The randomized concentration-controlled trial: Mathematical definitions, a dose-adjusting algorithm, and sample size efficiency

A randomized concentration-controlled trial (RCCT) (Peck 1990) is distinguished from a randomized dose-controlled trial (RDCT) by controlling the plasma concentration and allowing the dose to vary, as opposed to controlling the dose and allowing the concentration to vary. Monte Carlo studies have suggested that the RCCT is more efficient (in terms of sample size) than the more traditional dose-controlled trials. This paper presents a general mathematical defininition and investigates the properties of an RCCT. A Bayesian adjustment algorithm used in an RCCT is presented and discussed. Expressions for the sample-size efficiency of the RCCT relative to the RDCT is found for both Gaussian and Bernoulli efficacy variables. The RCCT is found to be at least as efficient in terms of sample size. The RCCT can be considerably more efficient than the RDCT if the interpatient pharmacokinetic variation is large; if the interpatient pharmacokinetic variation is small, the RCCT is not much more efficient than the RDCT.