COMPARISON OF CLINICAL STATE, RETROSPECTIVE INFORMANT INTERVIEW AND THE NEUROPATHOLOGIC DIAGNOSIS OF ALZHEIMERS-DISEASE

This study compared clinical diagnosis of dementia derived from medica l records and a retrospective version of the Informant Questionnaire o n Cognitive Decline in the Elderly (IQCODE) against the criterion of n europathology. Subjects comprised 50 consecutive patients who died in a geriatric hospital in Melbourne and underwent necropsy. All informan ts were seen within 3.5 years of death. Discordance between clinical a nd pathological diagnosis occurred in seven (14%) cases. The IQCODE sc ores for the group with a clinical diagnosis of Alzheimer's disease (m ean +/- SD) were 4.28 +/- 0.78 and were significantly different from t he non-demented group, 3.39 +/- 0.51, p < 0.001. Similarly, IQCODE sco res were significantly different between the groups with and without a pathologic diagnosis of Alzheimer' s disease, 4.21 +/- 0.70 and 3.51 +/- 0.79 respectively, p < 0.01. Using immunohistochemical techniques instead of conventional methods to make a diagnosis of Alzheimer's dis ease, there were significant differences in IQCODE scores between the Alzheimer's group, 4.13 +/- 0.83, and the group with normal neuropatho logy, 3.45 +/- 0.54, p < 0.01. These data support the conclusion that when used retrospectively, the IQCODE is a valid screening test for de mentia using neuropathological diagnosis as the criterion.