A method for estimating duration of illness in Alzheimer's disease.

We developed a set of questions for generating an estimate regarding the date of first symptoms to the nearest half-year. Physicians then revised this estimate in conjunction with medical record review and patient/informant interviews, and by testing the estimate by recall of life events. One experienced examiner rated 36 patients, and each was independently rated by a second, less experienced rater. The physician ratings were compared to each other and to an unstructured caregiver estimate of duration using Lin concordance coefficients. There was excellent agreement between independent physician raters (rho = 0.95, p < 0.001). Caregiver estimates of duration were usually shorter because of failure to relate the first symptoms to the onset of disease.