Abstract
The objective of this study was to examine the diagnostic accuracy of a primary care screening procedure for identifying cognitive impairment in elderly veterans, in comparison with 4 brief standardized neuropsychological tests. The sample included 100 primary care patients who met age and other criteria requiring screening for cognitive impairment. The results indicated that 3 of the tests significantly discriminated normal from mildly impaired status on the Dementia Rating Scale, but the existing procedure failed to correctly identify any cases in the entire sample. Correct classification rates were near 80% for the Mini-Mental State Exam, Clock Drawing Test, and both Trail Making Test (TMT)-A and TMT-B, with high specificity but variable sensitivity. TMT-B produced good results across eight predictive validity indicators when a cutoff of 3 minutes to completion (1 SD) was used to identify cases. There was no evidence to support the current interview-based screening procedure. Additional research with brief standardized screening is encouraged.
