“…The sampling frame for most studies involved administrative databases, such as voter registration lists, telephone listings, Medicare registries, or a private company's marketing list 2,3,5,7,12,13,16–20,26,29–34 . Fewer studies used convenience sampling with subjects recruited from such sources as physician practices, hospital patients being discharged, or seniors' clubs 1,6,8,11,27,28 . A number of articles reported on the results of secondary data analysis from previous large population‐based surveys, such as the Established Populations for Epidemiologic Studies of the Elderly (EPESE), 13,17,20 Monongahela Independent Elders Survey (MoVIES), 5,12,26,32,33 Pennsylvania Pharmaceutical Assistance Contract for the Elderly (PACE), 16,18 and others 2,7,14,19,29,31 …”