Patterns of seizure frequency reduction in clinical trial participants with lower baseline seizure frequency

The association of the magnitude of change in seizure frequency with baseline seizure frequency was inconsistent across trials. In eight of 11 trials, these patterns did not reduce the magnitude of difference between active treatment and placebo and thereby may not reduce statistical power. In only one trial did elevated placebo response reduce the difference between active treatment and placebo. In two trials, active treatment appeared more efficacious in lower seizure frequencies.