Validity and Clinical Utility of Different Clock Drawing Test Scoring Systems in Multiple Forms of Dementia

Abstract

The Clock Drawing Test (CDT) has a known potential for the detection of cognitive impairment in populations with dementia, especially Alzheimer disease (AD). Our aim was to compare the clinical utility of 3 CDT scoring systems (Rouleau, Cahn, and Babins) in several pathologies with cognitive compromise from a tertiary center memory clinic. We selected patients with a clinical diagnosis of mild stage AD, behavioral variant frontotemporal dementia (FTD), vascular dementia (VaD), dementia with Lewy bodies (DLB), and Parkinson disease with dementia (PDD). The results showed significant differences between the several diagnoses with the following pattern of results: AD, DLB < FTD, VaD, PDD. Qualitative analysis of clock drawing errors confirmed the stimulus-bound response as a hallmark of AD, while conceptual deficit was significantly more prevalent in patients with AD and DLB. Our results supported the CDT potential as a cognitive screening measure for mild dementia, particularly sensitive to AD and DLB, especially when we used the Cahn scoring system and its analysis of qualitative errors.

Cognitive screening instruments to identify vascular cognitive impairment: A systematic review

“…However, while the specificity of the MMSE was high (more than 90%) for all three, sensitivities were lower and differed between the three studies from 62% with a cut‐off of less than 26 out of 30 in Portugal to 91.8% taking a cut‐off of less than 26.5 in China . The accuracy of the CDT was fair to good, AUC of 0.73 to 0.83, and it had variable sensitivity depending on the scoring approach, ranging from moderate (69%) for the Rouleau method to high (91.2%) for the CLOX1; but generally had low specificity, 58% with the Babins method and 66% using the Rouleau scoring system . The accuracy of other instruments were based on single studies and varied markedly (Table ).…”

“…All the studies were conducted in an outpatient setting, where subjects were recruited by convenience or consecutive sampling from those attending geriatric medicine or neurology‐based memory clinics and dedicated referral centres (eg, a National CADASIL clinic), with additional participants including controls obtained from general practice or other community settings such as volunteer groups. Most studies were conducted in Europe (seven studies), with five in Asia, two in North America, and one in South America …”

“…These included short batteries of cognitive tests such as the MMSE (n = 4 papers), the MoCA (n = 4) and its shortened version, the MoCA Short (n = 1), the BMET (n = 3) and the Executive functioning and Memory Ratio (EMR) derived from the BMET (n = 1), the ACE‐Revised (ACE‐R) (n = 1), and the Cognitive Assessment Battery (CAB) (n = 1), which includes the following subtests: Symbol Digit Modalities Test (SDMT), immediate and delayed recall, the Boston Naming Test, the Stroop Colour Word Test, and the Clock Drawing Executive Test (CLOX) combined with cube drawing; each subtest was examined separately . The most commonly studied single‐item instrument was the CDT ( n = 5) with six different scoring systems reported: the Rouleau, Cahn, Babins and the AD Cooperative Study scoring methods, and two versions of the CLOX, ie, CLOX1 (free drawing) and CLOX2 (copying); the CDT was also combined with the MMSE and the CLOX was combined with the Cube Copying Test . Other individual instruments included the TMT (types A and B) (n = 1), tests of delayed recall (n = 3), including the Delayed Memory Index (DMI) subtest from the Repeatable Battery for the Assessment of Neuropsychological Status Update (RBANS), and four tests of verbal fluency: the F‐A‐S Phonemic (letter) Fluency (n = 1), Categorical Fluency (n = 1), the combination of the Controlled Oral Word Association Test (COWAT) and the Rey Auditory Verbal Learning Test (RAVLT) (n = 1), and the Orthographical fluency subtest from the Taiwanese version of the FAB (n = 1) .…”

“…The BMET components also had acceptable ( α = .71) to good internal consistency ( α = 0.85) when separating VaD from AD . The internal consistency of the CDT varied according to scoring approach with Babin's method giving an α of 0.88 compared with 0.50 using the Cahn and Rouleau scores . Few studies presented IRR or TTR values (see Tables ).…”

“…Two studies examined the discriminant properties of the BMET to separate VMCI from normal cognition, both showing that the BMET had excellent accuracy (AUC of 0.94 at cut‐off value of 13 out of 16). The CDT had fair (AUC 0.66) to good (AUC 0.81) accuracy; used as a stand‐alone test, its sensitivity to differentiate VMCI was low, ranging from 65% to 68.7% but improved to 93.7% when combined with the MMSE . Again, other instruments only had a single study available and results ranged widely (Table ).…”