Randomized trial of deep brain stimulation for Parkinson disease: thirty-six-month outcomes; turning tables: should GPi become the preferred DBS target for Parkinson disease?

Weaver et al. and Tagliati mistakenly infer clinical equivalence between globus pallidus interna vs subthalamic nucleus deep brain stimulation based on failure to demonstrate statistically significant differences. A clinically meaningful-not statistically significant-difference in outcome should be decided a priori, after which the sample size necessary to have a reasonable probability of detecting the difference could be determined. (4) Fortunately, the study by Weaver et al. had sufficient sample size to demonstrate a 1-point difference in motor outcomes. However, such comparisons presume optimal management so as not to produce a "ceiling effect" that would obscure differences.