The interaction paper reported descriptive concentration means and modeled geometric mean ratios. Percentages calculated from those summaries need not be identical.
Two calculations appeared in the report
For ketoconazole, descriptive mean THC peak concentration increased from 2.65 to 3.36 ng/mL, approximately 27%. The interaction analysis reported a geometric mean ratio of 1.252 for THC peak concentration, using paired data from eleven participants.
The formal model analyzed log-transformed exposure measures and included treatment, period and sequence terms. Its ratios and 90% confidence intervals are therefore not simply divisions of every arithmetic mean printed in the descriptive table.
A small discrepancy is not automatically an error
Arithmetic and geometric means weight a variable distribution differently. Analysis samples can also differ when values needed for a particular parameter are unavailable. A reader should identify the estimand and denominator before treating different percentages as contradictory estimates.
Use the interval with the estimate
The modeled THC peak ratio had a 90% interval extending from 1.043 to 1.503. That conveys uncertainty around the group interaction estimate; it is not the range of changes guaranteed for each person. The study’s pharmacokinetic comparison does not establish a proportional change in cognitive impairment or authorize extrapolation to every cannabis formulation.
Related reading: interpreting proposed CBD mechanisms.
Research cited
Stott and colleagues (2013). A Phase I, open-label, randomized, crossover study in three parallel groups to evaluate the effect of Rifampicin, Ketoconazole, and Omeprazole on the pharmacokinetics of THC/CBD oromucosal spray in healthy volunteers. DOI: 10.1186/2193-1801-2-236.
Leave a Reply