Use this URL to cite or link to this record in EThOS:
Title: Psychological problems in stroke : prevalence, risk factors, and assessment in the pre-stroke state
Author: Taylor-Rowan, Martin
ISNI:       0000 0004 8498 6902
Awarding Body: University of Glasgow
Current Institution: University of Glasgow
Date of Award: 2019
Availability of Full Text:
Access from EThOS:
Access from Institution:
Background: The psychological impact of stroke is well recognised as being of clinical importance. Historically, this field has received less attention than the physical consequences of stroke, but work designed to develop our understanding of post-stroke psychology is now well underway. Much of the current research of post-stroke psychology is overly limited however; little attention has been paid to the potential impact of the pre-stroke state on post-stroke psychology. As a consequence, fundamental information in relation to the pre-stroke state is lacking, ranging from the prevalence and relevant risk associations of various psychological and physical conditions, to the validity and optimal use of pre-stroke state assessment methods. The purpose of this thesis is to improve our understanding of the pre-stroke state in relation to these under-researched areas. Method: I conducted a series of studies designed to improve our understanding of the pre-stroke state in the areas of prevalence, risk association, and method of assessment. Specifically, I conducted a diagnostic test accuracy study to evaluate the psychometric properties of two informant questionnaires that can be used to assess pre-stroke cognition: the Informant Questionnaire for Cognitive Decline in the Elderly (IQCODE) and Acquired Dementia 8 (AD8). I conducted a systematic review and meta-analyses to establish pre-stroke depression prevalence and investigate its association with post-stroke depression. Based on the findings of this, I explored the potential use of informant tools for pre-stroke depression assessment by comparing the diagnostic test accuracy of the Stroke Aphasic Depression Questionnaire (SADQ) against the Geriatric Depression Scale (GDS), and the diagnostic test accuracy of the best performing informant questionnaire against that of medical records. I conducted secondary analysis of existing data held in two databases to investigate pre-stroke functioning and pre-stroke frailty. The Anglia Stroke Clinical Network Evaluation Study database was utilised to assess the validity of the pre-stroke modified Rankin Scale (mRS) as a measure of function and explore if reported predictive validity of the tool could be influenced by differences in post-stroke care pathway. The Glasgow Royal Infirmary research database was used to investigate the prevalence of pre-stroke frailty, the validity of a Frailty Index for pre-stroke frailty assessment, and a risk association between pre-stroke frailty and acute post-stroke cognition. Findings: I found that the IQCODE and AD8 are valid tools for assessing pre-stroke cognition. However, when utilised at recommended published cut-points the IQCODE is more specific, while the AD8 is more sensitive to cognitive impairment. There is also potential that application of differing cut-points could improve performance when used in a pre-stroke context. My systematic review and meta-analysis suggested that pre-stroke depression prevalence is around 17% and its presence significantly increases odds of post-stroke depression. In addition, there is evidence that the most commonly used method to assess pre-stroke depression, patient medical records, is likely to lack sensitivity to pre-stroke depression. I explored the use of the SADQ and GDS informant tools for assessment of pre-stroke depression. I found that both tools are valid measures of pre-stroke depression, but the GDS has favourable diagnostic test accuracy properties in comparison to the SADQ; comparative test accuracy performance with medical records is inconclusive, but seems to favour the GDS. Pre-stroke mRS evaluation suggests it has moderate validity as a measure of pre-stroke functioning and has predictive validity that could not be accounted for by differences in care pathway. Pre-stroke frailty prevalence is around 28%, rising to ~80% if the pre-frailty state is considered, and the Frailty Index is a valid measure of pre-stroke frailty that can be completed in almost all stroke patients. Pre-stroke frailty also has an association with lower acute post-stroke cognition that is independent of other established risk factors. Conclusions: In conclusion these findings develop our overall understanding of the pre-stroke state. The IQCODE and AD8 are both valid tools for assessment of pre-stroke cognition; however, they demonstrate contrasting strengths when employed at their recommended cut-points and these cut-points may not be the most optimal when these tools are utilised for pre-stroke assessment. Pre-stroke depression appears prevalent, existing in around one in six stroke patients, and it increases the odds of patients experiencing post-stroke depression. It is possible that informant assessment for detection of pre-stroke depression can outperform patient medical records and the GDS appears to outperform the SADQ in the pre-stroke context; however further work is required to confirm this. The pre-stroke mRS is a valid measure of function but has only moderate validity overall and may not be ideally suited to assessment of function in a pre-stroke context. Pre-stroke frailty may exist in around one quarter of stroke patients, and utilisation of a Frailty index approach appears to be valid. The presence of pre-stroke frailty may also contribute to the poor cognitive performance often observed in patients following acute stroke based on an independent association with lower cognitive performance; hence identification of pre-stroke frailty could be of importance to our understanding of post-stroke psychology.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
Keywords: Q Science (General) ; R Medicine (General)