronic fatigue syndrome.
A PDF of the slides presented can be found here: tinyurl.com/y9pc2awh
Part of the "Biostatistics in Action: Tips for Clinical Researchers" lecture series that is sponsored by the Irving Institute for Clinical and Translational Research - Biostatistics, Epidemiology and Research Design resource, which is supported in part by an NIH Clinical and Translational Science Award (CTSA) through its Center for Advancing Translational Sciences (Grant No, UL1TR001873).
The speaker, Bruce Levin, PhD, is a Professor in the Department of Biostatistics at the Mailman School of Public Health.
Sponsored by: The Irving Institute for Clinical and Translational Research: irvinginstitute.columbia.edu/
In affiliation with: The Department of Biostatistics at the Mailman School of Public Health: www.mailman.columbia.edu/beco...
Links mentioned:
www.ClinicalTrials.gov
CONSORT: www.consort-statement.org/
Equator: www.equator-network.org/

[@sophiamcl]

As someone suffering with ME/CFS, I feel great sadness at the immesurable negative impact this debacle has had and still has on thousands upon thousands of patients with this diagnosis. It has caused both physical and mental suffering for patients forced to undergo GET and CBT and as a result have lost function and suffered a worsening of symptoms and irreparable damage, as well as being belittled, dismissed and met with ignorance and disbelief from healthcare professionals. All this for what? Pride? Despicable.

[@ceecee4173]

Good to see a Stats POV on this, this seems a terrible case of data dredging given the lack of any appropriate blinding. Originally stats were given the groupings (ABCD), as stated in the limited SAP, but data collected on treatment times etc would probably have unblinded these. Why wasn’t this data handled separately? Preliminary results mean the blind was broken early, and as ABCD groupings known to Stats then all subsequent patients unblinded. So much potential for bias

[@Candyapplebone]

I was thinking it might be fun to conduct a randomized controlled trial, this seems like a good resource for learning about how to do it