Adults who remained symptomatic during previous amphetamine treatment improved after lisdexamfetamine in a trial subgroup. With only two placebo participants in that subgroup, the analysis cannot establish a reliable comparative switching benefit.
Babcock and colleagues reanalyzed a four-week randomized placebo-controlled adult ADHD trial. Of 414 efficacy participants, 41 had received amphetamine treatment at screening. Thirty-six of those remained symptomatic according to the specified rating threshold.
The symptomatic subgroup contained 34 lisdexamfetamine participants and two placebo participants. It was not a dedicated randomized trial designed to compare switching strategies in previous amphetamine nonresponders.
The study reported symptom changes from screening, when prior treatment was being assessed, and from baseline after medication washout. These changes answer different questions. An improvement from the post-washout baseline cannot automatically establish superiority over a person’s prior regimen.
Descriptive changes suggested benefit under lisdexamfetamine, but a placebo estimate based on two people is highly unstable. Similar-looking changes in this subgroup and the whole trial are also not evidence of equivalent treatment effects.
The authors noted that information about symptom severity before earlier amphetamine treatment was unavailable. Remaining above a symptom threshold does not show that the previous medicine provided no benefit, was optimally timed or was otherwise equivalent to the new regimen.
The post hoc design, selected population and short duration limit conclusions. This industry-funded report supports prospective investigation of switching, not a universal recommendation or a claim about drug concentrations and impairment.
See the importance of population and duration in active comparisons.
Efficacy of lisdexamfetamine dimesylate in adults with attention-deficit/hyperactivity disorder previously treated with amphetamines: analyses from a randomized, double-blind, multicenter, placebo-controlled titration study. doi:10.1186/2050-6511-13-18.