Interviewing the elderly about their health: validity and effects on family doctor contacts.

Health-related questions in a large interview schedule (designed to examine relationships between customary physical activity, physical and psychological well-being among the elderly) were found to be repeatable (overall agreement was 93%) when used by two independent interviewers. The sensitivity and specificity of the interview in discriminating between people with and without health problems were 80% and 92%, respectively, when compared with a physician's independent assessments. Family doctor case notes under-recorded many of the common health problems of elderly people and could not be used to validate the interviewees' responses. Comparisons between 41 interviewees and 75 age-, sex- and practice-matched controls demonstrated that interviewees were only slightly more likely to consult or request a repeat prescription in the week following the interview (19%) than controls (11%). The 95% confidence intervals of this 8% difference were wide (-6% to 22%) and it might well have occurred by chance. Practices intending to use screening or research interviews with elderly people should be reassured such activities are not associated with large increases in demand for their services.