Kaduszkiewicz H et al. (2005) Cholinesterase inhibitors for patients with Alzheimer's disease: systematic review of randomised clinical trials. BMJ 331: 321–327

A systematic review of 22 clinical trials involving cholinesterase inhibitors for the treatment of Alzheimer's disease was recently conducted in Germany by Kaduszkiewicz and colleagues. Their findings indicate that current recommendations for the use of such drugs are not supported by the evidence.

The authors used MEDLINE, EMBASE and the Cochrane Database of Systematic Reviews to identify randomized, double-blind, placebo-controlled trials that assessed the clinical efficacy of donepezil, rivastigmine or galantamine in terms of outcome in patients with Alzheimer's disease. Three researchers independently assessed each study for methodological quality using a pre-determined checklist of criteria, then discussed their findings to achieve joint assessments.

Although 19 of the 22 trials reported significant (though moderate) differences between the treatment groups and placebo, the authors found a number of methodological flaws in the trial designs. Trials commonly used several primary endpoints with no correction for multiple comparisons. Consequently, it was found that after correction for multiplicity, two of the five trials on rivastigmine did not show a significant benefit in terms of their primary endpoints. When trials used the last-observation-carried-forward method for endpoint analyses, at least eight had included data from dropouts, which might have influenced the endpoint analysis.

The authors suggest that because of poor methodology and limited clinical benefits reported by trials, the scientific basis for the current recommendations on the use of cholinesterase inhibitors in Alzheimer's disease is open to question.