Premium
P2‐431: Trends in hospitalization associated with Alzheimer's disease in the United States
Author(s) -
Gillum Richard,
Obisesan Thomas
Publication year - 2012
Publication title -
alzheimer's and dementia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.713
H-Index - 118
eISSN - 1552-5279
pISSN - 1552-5260
DOI - 10.1016/j.jalz.2012.05.2056
Subject(s) - medicine , demography , medical diagnosis , dementia , hospital discharge , mortality rate , population , disease , gerontology , pediatrics , environmental health , pathology , sociology
entiate normal aging from ADRD, and major depression. The Orange County Vital Aging Program (OCVAP) is a county-based program designed to educate public and professionals, plus provide effective screening for maintaining cognitive health. The present study examined the HBC’s screening performance in the OCVAP sample. Methods: 599 community subjects referred themselves to the OCVAP screening program, and were assessed using the HBC, the MCI Screen (objective cognitive test) and the Functional Assessment Staging Test procedure (FAST). Subjects were classified into FAST stages 1 (normal cognition: N 1⁄4 93), 2 (SCI: N 1⁄4 305), 3 (MCI: N 1⁄4 137) or 4 (Mild Dementia: N 1⁄4 64). Correspondence analysis was used to identify the HBC items with largest value separation between yes and no responses. These items were examined for accuracy classifying normal vs. cognitively impaired groups by dichotomizing FAST staging into groups of FAST 1-2 vs. 3-4. The classification scoring rule was very simple: if any of the items included for scoring were answered yes, then the subject was classified as cognitively impaired; otherwise, they were classified as normal. Results: While the full set of HBC items had sensitivity for cognitive impairment (Sn) 1⁄4 98.1%, specificity for normal cognition (Sp) 1⁄4 6.0%, the reduced item sets identified by correspondence analysis performed as follows: a four item set gave Sn 1⁄4 77.6%, and SP 1⁄4 75.1%, a six item set gave Sn 1⁄4 81.6% and SP 1⁄4 70.9%, and an eight item subset gave Sn1⁄4 90.1% and SP1⁄4 52.0%. Conclusions: The HBC can be reduced to a four-to-eight item questionnaire with very simple interpretation and separate NC/SCI from MCI/Mild dementia with 78%-90% sensitivity and 75%-52% specificity. This preliminary work will be useful for screening in primary care setting. The further optimization will improve the accuracy of its utility.