Aujla, Navneet
(2017)
The role of illness beliefs in understanding adjustment after stroke.
PhD thesis, University of Nottingham.
Abstract
Stroke is a highly prevalent condition. It is a principal cause of adult disability and a leading cause of death in the United Kingdom (UK). The lifetime risk of stroke recurrence is also high, particularly for those with specific risk factors (e.g., hypertension). The long-term management of stroke is multi-faceted and complex. Stroke survivors are often responsible for self-managing their secondary preventive treatment (e.g., oral medication(s)) alongside rehabilitation, which is most beneficial in the first six-months. However, patients undergo a period of psychological adjustment after stroke. This is particularly evident in the early days where mood problems, such as depression, are highly common and affect how well people engage with, and adhere to, their rehabilitation and treatment. The present PhD research was interested in further understanding adjustment after stroke, and in particular, by utilising key evidence from the field of Health Psychology (more specifically, the Common Sense Model (CSM)), examining whether survivors’ perceptions (or beliefs) about their stroke are influential in determining how well people recover.
The overall aims of this work were to:
• Examine and synthesise the current evidence, through a systematic review and meta-analysis on the CSM in relation to physical illnesses to determine whether individual illness belief domains were important predictors of adherence to self-management in adults with acute or chronic conditions;
• Elicit views regarding an assessment method (the Stroke Illness Perception Questionnaire-Revised (Stroke IPQ-R)) that would be suitable and acceptable to a stroke population for accurately measuring peoples’ beliefs in the acute phase (≤three-months post-stroke);
• Apply the CSM (utilising the Stroke IPQ-R) to a population within the acute phase of stroke, to determine whether the model provides a useful framework for understanding survivors’ recovery after stroke (e.g., Health-Related Quality of Life (HRQL), disability, and mood).
This research incorporated three standalone though inter-related studies, the first of which was a systematic review and meta-analysis. The key findings from this meta-analysis indicated that the majority of illness belief domains outlined by the CSM significantly predicted (albeit weakly) adherence to a range of self-management behaviours (e.g., attendance; medication adherence; diet; physical activity; and other disease-specific behaviours (e.g., blood glucose self-monitoring)). Pooled effect sizes ranged from 0.04 and 0.13. None of the relationships varied according to acute and chronic illnesses; the type of self-management behaviour; or the duration of follow-up.
A qualitative study was subsequently undertaken. A variety of approaches were employed (including, expert consultation and Think-Aloud interviews) to develop a stroke-specific version of the IPQ-R. Interview findings showed very few problems with completion of the instrument. Where there were problems, these related to the wording of items, specifically abstract, complex and negative wording; the response format for the identity sub-scale; and questions on the timeline-cyclical and treatment control sub-scales. The revised scale was used in the third and final study.
The final study examined the relationship of illness beliefs, assessed by the Stroke IPQ-R developed in study 2, and measures of post-stroke recovery, including HRQL, disability and mood. This was a longitudinal observational study involving 50 survivors (average age=66.9 years, standard deviation =14.5 years) within three-months of their stroke. The primary outcome for this study, HRQL, was measured using the EQ-5D-5L instrument. The secondary outcomes were: disability, which was assessed using the Barthel Index, Modified Rankin Scale, and Nottingham Extended Activities of Daily Living Scale; and mood, measured using the Patient Health Questionnaire-9. Various statistical approaches were employed in the analyses of these data, the results of which are as follows.
Spearman’s correlations indicated that participants who perceived their stroke to have fluctuating effects and considerable distress at baseline (<eight-weeks post-stroke) also reported worse mood one-month later. Multiple mediation analyses were conducted to examine whether baseline mood and coping (follow-up medication adherence, measured using the Medication Adherence Report Scale) mediated any of these relationships. The results indicated that baseline mood rather than coping was a significant mediator in the prediction of worse mood three-months post-stroke. Multiple linear regression analyses were carried out to determine whether baseline illness belief domains explained the additional variance in three-month HRQL, disability and mood, over and above that explained by socio-demographic (e.g., age, gender, and deprivation) and clinical variables (e.g., baseline co-morbidities, stroke severity, and indices of recovery). The findings indicated that baseline illness belief domains significantly added between 7.4 and 29.9% to the overall variance in models for the majority of the abovementioned outcomes. Last, the results from the study demonstrated that baseline illness belief domains were highly significantly inter-correlated at baseline, and with follow-up illness belief domains. In light of this, cluster analysis was carried out to explore whether stroke survivors in the acute phase could be grouped according to their illness belief schema. Three clusters were identified: ‘Low Adjusters;’ ‘Moderate Adjusters;’ and ‘High Adjusters.’ In contrast to ‘High Adjusters,’ Low Adjusters’ were participants who perceived their stroke to be chronic, with fluctuating effects, associated with a lot of symptoms, serious consequences and considerable distress. While baseline cluster membership was not significantly related to follow-up markers of post-stroke recovery, there were trends to suggest that ‘Low Adjusters’ had worse HRQL and mood, and greater disability one-month later compared to ‘High Adjusters.’
The present research had several important implications. The principal theoretical implication concerns treatment beliefs. The findings from all three studies highlighted that people’s beliefs about the effectiveness of their treatment in managing their condition are examined in the CSM in a limited way. There is currently a substantial focus on medication-taking behaviours despite treatments often being more complex (e.g., lifestyle behaviour or surgery). Therefore, the findings indicated that there is scope to further elaborate this domain, above and beyond what has already been undertaken in the Necessity and Concerns Framework. This examines the cost versus benefits decision-making that patients undergo when taking their medication(s) (Horne and Weinman 1999). In addition, suitable instruments need to be further adapted and validated to accurately measure these treatment beliefs.
There were also several clinical implications of this work. First, it was emphasised that mood assessments should be carried out in patients immediately after stroke. It was shown in this research that even mild depressive symptomatology affected peoples’ beliefs about their stroke, and thus how well they psychologically adjust in the acute phase. Second, there is a potential to develop brief belief-based interventions for modifying maladaptive beliefs about stroke. For instance, addressing perceptions around the fluctuating effects of stroke and the considerable distress caused by stroke could form the basis for such interventions. In addition, interventions can be targeted to a particular type of stroke survivor (i.e., ‘Low Adjusters’), who are likely to be the people to benefit most from receiving them. However, these findings need to be borne out in a larger sample beforehand. The primary methodological consideration in relation to this work was the logistics of recruiting stroke survivors in the acute phase, which led to issues including: a small sample size; sample non-representativeness; and statistical constraints (e.g., factor analysis).
Actions (Archive Staff Only)
 |
Edit View |
Tools
Tools
![[img]](/style/images/fileicons/application_pdf.png)