Integrating exercise interventions into routine care for mental illness and cancer : An implementation science approach

Exercise has been established as an effective intervention that can improve health outcomes in people living with a non-communicable disease (NCD), including mental illness and cancer. Despite the evidence, exercise is not routinely integrated into the treatment of most NCDs. This phenomenon is described as the research-to-practice gap with
implementation science seeking to increase the uptake
of evidence-based interventions (EBIs), such as exercise, in practice. Multiple factors and processes contribute to the suboptimal use of EBIs in practice. This includes the lack of formative approaches to establish research evidence that is relevant for implementation and scientific methods that explain EBI adoption, implementation, and sustainment. This thesis applies an implementation science approach to explore how exercise EBIs can be integrated into routine healthcare to treat NCDs. This is explored in two NCDs, mental illness and cancer, and through the two following sub-aims:
• Conduct novel evidence synthesis for mental illness and cancer to understand how different synthesis methods can support improved implementation in practice.
• Explore how healthcare organisations have successfully implemented exercise EBIs within the routine practice for treating mental illness and cancer.

To address the first sub-aim, unique approaches were applied to two systematic reviews that were conducted in mental illness and cancer. For the first systematic review, a meta-review design was used to synthesise the evidence on the effectiveness of exercise EBIs for mental illness. effectiveness was defined in clinically useful terms including the anticipated health benefits, safety and cost of
exercise EBIs. Although positive effects on health outcomes (i.e., symptoms of mental illness, quality of life, and physical health outcomes) were reported in the majority of reviews, limited safety information and no cost data were identified.

For the second systematic review, efficacy studies were excluded to investigate the real-world
implementation outcomes of exercise EBIs for cancer care. Implementation outcomes were aligned with Proctor and colleagues’ Implementation Outcomes Framework (IOF), and the review revealed that the most common implementation outcomes assessed were adoption and feasibility. Penetration and sustainability were infrequently measured, and implementation fidelity was difficult to establish because exercise protocols were poorly reported.

In sum, the unique methods used in the two systematic reviews enabled the synthesis of broad and contextually relevant information valuable for implementation practice. The research gaps identified suggest that there is significant scope to produce more practice-relevant evidence.
To address some of these gaps and the second sub-aim of this thesis, two implementation studies that explored how healthcare organisations have successfully implemented exercise EBI’s in mental health and cancer were conducted. For both studies, a case study design and theoretically-informed approach were used to develop an explanation for the implementation process that included the identification of determinants, implementation strategies, and implementation outcomes. Four data sources informed the studies: semi-structured interviews, document review, observations, and
administering the Program Sustainability Assessment Tool (PSAT). Framework analysis was applied, and a theory-informed logic model was developed. Linking implementation science frameworks through the logic model elucidated the causal pathways of implementation. Second, the methods facilitated synthesis across sites to support generalisable knowledge.

The first implementation study evaluated an exercise EBI implemented within a youth mental healthcare service. Over 40 determinants that influenced implementation of exercise EBIs and a similar number of implementation strategies were identified. Several activities aided implementation, including the creation of a new clinical team and the auditing and provision of feedback on physical
healthcare practices (including exercise). Exercise acceptability was high, and many strengths (identified via the PSAT) contributed to EBI sustainability. However, implementation fidelity was challenging to establish, and penetration was low.

The second implementation study was a multiple case study on the implementation of exercise EBIs across three cancer care settings. Across the sites, 18 determinants and 22 implementation strategies were consistent. Sixteen determinants, 24 implementation strategies, and implementation outcomes differed across the sites. Via the commonalities, 11 common causal pathways were developed,
wherein the mechanisms theorised to support implementation include: 1) developing knowledge; 2) building skills and capability; 3) securing resources; 4) generating optimism and 5) simplified decision-making processes associated with exercise; 6) developing relationships (social and
professional) and support for the workforce; 7) reinforcing positive outcomes; 8) developing capability to action plan through evaluations and 9) interactive learning; 10) aligning goals between the organisation and the EBI; and, 11) establishing a consumer-responsive service. These mechanisms
represent transferable elements of the implementation process that can inform future implementation efforts.

This thesis uses implementation science to increase our understanding of the evidence, factors, strategies and processes required to implement exercise EBIs in practice. Improved implementation knowledge will help shape healthcare so people living with a NCD can access evidence-based care,
such as exercise.