PURPOSE OF THE STUDY: This report synthesizes existing evidence to compare the accuracy of various Alzheimer's disease (AD) diagnostic approaches. DESIGN AND METHODS: Meta-analyses and reviews of diagnostic accuracy of AD were identified through a search of the PubMed and Cochrane Library databases using the keyword combinations of "sensitivity specificity Alzheimer's disease diagnosis" and "accuracy of Alzheimer's disease diagnosis." RESULTS: From 507 abstracts initially identified, 41 systematic reviews or meta-analyses were selected. Cerebrospinal fluid-tau demonstrated variable sensitivity (range 73.3%-100%) and specificity (range 70.0%-92.4%) in diagnosing AD when compared to neuropathological verification of clinical criteria for AD. Various positron emission tomography approaches showed a similar range of sensitivity (range 80.0%-100%) and specificity (range 62.0%-90%) as diagnostic protocols. IMPLICATIONS: Issues that remain in the study of AD diagnosis include the need to determine the comparative effectiveness of diagnostic approaches. Variations in study quality make empirically derived conclusions about the diagnostic accuracy of existing approaches tenuous.
PURPOSE OF THE STUDY: This report synthesizes existing evidence to compare the accuracy of various Alzheimer's disease (AD) diagnostic approaches. DESIGN AND METHODS: Meta-analyses and reviews of diagnostic accuracy of AD were identified through a search of the PubMed and Cochrane Library databases using the keyword combinations of "sensitivity specificity Alzheimer's disease diagnosis" and "accuracy of Alzheimer's disease diagnosis." RESULTS: From 507 abstracts initially identified, 41 systematic reviews or meta-analyses were selected. Cerebrospinal fluid-tau demonstrated variable sensitivity (range 73.3%-100%) and specificity (range 70.0%-92.4%) in diagnosing AD when compared to neuropathological verification of clinical criteria for AD. Various positron emission tomography approaches showed a similar range of sensitivity (range 80.0%-100%) and specificity (range 62.0%-90%) as diagnostic protocols. IMPLICATIONS: Issues that remain in the study of AD diagnosis include the need to determine the comparative effectiveness of diagnostic approaches. Variations in study quality make empirically derived conclusions about the diagnostic accuracy of existing approaches tenuous.