onthe assessment of dementia—ie, on thevalue of investigations for casemanagement. Copeland comments thatthis study was based on preselectedneurological referrals, and concludesthat studies on unselected samples arerequired. We conducted such a study,in which we systematically informed thelocal general practitioners andencouraged them to refer all theirpatients with suspected dementia aged65 and older. We reported on 200consecutively referred patients, of which170 met the