MacPherson and colleagues' article (1) reports outcomes that differ from those initially registered (2). Three primary outcomes (1 score at 3 time points) were prespecified. Although the article reports data for all 3 time points, it incorrectly states that the primary end point was at 12 months when the registry entry gives no such ranking. In addition, 17 secondary outcomes (9 outcomes to be measured at between 1 and 3 time points each) were prespecified; of these, 6 are reported in the paper, whereas 11 are not reported anywhere in the publication. Furthermore, the article reports various nonprespecified analytic approaches, such as the use of regression models, and declares only some of these as novel.