The use of surrogate respondents to obtain questionnaire data in case-control studies of neurologic diseases

In 1984-1985, the authors assessed the reliability of surrogate respondents to provide interview data for the specific items of a case-control study of Alzheimer's disease conducted in Italy. For all questions of the interview, responses of 52 non-demented subjects were compared to responses of their next-of-kin. In 21-27% of the pairs the next-of-kin was unable to answer questions about general anesthesia, antacid drug use, and age of mother and father at index birth. However, the surrogate respondent was able to answer 45 of 57 tested items with agreement greater than 80%. Questions about use of hard liquor and behavior pattern yielded agreement of 71-75%, while those about number of jobs, and number of cigarettes per day yielded 62-63% agreement. For those who provided information about antacid drug use, agreement was poor. These findings are encouraging for the use of next-of-kin respondents in case-control studies of Alzheimer's disease or other neurologic conditions for which the subject cannot provide historical information.