John Breitner

Publication Detail

The publication detail shows the title, authors (with indicators showing other profiled authors), information on the publishing organization, abstract and a link to the article in PubMed. This abstract is what is used to create the fingerprint of the publication. If any grants are referenced by the publication, they will be listed here as well.

Performance characteristics of a two-stage dementia screen in a population sample.

A S Khachaturian; J J Gallo; J C Breitner (Profiled Author: John Breitner)

Department of Mental Hygiene, School of Hygiene and Public Health, The Johns Hopkins University, 624 North Broadway, 8th Floor, Baltimore, MD 21205, USA.
Journal of clinical epidemiology 2000;53(5):531-40.

We analyzed the performance of a two-stage screening protocol [the Modified Mini-Mental State Exam (3MS) or the Informant Questionnaire for Cognitive Decline (IQCODE), and the Dementia Questionnaire (DQ)] in a weighted stratified sample of 839 subjects from a population survey of dementia in Cache County, UT. The subjects were subsequently examined using a standardized diagnostic assessment protocol. Using the method of receiver operating characteristic (ROC) analysis, the main outcome measure was area under the ROC curve (AUC). The overall AUC estimates were 0.956 (95% confidence interval 0.943-0.968) for the 3MS/IQCODE and 0.945 (0.931-0.960) for the DQ. After adjustment for age, the efficiency of the both 3MS/IQCODE and the DQ was better for subjects with genotype epsilon4/epsilon4 at APOE, better among women, and better in those with two or more years duration of dementia. The optimal threshold for this two-stage screen yielded sensitivity and specificity estimates of 91.0% and 92.0%, respectively.

Scientific Context

This section shows information related to the publication - computed using the fingerprint of the publication - including related publications, related experts and related grants with fingerprints representing significant amounts of overlap between their fingerprint and this publication. The red dots indicate whether those experts or terms appear within the publication, thereby showing potential and actual connections.