Task Exclusions Changed the CBD Imaging Analysis Sample

Two high-risk participants were excluded from imaging analysis after failing to perform the task correctly. The randomized enrollment count and the final imaging denominator were therefore different.

The study reported both counts

Thirty-three antipsychotic-naive high-risk participants were assigned to CBD or placebo. One from each treatment group was excluded for task performance, leaving fifteen CBD participants and sixteen placebo participants for imaging, alongside nineteen healthy comparators. Baseline tables and analytic results need to be read with their respective denominators.

Equal exclusion counts do not answer every bias question

Removing one person per group avoids a simple numerical imbalance in exclusions. It does not establish that the excluded participants were interchangeable or that they would have shown the same activation patterns as completers. The report does not supply their missing valid task scans, so those outcomes cannot be reconstructed by assumption.

Interpret the analyzed population

The findings concern participants able to complete the imaging task under the protocol. They do not represent every person at clinical high risk with equal certainty. A report should separate recruitment, treatment assignment, task completion and usable imaging. The absence of significant performance differences in the analyzed experiment also prevents the imaging result from being presented as proven memory enhancement.

Related analysis: brain activation and the null task-performance result.

Research cited

Bhattacharyya and colleagues (2018). Effect of Cannabidiol on Medial Temporal, Midbrain, and Striatal Dysfunction in People at Clinical High Risk of Psychosis. DOI: 10.1001/jamapsychiatry.2018.2309.