Joined: Fri Aug 11, 2006 1:46 pm
Location: SF Bay Area (Northern CA)
Using Bender Gestalt Test to Differentiate LBD from AD
This abstract (just published) is about using something called the Bender Gestalt Test to differentiate whether someone has DLB or AD. I don't know if this is a test often given during neuropsych exams. If anyone has experience with this test, please share!
Dementia and Geriatric Cognitive Disorders. 2007 Mar 12;23(4):258-263
Utility of the Bender Gestalt Test for Differentiation of Dementia with Lewy Bodies from Alzheimer's Disease in Patients Showing Mild to Moderate Dementia.
Murayama N, Iseki E, Yamamoto R, Kimura M, Eto K, Arai H.
Department of Psychiatry, Juntendo Tokyo Koto Geriatric Medical Center, Juntendo University School of Medicine, Tokyo, Japan.
Aims: We examined the utility of the Bender Gestalt Test (BGT) for the differentiation of dementia with Lewy bodies (DLB) from Alzheimer's disease (AD), comparing BGT scores between DLB and AD patients showing mild to moderate dementia.
Methods: Eighteen DLB patients, 36 AD patients controlled by age, years of education, Clinical Dementia Rating and Mini Mental State Examination scores, and 21 nondemented elderly participants controlled by age and years of education were subjected to the BGT. Their BGT performances were scored according to the Pascal-Suttell method.
Results: The DLB group showed significantly higher (that is worse) BGT scores than the other groups. When a cutoff point of 98 was used to differentiate DLB from AD, the patients exceeding 98 were 94% in the DLB group, 17% in the AD group and 0% in the control group. The sensitivity and specificity of this cutoff point were 0.94 and 0.89, respectively.
Conclusion: The BGT is a useful neuropsychological test to differentiate DLB from AD.
Pub Med ID: 17351317
A note about the term "sensitivity"... This is a statistical term that refers to accuracy. The term describes how well a test identifies everyone who really has the underlying disorder. For example: let's suppose I have 1000 people and I test all of them for LBD. If 50 of the people in that group "really" have LBD (meaning, in this case, that an autopsy would reveal the pathological indicators that doctors accept as indicating LBD), and if the test identifies all 50 of those folks, then we would say that the test has 100% sensitivity. Another way scientists sometimes put this is to say that the test, under these circumstances, has no "false negatives." In other words, everyone who really has LBD turned up positive. We can be sure that anyone with a negative result does NOT have LBD. (If anyone has a different - more accurate? - view of sensitivity, please share that too!)
In the particular study above, I don't think brain pathology tests were run to "confirm" the LBD diagnosis. The only "confirming" that could be done was to confirm that those with LBD based on their BGT test scores were clinically diagnosed with LBD. This is not especially compelling since many people are diagnosed with LBD pathologically but in life their diagnosis was not LBD.
I'm not sure what specificity means. Does anyone know?