Cmax as covariate in mixed model?


New Member
I am involved in a study looking at drug effects on HR. We have one treatment and one placebo group, and are testing patients once/week for 8 weeks. One of the measures we are interested in is Cmax, and whether it affects efficacy (via HR), and if so, at what approximate range. The idea would be to determine at what Cmax range can we expect the greatest efficacy. We also have Cmax measurements once/week. Would it make sense to hold Cmax as a covariate in a mixed model, with or without transformation? What other options would you recommend? Thank you in advance!