Results of a UK-wide vignette study with occupational therapists to explore cognitive screening post stroke

There is a paucity of evidence in the UK regarding cognitive screening performed by occupational therapists and whether, and how, cognitive impairments are identified and assessed. The aim of this study was to identify current occupational therapy practice for the assessment of cognitive problems in patients following stroke. Occupational therapists were invited to complete an online vignette study. Participants were asked to: identify any presenting cognitive problems; decide whether to complete cognitive assessments; and list any assessments they would use. Data were analysed using descriptive analysis. A total of 53 occupational therapists from across the UK participated. Participants identified key cognitive issues, but some problems, such as apraxia and attention, were overlooked. A large number of potential assessments were suggested, the most common being the Montreal Cognitive Assessment and Oxford Cognitive Screen. The variation found in occupational therapists' recognition and assessment of cognitive problems has the potential to impact on management and rehabilitation in stroke services, survivor outcomes, education and research.


Introduction
It is estimated that over 100,000 people have a stroke in the UK every year (Stroke Association 2018). The prevalence of first time stroke in adults aged 45+ years is expected to increase by 59% by 2035 (Patel et al, 2018) with people most likely to have a stroke aged 55 years and over (Wang et al, 2013). Cognitive problems are common after stroke (Lincoln et al, 2011) and are reported in over half of stroke survivors six months post stroke (Mellon et al, 2015). Also, importantly, they occur in approximately half of those who have a transient ischemic attack (TIA) (Fens et al, 2013). Cognitive problems encompass impairments of attention, memory, language, visuospatial and executive abilities. Such impairments may vary in severity and indeed may not be significant until after discharge from hospital when the patient returns home and undertakes daily living activities (Patel et al, 2017;Zinn et al, 2004), or when they return to work  and find that they are not independent or need substantial help and support with activities UK clinical guidelines advise screening patients early for cognitive problems (Department of Health, 2007;British Psychological Society, 2010; Royal College of Physicians [RCP], 2016a). The RCP clinical guideline for stroke states: 'People with stroke should be considered to have at least some cognitive impairment in the early phase. Routine screening should be undertaken to identify the person's level of functioning, using standardised measures' (4.3.1.1.A) (Royal College of Physicians, 2016a). Current UK audit data suggests that many stroke survivors are screened before being discharged from hospital (Sentinel Stroke National Audit Programme, 2017). The increasing emphasis on discharging stroke survivors from hospital early has important implications for rehabilitation management within community stroke teams, particularly with regard to how information is passed on and how results of tests are used. There can also be issues around who should take responsibility for follow-up of stroke survivors with cognitive problems. Tang et al (2017), in a study of primary and secondary care, found professionals were unsure who was ultimately responsible for longer term care and highlighted problems around lack of experience in some clinicians in conducting assessments and difficulties in communicating findings.
Occupational therapists (OTs) play an important role in the rehabilitation of stroke survivors, and are regarded as having a key role in cognitive rehabilitation (Govender and Kalra, 2014) particularly amongst Allied Health Professionals. There has been some exploration internationally of cognitive assessment practices undertaken by OTs with stroke survivors.
A previous telephone survey by Korner-Bitensky et al (2011) of OTs in Canada explored assessment and management of cognitive impairments following stroke; they found detection of problems was lower in community-based patient scenarios. They also found that a large number of assessments were used for screening, but highlighted that the most frequently cited was the Mini-Mental State Examination (MMSE). Korner-Bitensky et al (2011) found that few clinicians indicated that they would reassess cognitive function after the initial assessment which may result in missed opportunities to identify changes in cognition, which in turn, impacts on missed opportunities to adjust rehabilitation. In addition, the authors found that, of the clinicians who would repeat a cognitive assessment, they reported that they would use MMSE, which is not an appropriate measure for monitoring cognition post stroke as it is does not predict deterioration or improvement in cognitive function over time (Bour et al 2010).
Korner-Bitensky et al (2011) noted the marked variation in practice between OTs participating in their survey and felt that less emphasis was placed on cognition in community services. Similarly, findings from interviews in the qualitative phase of the present study,  also showed marked variation in the cognitive screening of stroke survivors by OTs working in the community. It seemed that cognitive assessments were selected based on availability and familiarity rather than evidence for appropriateness to the suspected cognitive problem. Similarly, a Danish study that surveyed the choices made by OTs when selecting assessments for screening cognitive deficits during the immediate postacute phase of stroke found only 9% of participants were using standardised assessments (Pilegaard et al, 2014). The authors also found that 13 different screening methods were being used.
The impact of cognition on recovery after stroke has been recognised (McKevitt et al, 2011) and, indeed, the James Lind Alliance cited it first in their list of the top ten priorities relating to life after stroke (Pollock et al., 2012). In the recent European Stroke Action Plan (ESAP) for 2018-2030 the variability in access to rehabilitation following stroke within and between countries (Norrving et al, 2018) was highlighted and the effective management of cognitive impairment following stroke was identified as a research and development priority. In addition, ESAP also agreed that the improvement of community rehabilitation is a target.
Despite emerging international research findings, relatively little is known about the identification of cognitive problems by OTs in the UK, the assessments that OTs routinely undertake or how these are used to shape rehabilitation, particularly in the community. Thus the aim of this study was to identify current OT practice in the UK for the assessment of cognitive problems in stroke survivors, in the community. We wanted to specifically explore: (i) OTs' identification of common cognitive problems after stroke, (ii) OTs' intention to screen stroke survivors, and (iii) which cognitive assessments were suggested and the reasons for this choice.

Recruitment
OTs were invited to take part in the study via advertisements circulated within the Royal College of Occupational Therapists' Specialist Section in Neurological Practice and via a professional Twitter account. We aimed to recruit 40 OTs; this was a pragmatic target in accordance with the time and resources available for this exploratory study.
The eligibility criteria were that participants were OTs, working primarily in UK community stroke services and willing to give informed consent. social support systems. The vignettes were generated to identify OTs' current knowledge and understanding of cognition and cognitive assessment. OTs were asked to record whether the patient had a cognitive problem and their thoughts around screening assessment. In some scenarios, they were asked what further assessments might be administered, and to describe the likely implications of the given test results for patient management. The vignettes were circulated amongst the steering group for comments and iteratively developed.

Development of questionnaire
Each vignette was set out slightly differently to minimise response bias and fatigue. In some vignettes, we provided a Montreal Cognitive Assessment (MoCA) (Nasreddine et al, 2005) score from a previous assessment in hospital, and in others the scenario unfolded over several statements in order to reflect gathering information over time and to find out at what point (if any) the OTs would complete a cognitive assessment with the stroke survivor. Each vignette contained signs and symptoms of two common cognitive problems after stroke such as a memory problem an executive function problem (see Table 3 for further details on this). An example of a 'closed' vignette is shown in Table 1 and an 'unfolding' vignette is shown in Table 2.
Insert Table 1 here Insert Table 2 here.
The question response formats included a 6-point Likert scale (6 was deemed 'very likely') to indicate how likely participants felt the presenting issues were due to cognitive problems and yes/no response formats to record whether participants would administer a cognitive assessment. Participants were asked to complete free text responses to identify any cognitive problems, note any relevant issues for screening, or any other factors they considered relevant. Responses were categorised into cognitive domains based on the definitions provided in the Stroke Association Guide (Stroke Association, 2018). In addition, OTs were asked to provide information about: their NHS banding to note their seniority, the higher the band, the more senior the occupational therapist, (explanation of the NHS clinical bands can be found at https://www.nhsemployers.org/your-workforce/pay-and-reward/jobevaluation/national-job-profiles/allied-health-professionals), how long they have worked clinically with stroke patients, whether the bulk of their work was in a hospital or community setting and their geographical location.
The questionnaire underwent a review with three clinical psychologists (who were not part of the research team or steering group) who had expertise in cognitive rehabilitation in stroke.
This was to ensure the cognitive issues described were accurate, relevant and could be identified from the vignette. Some changes were made at this stage to clarify the wording, to ensure cognitive issues were central. A final review was conducted by the research team before formally pilot testing the questionnaire.
Three clinical OTs, recruited via contacts within the East Midlands, agreed to pilot test the questionnaire. Further minor changes were made to the wording of certain vignettes, based on their feedback. Furthermore their feedback on the length of the questionnaire and the time needed to complete it, led the research team to remove two vignettes. This was in order to ensure the questionnaire took approximately 30 minutes to complete. The data collected in the pilot study was not included in the final data analysis.
The vignettes and questionnaire were uploaded onto Bristol Online Survey. Participants were emailed a participant information sheet, informed consent was taken online and a link to the online survey was provided. A written version of the questionnaire was also available on request.

Data Analysis
Descriptive statistics were used to examine the numbers of participants who identified cognitive problems within the vignettes, and the number of participants who, based on the information in the vignettes, indicated an intention to administer a cognitive assessment. We also examined the proportion who stated that they required additional information prior to deciding whether to administer a cognitive assessment, although the research team felt there was sufficient information in the scenarios. We also analysed the combination of proposed cognitive assessments with cognitive problem included within the vignette. Quantitative data were analysed using Statistical Package for the Social Sciences (SPSS) version 24. Free text responses were explored for key themes. The free text responses were coded and analysed using content analysis, following Kippendorff (2004). We coded free text segments into meaning units, which were then condensed to specific codes that coalesced around a meaning unit. These codes were then collapsed into categories and higher level themes.

Vignette Reponses
Identification of cognitive problems and intention to conduct screening assessment Between 85% and 95% (n=45-51) of OTs correctly identified that the issues presented in vignettes 1, 3 and 8 were due to cognitive problems. Vignettes 1, 3 and 8 were also the stroke survivor scenarios where OTs indicated that they would cognitively assess.
Vignette 4 was least likely to be considered as having a cognitive issue, with 60% (n=32) of the participants identifying a cognitive problem. Despite this, 79% (n=42) indicated that they would administer a cognitive assessment in this scenario.
In each of the vignette responses, a higher number of OTs indicated they would administer a cognitive assessment compared to the number of OTs who identified the stroke survivor to have a cognitive issue (Figure 1).
For most of the vignettes, participants indicated that they required additional information regarding the stroke survivor in order to decide whether or not to administer cognitive assessments, with the exception of vignettes 1 (89%, n=47), vignette 3 (81%, n=43) and 8 (89%, n=47) ( Figure 2).
Insert Figure 2 here.

Proposed cognitive screening assessments
Participants proposed a number of assessments they would use to screen for cognitive problems. Table 3 presents the most common assessments proposed by OTs to screen for cognitive problems and the percentages of OTs who would use these assessments in each vignette.
Insert Table 3 As shown in table 3, for each of the vignettes, the OTs said they would use a 'functional' assessment. Although they did not define what they meant by this, the impression given was that they were observing patients undertaking a functional activity (such as a kitchen task).
This was particularly in vignettes 2 and 6 that included a scenario with visuospatial and attention problems.

Free text comments
Participants had the opportunity to provide free text comments. The key issues raised are listed in Table 4 with representative quotations to illustrate the points raised. Table 4 Discussion In this exploratory study, data were gathered on UK OTs' recognition of cognitive problems in stroke survivors (based on eight vignettes), their intention to administer cognitive assessments based on that recognition, and the assessments most frequently used. While there are research findings from other countries, this study and other results , provides a foundation for understanding UK OTs' cognitive assessment practices.

Insert
Although the aim was to recruit 40 OTs working in stroke teams primarily in the community, we recruited 53 OTs from throughout the UK. We believe this demonstrated the interest of the OTs who responded in the topic of cognitive problems after stroke.
Our results demonstrated inconsistencies in the identification of common cognitive problems.
Although the majority of the OTs identified difficulties such as memory loss and visual neglect, apraxia and attention were less well recognised. This has implications for rehabilitation as these cognitive deficits may not be factored into or addressed in the rehabilitation programme, which suggests variability in rehabilitation as identified in the development of ESAP (Norrving et al, 2018). There were also discrepancies in the identification of the cognitive problems within the same domain, which might be explained by the presentation of other symptoms in the vignette (such as aphasia) which might have confused the OTs.
In some vignettes, where the OTs were able to identify a cognitive problem, they felt that the stroke survivors' presenting difficulties were not a result of cognitive problems. This suggests that there are individual differences in the knowledge of cognitive problems, and that there is some uncertainty in how to assess cognition when stroke survivors display perceptual or language issues.
Our findings also showed that not all OTs would consider routinely assessing all stroke survivors for cognitive problems and that the likelihood of assessment varied depending on the signs and symptoms described. OTs were most likely to conduct an assessment when they felt stroke survivors had memory loss and executive function difficulties, and least likely when they believed stroke survivors had visuospatial issues. The most common reasons for not administering assessments were because the OTs felt 'functional' assessments would be more suitable for older patients or those with speech or perceptual issues would find assessments difficult. Although the OTs did not define what they meant by 'functional assessments', the impression is that they were observing patients undertaking a functional activity (such as dressing). This impression would be very much in keeping with what Koh et al (2009) reported in their survey of Australian OTs whereby a number of OTs felt that performance on paper and pencil activities may not be transferrable to activities of daily living. Some OTs did not consider assessments to be appropriate when the stroke survivor was demonstrating perceptual issues as they did not perceive this to be a cognitive issue.
Consistent with other study findings (Pilegaard 2014, Ablewhite et al 2019) a variety of assessments were suggested by OTs. The most commonly suggested were the MoCA and the OCS, followed by the ACE and the CAM (Addenbrooke's Cognitive Assessment (Mioshi et al, 2006)and Cognitive Assessment of Minnesota (Rustad et al, 1993)). Key reasons for choosing assessments included speed and ease of use, and familiarity which is consistent with other findings . The MMSE was only proposed once and for only one vignette, unlike findings by Koh et al (2009) andKorner-Bitensky et al (2011) in which the MMSE was one of the most popular tools reported. Functional assessments were also a popular choice when OTs considered exploring cognitive problems after stroke either, instead of, or as well as, another assessment. Functional assessments were most commonly suggested if the stroke survivor was retired or had language issues, or because they believed they would highlight both cognitive issues and rehabilitation goals. This was also similar to the results from the Koh et al (2009) who found that OTs relied heavily on clinical observations. These findings offer some preliminary insights into the identification and management of cognitive problems after stroke in the UK. They must, however be considered in light of the study's limitations. There is a possibility of sampling bias, as participants with a particular interest and knowledge of cognitive problems after stroke would have self-selected to participate in the research. We were also very interested in gaining responses from OTs who worked in the community but, as this was an online tool, we could not control this. However, it was clear from the responses that the majority of the respondents were currently employed in, or had recent experience of working in the community. We also did not define any of the terms used but left the OTs to use these as they would routinely and as they thought appropriate. It was clear from the findings that OTs used many terms interchangeablynotably assessments, tools, and screening measures.
Notwithstanding these limitations, this is the first UK study to examine the day to day assessment of cognitive problems by OTs and the findings are in keeping with literature from other countries where similar research has been undertaken. The results suggest there is interest in the topic and the real possibility of professionally developing OT skills further in identifying cognitive problems, which could potentially improve the rehabilitation of people with stroke. Not only would this have the potential to improve the independence and quality of life for stroke survivors but would have implications for caregivers and support services.
Given the current issues with clinical psychology staffing and provision for stroke in the UK (RCP, 2016b), and the role of OT in providing ecologically valid assessments (Crist, 2015).
OTs would be ideally placed to manage day to day cognitive issues in people with stroke.

Conclusion
There is variation within UK OTs recognition and assessment of cognitive problems which has potential to impact on rehabilitation in stroke services, survivor outcomes, education and research.

Table 1-Example of a closed vignette
Mr Taylor is 55 years old and works full-time as a bank-clerk. Mr Taylor had a right frontal hemisphere stroke and had a MoCA score of 22 prior to discharge; he dropped points on trail making, letter fluency, attention and abstraction. He has coped well with most aspects of his phased return to work, but seems to cope less well when things don't go to plan, e.g. if he has to tackle an unexpected problem or change his plans at the last minute. His daughter reports that Mr Taylor also sometimes does things that seem a little impulsive, such as making large purchases over the internet.