A qualitative descriptive study involving semi-structured telephone interviews and focus groups was undertaken as part of pilot randomized controlled trial (6). The trial methodology is described in detail elsewhere (6). We used the consolidated criteria for reporting qualitative research (33) (COREQ) checklist to guide reporting. The research ethics board at the University of Toronto and at each participating hospital site approved the study protocol.

Table 1 describes stakeholder qualifications and responsibilities. Across the three study sites, 5 recreation managers, 3 healthcare managers, 8 fitness instructors, 3 healthcare partners, and 3 volunteers (1 site did not employ volunteers) were involved with TIME™ program implementation during the study and available for recruitment. On completion of the 12-week TIME™ program in the experimental group at a site, the central study coordinator contacted each stakeholder by email and invited them to participate.

The TIME™ program offers two one-hour exercise classes per week for 12 weeks. Group classes include a seated warm-up, repetitive and progressive practice of functionally relevant balance and mobility tasks organized in a 3-station circuit, and a seated cool-down. An instructor-to-participant ratio of 1-to-4 is required to optimize supervision and safety; volunteers can be used to supplement instructors to attain this ratio. Healthcare partners visited the first two classes and three additional classes interspersed across the remaining 22 classes (6).

In the trial, community centres first signed the TIME™ license, and then received an electronic toolkit with materials (e.g., liability waiver, exercise guidelines, equipment list, etc.) required to implement the program. Fitness instructors, healthcare partners, and volunteers completed training which involved a combination of e-modules, and in-person and virtual sessions (6).

We conducted interviews with managers, and focus groups, consisting of two to three participants, with fitness instructors or volunteers. A Toronto-based, female, interviewer with six years of experience with conducting qualitative research obtained informed consent and conducted focus groups and interviews. Interview or focus group guides were tailored to each stakeholder group. We selected the CFIR to inform interview guides for managers as this framework incorporates factors at the organizational level relevant to managers’ experiences and decision-making. We selected the TDF to inform the interview and focus group guides tailored for program providers as this framework emphasizes factors relevant to individual experiences and decision-making. Across sites, interviews and focus groups were held up to 8 months post-intervention and lasted 45–60 min. The interviewer used probing questions to obtain greater detail or explanation. Sessions were conducted by telephone to allow for flexibility given geographical distance. Sessions were audio recorded and professionally transcribed.

To inform the description of costs, we documented the time required to complete the activities involved with training and delivery of the first program. We also documented the hourly wages of the staff involved in program delivery and the cost of the equipment that was purchased for the sites. We requested fitness instructors to document the travel modes used by participants to attend the class using a standardized form, while healthcare partners were requested to document the time spent on class visits.

A directed content analysis that incorporated the CFIR and TDF was undertaken. We developed a codebook composed of CFIR and TDF domains. Two researchers (NMS and KB) used an inductive and deductive approach to independently open-code one transcript per stakeholder group (five in total). Researchers then met to compare and contrast codes to ensure consistency and to revise the coding scheme before KB proceeded to independently code the remaining transcripts using NVivo 10 software. Subsequently, for each CFIR and TDF domain, KB reviewed coded text from all stakeholder groups first within each site and then compared and contrasted coded text across sites. NMS, GA, and KB met and reviewed within-site node summaries to identify emerging categories as they related to facilitators, barriers, and recommendations for the process of first-time implementation of the exercise program. After recognizing that experiences were similar across sites, KB combined site-specific data to develop summaries of facilitators and barriers within each CFIR and TDF domain, and recommendations, with supporting quotes. Through a series of meetings, NMS, GA, and KB met and refined summaries of the facilitators and barriers to ensure alignment with CFIR and TDF domains. An inductive approach was then used to identify overarching themes across the data through discussion and consensus. Similarly, the recommendations and supporting quotes were reviewed and classified by the activity within the implementation process. To maintain records and encourage reflexivity about data collection procedures, an audit trail was used. The interviewer took reflexive notes after completing each session related to similarities across interviews and focus groups and new ideas arising from the data. Preliminary results were presented and discussed with the interdisciplinary research team to consider alternative interpretations of the data. Throughout the analysis, discrepancies around emerging ideas, codes, and themes were resolved by discussion and through reference to the original transcripts to ensure authors were in agreement. Lastly, we presented findings with direct quotations from the participants as exemplary anecdotes that support the findings of the study.

We developed a sample budget that presents personnel and travel costs for the healthcare partner, recreation partner and the participant to run a 12-week TIME™ program. For healthcare and recreation partners, we distinguished costs for planning and implementation activities. We took the mid-point value of pay rates observed across the three sites during the study to generate cost estimates; otherwise we used and indicated an alternate source.

Table 3 presents recommendations made by participants on how to improve the process of implementating the TIME™ program. Recommendations related to engaging healthcare and recreation organizations, selecting staff and volunteers, training, healthcare partner visits and referral, intake, and class instruction.

Table S1 in Supplementary file S1 presents a sample budget of estimated personnel and travel costs for the planning and first-time implementation of a 12-week TIME™ program (two classes per week) for a group of 12 participants with two fitness instructors and one volunteer per class. For the healthcare centre, the estimated total cost of the healthcare partner, including selection, training, class visits, and consultations, is $680 CAD. For the recreation centre, the estimated total cost of running one program, including purchasing the TIME™ license, and selection, training, and use of fitness instructors, is $3,153 CAD. Therefore, the sum total of healthcare and recreation costs was $3,833 CAD. The transportation cost for participants to attend the entire program is $283 CAD based on mileage for using a personal vehicle, and $110 CAD for taking public transit.

Recreation and healthcare managers reported similar priorities and organizational considerations that influenced their decision to implement the program. The reputation of program developers, scientific evidence, successful implementation in other centres, interaction with program experts, and alignment with organizational mandates, were influential. This suggests that even when implementing an evidence-based program such as the TIME™ program, management places importance on how it is implemented, who is involved with the implementation, and program compatibility with the characteristics of the organization and the community. Determining who presents the program is important as someone with comprehensive knowledge of the program can highlight key advantages of the program and address any concerns. When organizational leaders have confidence in the program and display a readiness for change, it is reflected in the program staff's attitudes and involvement (34) which in turn influences the participants' experiences with the program (35).

Recreation and hospital managers confirmed that organizational capacity to bear program costs, a commonly reported barrier to program implementation (17, 29, 36, 37), impacted their decision to deliver the program. Findings from this study add new information on the timing and extent of costs. When implementing a licensed CBEP-HCP, an established community recreation centre will encounter costs related to the license and instructor training and program delivery. If the recreation centre employs the healthcare partner, then additional costs would include healthcare partner training and visits. The healthcare partner costs steadily decline over time as instructors gain experience and skill, requiring fewer sessions of supervision/consultation (1). Training the instructors is a single time cost, but additional instructor training costs may arise if refreshers are needed or if new instructors require training because of program scale-up or staff turnover (1). The TIME™ program utilizes equipment that is commonly available in most recreation centres (38).

Community organizations can establish cost-recovery mechanisms to offset the cost of implementing and delivering the program. This includes program fees, membership fees, donations or applying for and obtaining funding to support program activities (1, 15, 29). Program duration may vary across sites, and can be further influenced by parameters such as use of volunteers or regional cost of living (1). Using the cost analyses presented in this study, community organizations can estimate the anticipated costs at the time of implementation based on their program design. For healthcare systems, supporting CBEP-HCPs through referrals and providing a healthcare partner has the potential for cost-savings in the long-term. The availability of local CBEP-HCPs may help to expedite discharge from outpatient services and decrease costs by reducing the duration of outpatient care. Participation of people post-stroke in community exercise programs may be associated with significant cost savings to the healthcare system resulting from the reduced use of healthcare services (39–41). Hospital managers and healthcare partners can consider these factors to justify the allocation of time and personnel resources towards such programs. The public transportation cost for participants is sufficiently high that it may pose a barrier to attending the program. Recreation centres should consider transportation cost and ability to pay when offering community exercise programs for people with stroke and provide subsidies or identify low-cost alternative transportation options in their local communities.

In this study, healthcare and recreation managers placed value on inner setting variables such as organizational readiness to implement the program, and the potential growth in organizational impact on the community brought on by program implementation. Since the program could be scheduled during “off-peak times”, it was not challenging for the recreation manager to allocate space, equipment, and instructors to the program. Interestingly, despite their mandate to include accessible programming for all community residents, it was the implementation of the TIME™ program that allowed the recreation organization to serve residents who experience balance and mobility limitations resulting from a stroke. This intersects with hospital managers' identified need for programs in the community that support people with stroke as they transition from the hospital to home. The partnerships created through the process of implementation were nurtured through the complementary mandates and a shared culture that values innovation and inclusivity. While some of these factors have been identified in previous studies on the implementation of CBEP-HCPs (1, 16), this is one of the first studies to present managerial perspectives. Understanding managers’ perspectives is critical because, in addition to overseeing allocation of staff and resources, managers are key decision-makers about program implementation, delivery and long-term sustainability (29).

Program staff were aware of the influence of their skills, knowledge, and beliefs on the delivery and impact of the program. In the absence of previous experience working with individuals with disabilities, program staff described a lack of confidence in managing people of varying abilities. At the same time, their beliefs about the positive consequences of participating in the program for themselves (obtaining/expanding skills) and the participants (affecting a positive change in their health) motivated them to remain involved. Recommendations to improve support included adapting the training material to cater to inexperienced instructors and discussing specific challenges, such as lack of resources, participant safety, and delivering a class for people with varied ability levels. Importantly, program providers recognized having previous experience working with individuals with disabilities and the eagerness to work with them as important attributes in a TIME™ fitness instructor.

Study findings reinforce the role of the manager in identifying staff who fit the profile of a suitable instructor. This would include someone with experience and eagerness in working with individuals with mobility issues, ability to troubleshoot and innovate on-the-go, or the willingness to learn these skills. To assist instructors with program delivery, the healthcare partner role would require an individual who is able to empower and work with the instructor to adapt the program to the needs and abilities of the participants. When program staff can internalize the benefits of associating with the program and feel supported through the challenges, turnover may decline (42). This is important because participants rely on consistency and trust developed over time for their continued participation in community-based programs (43).

The factors identified as facilitators and barriers to first-time program implementation in the current study also impact program continuity and sustainability. In a study examining factors influencing sustainability of CBEP-HCPs (29), CBEP-HCPs that discontinued had common features including inexperienced instructors. Sites that sustained the CBEP-HCP for four or more years had common features including stable funding models, experienced and motivated instructors, supportive management, dedicated resources, support of a local champion and a strong network of local partners (29). Planning for these determinants at the time of implementation can set the stage for program sustainability.