Telephone Assessment of Cognition After Transient Ischemic Attack and Stroke
Modified Telephone Interview of Cognitive Status and Telephone Montreal Cognitive Assessment Versus Face-to-Face Montreal Cognitive Assessment and Neuropsychological Battery
Background and Purpose—Face-to-face cognitive testing is not always possible in large studies. Therefore, we assessed the telephone Montreal Cognitive Assessment (T-MoCA: MoCA items not requiring pencil and paper or visual stimulus) and the modified Telephone Interview of Cognitive Status (TICSm) against face-to-face cognitive tests in patients with transient ischemic attack (TIA) or stroke.
Methods—In a population-based study, consecutive community-dwelling patients underwent the MoCA and neuropsychological battery >1 year after TIA or stroke, followed by T-MoCA (22 points) and TICSm (39 points) at least 1 month later. Mild cognitive impairment (MCI) was diagnosed using modified Petersen criteria and the area under the receiver-operating characteristic curve (AUC) determined for T-MoCA and TICSm.
Results—Ninety-one nondemented subjects completed neuropsychological testing (mean±SD age, 72.9±11.6 years; 54 males; stroke 49%) and 73 had telephone follow-up. MoCA subtest scores for repetition, abstraction, and verbal fluency were significantly worse (P<0.02) by telephone than during face-to-face testing. Reliability of diagnosis for MCI (AUC) were T-MoCA of 0.75 (95% confidence interval [CI], 0.63–0.87) and TICSm of 0.79 (95% CI, 0.68–0.90) vs face-to-face MoCA of 0.85 (95% CI, 0.76–0.94). Optimal cutoffs were 18 to 19 for T-MoCA and 24 to 25 for TICSm. Reliability of diagnosis for MCI (AUC) was greater when only multi-domain impairment was considered (T-MoCA=0.85; 95% CI, 0.75–0.96 and TICSm=0.83, 95% CI, 0.70–0.96) vs face-to-face MoCA=0.87; 95% CI, 0.76–0.97).
Conclusions—Both T-MoCA and TICSm are feasible and valid telephone tests of cognition after TIA and stroke but perform better in detecting multi-domain vs single-domain impairment. However, T-MoCA is limited in its ability to assess visuoexecutive and complex language tasks compared with face-to-face MoCA.
- mild cognitive impairment
- modified Telephone Interview of Cognitive Status
- Montreal Cognitive Assessment
- telephone testing
- vascular cognitive impairment
Cognitive impairment is prevalent after stroke,1,2 and there is a need for short tests of cognition, including telephone tests, because lengthy neuropsychological batteries often are not feasible in large studies.
The Montreal Cognitive Assessment (MoCA)3 recommended by the National Institute of Neurological Disorders and Stroke–Canadian Stroke Network Vascular Cognitive Impairment Harmonization Standards working group4 is sensitive to mild cognitive impairment (MCI) in transient ischemic attack (TIA) and stroke5–7 and may be administered by telephone, but there are no published validations. The Telephone Interview of Cognitive Status (TICS) was originally developed as a short cognitive test for face-to-face or telephone administration,8,9 but there are few reports after stroke.10,11
Therefore, we assessed the telephone MoCA (T-MoCA; items not requiring the use of a pencil and paper or visual stimulus) and the modified TICS (TICSm) in community-dwelling patients with TIA or stroke who had recently undergone face-to-face MoCA and National Institute of Neurological Disorders and Stroke–Canadian Stroke Network Vascular Cognitive Impairment neuropsychological battery.4
Subjects and Methods
Patients were participants in the Oxford Vascular Study (2002–), a prospective population-based study of all acute vascular events12 approved by the local ethics committee. Between August 2009 and November 2010, consecutive community-dwelling patients were invited at their routine 1-or 5-year follow-up (at which they had the MoCA)3 to undergo the National Institute of Neurological Disorders and Stroke–Canadian Stroke Network Vascular Cognitive Impairment Harmonization Standards Neuropsychological Battery,4 which was used to define MCI by modified Petersen criteria as single-domain (amnestic and nonamnestic) and multiple-domain (amnestic and nonamnestic) as described previously.6
At least 1 month after neuropsychological testing, subjects were telephoned and invited to complete the T-MoCA and TICSm presented in counterbalanced order (Supplementary Table I). For the MoCA sustained attention task in which subjects tap on the desk during face-to-face testing, subjects were instructed to tap the side of the telephone with a pencil. Telephone testing was performed by researchers (S.W. and F.C.) blinded to face-to-face data.
For MoCA data acquired by telephone, we considered the T-MoCA (total 22 points) and, separately, the T-MoCA-Short (verbal fluency, recall, and orientation; total 12 points) recommended by the National Institute of Neurological Disorders and Stroke–Canadian Stroke Network Vascular Cognitive Impairment Harmonization Standards Working Group.4 The additional point for low education3 was added to the face-to-face score but not to telephone scores.
Differences between scores on MoCA subtests performed over the telephone vs face-to-face were assessed using the Wilcoxon signed-rank test. Prediction of MCI by T-MoCA, T-MoCA-Short, and TICSm was assessed using the area under the receiver-operating characteristic curve.
Ninety-one nondemented patients (73.4±7 years; stroke 56%; National Institutes of Health Stroke Scale score 0.95±1.5; 63% had education <12 years) completed face-to-face MoCA and neuropsychological battery at a mean of 3.1±1.9 years after index event, and 73 had telephone testing at least 1 month after face-to-face test. MoCA repetition, abstraction, and verbal fluency were worse (all P<0.02) by telephone than face-to-face (Supplementary Table I), even after exclusion of patients with overt hearing problems.
Twenty-seven (40%) of 68 patients with neuropsychological battery and telephone test had a diagnosis of MCI (single domain, 15; multiple domain, 12). MCI patients were older (77.8±9.1 vs 70.0±10.1 years; P=0.002), had low education (22/27 vs 23/41; P=0.03), and had experienced a stroke (17/27 vs 16/41; P=0.08). All telephone tests differentiated between no MCI, single-domain MCI, and multiple-domain MCI (Supplementary Table II).
Reliability of detection of MCI (area under the curve; Figure) for T-MoCA was 0.75 (95% confidence interval [CI], 0.64–0.87; 0.83, 0.73–0.92 for the same 22 items extracted from face-to-face MoCA), for T-MoCA-Short was 0.72 (95% CI, 0.60–0.84), for TICSm was 0.79 (95% CI, 0.68–0.90), and for face-to-face MoCA was 0.85 (95% CI, 0.76–0.94).
For multiple-domain MCI only, areas under the curve were 0.85 for T-MoCA (95% CI, 0.75–0.96), 0.85 for T-MoCA-Short (95% CI, 0.75–0.96), 0.83 for TICSm (95% CI, 0.70–0.96), and 0.87 for face-to-face MoCA (95% CI, 0.76–0.97).
Optimal sensitivities and specificities for MCI were achieved for T-MoCA cut-offs at around 18 to 19, for T-MoCA-Short around 10 to 11, and for TICsm around 24 to 25, with cut points 1 to 2 points lower for multi-domain impairment (Table).
The T-MoCA and the TICSm had similar area under the curve curves and reasonable sensitivity and specificity for MCI, although the T-MoCA-Short performed less well and face-to-face MoCA was best. Performance was better and more similar for all tests in detecting multiple-domain impairment.
Our findings show that both the T-MoCA and the TICSm are valid tests for assessing cognition after TIA and stroke, although certain MoCA subtests (abstraction, verbal fluency, and repetition) seemed to be adversely affected by telephone administration. Sensitivity/specificity of the T-MoCA for MCI thus was better when the same 22 items were extracted from the face-to-face MoCA, and this would be a valid test for face-to-face assessment of visually impaired patients. The T-MoCA-Short performed less well than the other tests in detection of single-domain MCI.
Our study has several limitations. First, we included only community-dwelling subjects who were tested some time after a relatively mild cerebrovascular event, and telephone testing may be more difficult for patients with more severe stroke. Second, optimal cut-offs for the T-MoCA and TICSm will vary with different definitions of MCI.6 Finally, neuropsychological battery/telephone testing was only performed with a small number of patients because of resource constraints, and further large studies are needed to confirm our findings.
In conclusion, both the T-MoCA and the TICSm seem to be valid telephone tests of cognition but perform better in detecting multiple-domain vs single-domain MCI in community-dwelling patients with TIA and stroke. The T-MoCA is limited by inability to test visuoexecutive items and by the impact of telephone administration, and face-to-face MoCA is preferred when clinic assessment is feasible.
Sources of Funding
The Oxford Vascular Study is funded by the UK Stroke Association, the Dunhill Medical Trust, the National Institute of Health Research (NIHR), the Medical Research Council, and the NIHR Biomedical Research Centre, Oxford (which also supports Dr Pendlebury).
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.112.673384/-/DC1.
- Received August 7, 2012.
- Accepted August 31, 2012.
- © 2012 American Heart Association, Inc.
- Narasimhalu K,
- Ang S,
- De Silva DA,
- Wong MC,
- Chang HM,
- Chia KS,
- et al
- Hachinski V,
- Iadecola C,
- Petersen RC,
- Breteler MM,
- Nyenhuis DL,
- Black SE,
- et al
- Pendlebury ST,
- Cuthbertson FC,
- Welch SJ,
- Mehta Z,
- Rothwell PM
- Pendlebury ST,
- Mariz J,
- Bull L,
- Mehta Z,
- Rothwell PM
- Brandt J,
- Spencer M,
- Folstein M