[Memory: therapeutic approach. Clinical evaluation]

Memory assessment in pharmaceutical trials has to date been mainly performed in Alzheimer's disease. North American and European Medical Evaluating Agencies recommend guidelines. Tests need to be simple and short, and sensitive enough to assess changes over a wide range of severity to avoid floor or ceiling effect. They require inter-rater reliability and face validity. They should be at least appropriate for different languages and culture, correlate well with universally accepted estimates of a function and have several parallel forms available to avoid a training effect. Memory is assessed with sub-tests of composite scales assessing a number of cognitive functions including episodic memory. The ADAS-Cog has become the standard instrument since it was used for the approval of tacrine and other cholinesterase inhibitors. Some studies have assessed more specifically different sub-systems of memory. However, autobiographical, remote or prospective memory, as well as metamemory, has not been explored in pharmaceutical trials.