Domainspecific versus generalized cognitive scree

Authors: Nele Demeyere M J Riddoch E D Slavkova K Jones I Reckless P Mathieson G W Humphreys

Publish Date: 2015/11/20

Volume: 263, Issue: 2, Pages: 306-315

PDF Link

Abstract

Cognitive assessments after stroke are typically short form tests developed for dementia that generates pass/fail classifications eg the MoCA The Oxford Cognitive Screen OCS provides a domainspecific cognitive profile designed for stroke survivors This study compared the use of the MoCA and the OCS in acute stroke with respect to symptom specificity and aspects of clinical utility A crosssectional study with a consecutive sample of 200 stroke patients within 3 weeks of stroke completing MoCA and OCS Demographic data lesion side and Barthel scores were recorded Inclusivity was assessed in terms of completion rates and reasons for noncompletion were evaluated The incidence of cognitive impairments on both the MoCA and OCS subdomains was calculated and differences in stroke specificity cognitive profiles and independence of the measures were addressed The incidence of acute cognitive impairment was high 76  of patients were impaired on MoCA and 86  demonstrated at least one impairment on the cognitive domains assessed in the OCS OCS was more sensitive than MoCA overall 87 vs 78  sensitivity and OCS alone provided domainspecific information on prevalent poststroke cognitive impairments neglect apraxia and reading/writing ability Unlike the MOCA the OCS was not dominated by left hemisphere impairments but gave differentiated profiles across the contrasting domains The OCS detects important cognitive deficits after stroke not assessed in the MoCA it is inclusive for patients with aphasia and neglect and it is less confounded by cooccurring difficulties in these domainsFollowing stroke cognitive deficits are frequent 1 2 3 4 predictive of recovery 5 6 7 8 9 10 11 12 and interfere with rehabilitation eg due to poor comprehension or spatial attention Cognitive deficits after stroke are also associated with a reduced quality of life 13 14 15 and depression 8 Due to their prevalence and importance early detection is required to facilitate rehabilitationTo facilitate early detection short generalized cognitive screening tools are increasingly adopted The Montreal Cognitive Assessment MoCA 16 17 is one tool which is freely available and easy to administer returning a pass/fail generalized cognition score Though developed for dementia the MoCA has been shown to have better sensitivity in detecting poststroke cognitive impairments than the traditionally used MiniMental Status Examination MMSE 18 19 20 21 However neither the MMSE nor the MOCA assesses common poststroke domainspecific impairments including aphasia visual loss visuospatial inattention neglect apraxia and reading/writing problems Furthermore performance on the tests that are included can be confounded by cooccurring problems For example arguably all of the MoCA subtests require substantial verbal abilities and aphasic patients will fail tests of nonlanguage domains eg memory because of language impairments Similarly patients can fail subtests because they neglect one side of the page eg in the trail making testClinical guidelines emphasize the need to assess performance across different domains of cognition after stroke eg “attention memory spatial awareness apraxia perception”—UK National Institute for Clinical Excellence guideline for stroke care 2013 highlighting the need for domainspecific cognitive assessments Detailed neuropsychological examinations can detect specific cognitive impairments 2 22 Not surprisingly when comparing a short MoCA screen to a detailed battery of neuropsychological assessments the detection rate of cognitive problems was demonstrably lower in the MOCA 23 However detailed batteries are often impractical not designed for acute stroke and very time consuming and need trained examiners for administration who cannot routinely see all patientsA recent review and metaanalysis of test accuracy of cognitive screening tests concluded that there was no clearly superior screening test comparing MoCA ACER MMSE and CAMCOG It should be noted however that none of the screens were strokespecific and the studies that were included focussed on generalized impairments equating cognitive impairments to dementia In addition only 11 of the 35 included studies were conducted in acute stroke 24The Oxford Cognitive Screen—OCS 25 was specifically developed to measure domainspecific cognitive deficits in acute stroke It provides a short cognitive screen covering five cognitive domains including the assessment of important and commonly found strokespecific cognitive problems such as unilateral neglect aphasia and apraxia The reporting structure emphasizes the domain specificity of problems going beyond an overall pass/fail outcome It also goes beyond other measures by being designed to avoid confounding effects within the separate cognitive domains providing ‘aphasia and neglect friendly’ measures of performanceIn this study we compared domainspecific cognitive screening OCS with generalized screening provided through the MoCA in an acute stroke population We examined 1 how well the tools detected strokespecific cognitive impairments and 2 their clinical utility in terms of patient inclusion and generating accurate cognitive profiles for patients with cooccurring deficitsThe Oxford Cognitive Screen OCS is a recently developed strokespecific cognitive screen see 20 for normative data validation and reliability and sensitivity measures of the OCS The OCS is structured around five domains 1 attention and executive function 2 language 3 memory 4 number processing and 5 praxis The tests were designed to be inclusive and uncontaminated by aphasia and neglect when respectively language and spatial attention are not assessed The test is freely available for clinical use and licensed through the University of Oxford’s technology transfer office http//wwwocstestorg The OCS as a domainspecific assessment provides a ‘visual snapshot’ of a patient’s cognitive profile for easy domain level see 25

Keywords: