Abstract
Background: A rise in life expectancy has increased the likelihood of complex co-morbidities and disability resulting in a change in the make-up of hospitalised older adults and their rehabilitative requirements, including physiotherapy. Studies to date of the impact of physiotherapy on older adults admitted to hospital have investigated the intervention in isolation from context and have resulted in inconsistent outcomes. There remains a lack of evidence and theory regarding why physiotherapy is effective in some instances and not in others. Purpose: The thesis aims to examine the mechanisms might or might not work for patients, the influence of context on the actions of stakeholders to impact on outcomes and finally the appropriate methodology with which to investigate physiotherapy intervention. Method: Realistic evaluation (RE) was used as a framework for the study. Initial theories about the impact of physiotherapy on medically unwell older adults were expressed terms of context-mechanism-outcomes configurations and developed into hypotheses with stakeholders. Testing of hypotheses utilised a sequential mixed methods design including quantitative and qualitative measures, obtained from a convenience sample of older medically unwell patients (n=75) and a purposive sample of physiotherapists (n=6), patients and carers (n=8) respectively. Outcomes included measures of physiotherapy intervention, patient health status and hospital performance. Descriptive data analysis with subgroup comparison (frail/non-frail or cognitive impairment/cognitive impairment) was undertaken for outcome measures using Mann-Whitney U and Spearman’s correlation co-efficient. Qualitative data were analysed using framework analysis. Quantitative and qualitative study findings were reviewed, integrated and context-mechanism-outcome threads identified. Initial theories were refined in light of study outcomes. Findings: The sample mean age was 84.8 years (SD±7.1), 73% of individuals were identified as frail and 52% had cognitive impairment. Median hospital length of stay was 14 days (9-26). Median (IQR) physiotherapy interventions were 5 (3-10), total amount of physiotherapy 3.5 hours (1.8-7.21 hours) and frequency of physiotherapy intervention was 0.4 interventions/day (0.3-0.5). Subgroup analysis revealed patients had no between group differences in physiotherapy interventions. Moderate to large associations relating to a positive effect of physiotherapy on change in health status outcomes were present in unimpaired subgroups only. Physiotherapist and patient accounts indicated that organisational commitment to physiotherapy service provision, patient presentations of frailty and cognitive impairment and the priority of rehabilitation activities in the clinical ward setting triggered specific actions in frontline staff and patients. Patient and staff actions influenced by contextual factors affected the dose of physiotherapy and independent patient activity on the ward and contributed to explanation of patient health status outcomes in this study. Conclusion: The use of RE has resulted in the emergence of important elements related to context and stakeholder actions, which act as barriers or facilitators to physiotherapy implementation and effects of physiotherapy on medically unwell older adults admitted to hospital. Further testing of revised theories is required to enhance understanding of the impact of physiotherapy in terms of what works or doesn’t work for older adults. However, the in-depth knowledge revealed in this study has utility for informing models of physiotherapy and rehabilitation provision in the acute setting to best meet the needs of patients.