Memory Tests for the Diagnosis MCI - Systematic Review and Meta-Analysis

Abstract

Mild cognitive impairment is regarded as a prodrome to dementia, involving subjective memory complaint, greater forgetfulness than one would expect for age and intact activities of daiy living in a patient not meeting criteria for dementia. Numerous short bedside cognitive tests can be used to assess cognition, and they have varying characteristics. We conducted a systematic review of the literature for ‘MCI’ and ‘diagnosis’ or ‘screening’, seeking studies evaluating the diagnostic accuracy of a cognitive test for MCI diagnosis as per published criteria compared to a reference standard diagnosis made by psychiatrist/MDT consensus. Risk of bias in each study was assessed using the QUADAS-2 tool. Where a single test had been evaluated in at least 5 published studies, we conducted bivariate random effects meta-analysis to calculate a summary sensitivity and specificity, and summary receiver-operated curves. Eight cognitive tests (ACE-R, CERAD, CDT-Sunderland, IQCODE, Memory Alteration Test, MMSE, MoCA, and Qmci) were considered for meta-analysis. ACE-R, CERAD, MoCA, and Qmci were found to have similar diagnostic accuracy (sensitivity approximately 80%, specificity approximately 75%) while the MMSE had lower sensitivity (66%), but similar specificity to the other tests. Memory Alteration Test had the highest sensitivity (91%) and equivalent specificity to the other tests. The diagnostic accuracies of the CDT-Sunderland and IQCODE were too heterogeneous for quantitative meta-analysis, and we can only comment that their diagnostic accuracy for MCI is unclear. The present meta-analysis suggests that a) MMSE should not be used for MCI diagnosis b) novel cognitive tests aiming to diagnose MCI should not be compared to the MMSE c) Memory Alteration Test, a short test with a more detailed focus on verbal learning and recall, has the highest sensitivity for MCI. Its short duration of administration makes it worthy of further study for further application in primary care.
Original Publication doi: 10.1002/gps.5016

Full text

Background
MMSE
MoCA
M@T
ACE-R
False Positive Rate
Sensitivity
>10%
1) More than 10% of
MCI patients are
being underdiagnosed
when you use the
MMSE.
2) Clinicians should
avoid using it.
3) MMSE is still
copyrighted.
MCI is regarded as a prodrome to dementia,
involving subjective memory complaint, greater
forgetfulness than one would expect for age and
intact activities of daily living in a patient not
meeting criteria for dementia. Numerous short
bedside cognitive tests can be used to assess
cognition. MMSE is a gold standard diagnostic
validation test.
Faculty of Old Age
Psychiatry Annual
Conference
13-15 March 2019
Nottingham, UK
We conducted a systematic review of the literature for
‘MCI’ and ‘diagnosis’ or ‘screening’, seeking studies
evaluating the diagnostic accuracy of a cognitive test for
MCI diagnosis as per published criteria compared to a
reference standard diagnosis made by psychiatrist/MDT
consensus. Search strategy detected 9823 initial studies
(Fig.1). Where a single test had been evaluated in at least
five published studies (n=59), we conducted bivariate
random effects meta-analysis to calculate a summary
sensitivity and specificity, and SROC curves1.
Methods
Eight cognitive tests were considered for meta-
analysis (Tab.1).
ACE-R, CERAD, MoCA, and Qmci were found to have
similar diagnostic accuracy (sens. approx. 80%,
specif. approx. 75%). The MMSE had lower
sensitivity (66%), but similar specificity to the other
tests (Fig.2). Memory Alteration Test had the
highest sensitivity (91%).
The diagnostic accuracies of the CDT-Sunderland
and IQCODE were too heterogeneous for quantitative
meta-analysis, and diagnostic accuracy is unclear.
Test N Participants
ACE-R 6563
CDT-Sunderland 74263
CERAD 54076
IQCODE 51372
M@T 51485
MoCA 24 4095
MMSE 46 17749
Qmci 51206
1) Memory Alteration
Test2, is a short test
with a more detailed
focus on verbal
learning and recall.
2) M@T has the highest
sensitivity for MCI.
3) QMCI4 is an effective
quick way to screen
for prodromal stage
of dementia
Discussion and Conclusions
1) ACE-R, MOCA and
CERAD have similar
diagnostic
accuracy3,5.
2) Multi-domain tests
are still desirable in
tertiary and memory
clinic settings.
3) ACE-III and MoCa are
copyright free for
clinical use.
Meta-analysis
Tab.1
Fig.2
Fig.1
Results
Memory Tests For The Diagnosis of MCI
A Systematic Review And Meta-analysis
Dr. Alexandre Breton1,2, Dr. Daniel Casey3, Prof. Klaus Ebmeier3,4, Dr. Nikitas Arnaoutoglou5,6
1University Hospitals Plymouth NHS Trust, 2Trinity College, 3University of Oxford, Oxford Health NHS FT,
4Dept of Psychiatry, University of Oxford, 5Cambridgeshire and Peterborough FT, 6Dept of Psychiatry, University of Cambridge
References
1. Breton, A., Casey, D. & Arnaoutoglou, N. A. Cognitive tests for the detection of mild cognitive impairment (MCI), the prodromal stage of dementia: meta-
analysis of diagnostic accuracy studies. Int. J. Geriatr. Psychiatry (2018). doi:10.1002/gps.5016
2. Rami, L., Bosch, B., Sanchez-Valle, R. & Molinuevo, J. L. The memory alteration test (M@T) discriminates between subjective memory complaints, mild
cognitive impairment and Alzheimer’s disease. Arch. Gerontol. Geriatr. 50, 171–174 (2010).
3. Mioshi, E., Dawson, K., Mitchell, J., Arnold, R. & Hodges, J. R. The Addenbrooke’s Cognitive Examination Revised (ACE-R): a brief cognitive test battery for
dementia screening. Int J Geriatr Psychiatry 21, 1078–1085 (2006).
4. O’Caoimh, R. et al. Comparison of the quick mild cognitive impairment (Qmci) screen and the SMMSE in screening for mild cognitive impairment. Age
Ageing 41, 624–629 (2012).
5. Nasreddine, Z. S. et al. The Montreal Cognitive Assessment, MoCA: A Brief Screening Tool For Mild Cognitive Impairment. J. Am. Geriatr. Soc. 53, 695–699
(2005).
nikitas.arnaoutoglou@cpft.nhs.uk