Data were derived from the Cochrane Collaboration meta-analyses of the efficacies of ginkgo, donepezil, rivastigmine and galantamine on changes in cognitive function in patients with dementia and, where necessary, were transformed to standardized mean differences. The proportion of patients discontinuing trials was used as a proxy measure of tolerability. Outcomes were assessed after 6 months of treatment. Trial data for cholinesterase inhibitors were more consistent than those for ginkgo, particularly regarding patient populations and outcome measures. Significant benefits on cognition vs. placebo were seen with donepezil, 5 and 10 mg, rivastigmine, 6-12 mg, and galantamine, 16 and 24 mg. Significant benefit vs. placebo with ginkgo was seen only when all doses were pooled. Similar proportions of patients discontinued treatment with ginkgo and placebo. Cholinesterase inhibitors were also well tolerated, although a significantly greater proportion of patients receiving active treatment discontinued vs. placebo with some doses. An evidence-based medicine approach, taking into account the quality of clinical trials, is essential when assessing the safety and efficacy of medications.
Copyright 2004 S. Karger AG, Basel