Key methodological features of randomized controlled trials of Alzheimer's disease therapy - Minimal clinically important difference, sample size and trial duration

Background: The results of clinical trials are routinely presented in terms of statistical significance, which may or may not indicate clinical significance. Analysis of the minimal clinically important difference (MCID) of cognitive scales has received little attention to date. Objectives: By reviewing the key methodological features (sample size, duration, statistical and clinical significance) of clinical trials examining the efficacy of tacrine in the treatment of Alzheimer's disease (AD), we assessed their ability to detect clinically important changes in cognition. Design: The value for the MCID of the Mini-Mental State Examination (MMSE) was determined by surveying specialists in neurology and geriatric medicine. This value was then used to interpret the clinical significance of the results of published randomized controlled trials (RCTs) assessing the efficacy of tacrine in the treatment of AD and to retrospectively determine their optimal sample size and trial duration. Results: The mean survey MCID for the MMSE was 3.72 (95% confidence interval 3.50-3.95) points. Only 2 of 12 tacrine RCTs using the MMSE found a statistically significant difference in MMSE scores for patients taking tacrine compared with those taking placebo. These improvements were not clinically significant when compared with the survey MMSE MCID. For parallel trials of tacrine in AD, the smallest sample size and minimum trial duration required to demonstrate a clinically significant difference were calculated to be 53 subjects and 1 year, respectively. Five of the 7 parallel trials met the required sample size; however, none of them met the criteria for trial duration. Conclusions: When using the MMSE as an outcome measure, no tacrine trial reported results that were clinically significant as perceived by clinicians working with dementia patients. Application of a range of plausible MCIDs to the parallel design RCTs also demonstrated that 2 of 7 of these trials did not have sufficient sample size, and none had sufficient duration of treatment to reliably detect clinically meaningful changes in cognition. Future clinical trials in this area will need to incorporate the evolving knowledge of MCIDs in order to increase their chance of detecting clinically relevant results. Copyright (C) 1999 S. Karger AG, Basel.