Abstract
In the light of prevailing views about the difficulty of analysing data from cross-over trials in Alzheimer''s disease and the increasing emphasis being placed on the possible clinical relevance of statistically significant changes in test scores we have examined the results of a recently reported clinical trial of tacrine. For this purpose we have analysed the first half of our cross-over trial as a parallel group study involving 73 patients. The previously reported significant changes in MMSE in drug versus placebo groups was held up in this analysis (p < 0.001). We have also extracted clinician and carers'' comments from the case notes. These comments were rated blindly by two independent psychologists on a 5-point scale. These retrospective ratings clearly favoured the tacrine group (p = 0.0004) although they have to be viewed critically in view of the post-hoc nature of the procedure. It is concluded that tacrine did indeed appear to produce clinically relevant improvements when compared to placebo and that these were evident in parallel groups as well as in our previous cross-over trial. They could not be attributed to changes in mood.