[Variability of efficacy measures in Alzheimer's disease].

The efficacy of medical interventions is their capacity of inducing positive modifications of the natural history of diseases. The natural history of dementia is marked by specific events related to the cognitive and functional decline, but their occurrence is poorly predictable in individual patients being highly variable from patient to patient. For this reason it is difficult that the modest efficacy of available interventions for dementia, or their entity measured by clinical scales, may be perceived in clinical practice or in observational studies. Moreover in randomized clinical studies, the effect of this variability, in analogy to misclassification of exposition and/or disease in case control or cohort epidemiological studies, is that of an underestimation of the true efficacy of interventions.