What is the outcome we seek? A commentary on Keith et al. (2002)

The article by J. R. Keith et al. (2002) raises several issues that are frequently pertinent when neuropsychological tools are used as outcomes measures. These issues include whether the main interest is the mean change on a neuropsychological measure, the number of people who show a significant change on that measure, or both. Also at issue is whether the outcome of interest is either cognitive change or cognitive impairment. A corollary is determining when the failure to produce a practice effect is of clinical significance. Finally, it may be time to move beyond asking whether coronary artery bypass grafting (CABG) influences neuropsychological outcome. Rather, the pertinent question may now be, under what conditions and for which people does CABG have the least neuropsychological impact?