In ten Italian centers--hospitals and geriatric institutions--130 outpatients and inpatients with a diagnosis of Alzheimer's disease were recruited for a randomized, double blind, placebo controlled clinical trial with 1-acetylcarnitine. In planning the trial, we had to deal with some important and largely open issues.
Diagnosis: we decided not to use neuropsychological tests (NPT) in the diagnostic process, both for the unknown risk of false positive and false negative rate and to improve the feasibility of the trial.
Follow-up: it must be representative of the disease and consequently we chose to follow the patients for one year (assessment at baseline, 3rd, 6th and 12th month) in spite of a possible high rate of drop-outs.
Assessment: to assess the patients' outcome we used NPT and behavioural scales. However, the validity, reliability and feasibility of these instruments are rarely discussed and their usefulness as indicators of the relevant aspects of the disease needs a careful evaluation.
Statistical analysis: in this type of trial there is both a high risk of alpha-error, in view of the high number of NPT and behavioural scales used to assess the drug efficacy, and a high risk of beta-error connected with the usually small sample size. Thus, the role of small trials in defining the risk/benefit ratio of a treatment needs to be discussed.