The Community Assets Supporting Transitions (CAST) intervention

The CAST intervention was developed for older adults with multimorbidity and depressive symptoms. Building on a previous feasibility study [12], CAST was tested in 3 Ontario communities in Canada with the aim to improve the quality and experience of hospital to home transitions for older adults with multimorbidity and depressive symptoms.

The CAST intervention [28] consisted of an individualized patient- and caregiver-centred intervention for older adults with multimorbidity and depressive symptoms delivered by a community-based Registered Nurse acting as a Care Transition Coordinator (CTC) in addition to usual care. The CTCs participated in a standardized training program developed by the research team regarding intervention implementation. Clients of the CAST intervention (i.e., older adults with multimorbidity and depressive symptoms who were transitioning from hospital to home) received at least two home visits and a minimum of four telephone calls over the six-month intervention period. As part of the intervention, CTCs: (a) identified the health and social care professionals (e.g., family physician, specialists, home care providers, allied health professionals, community-based service providers) involved in the participant’s circle of care and initiated a coordinated care plan with these individuals and organizations, including sending alerts to the participant’s primary care provider regarding the presence of depressive symptoms, dementia, delirium, suicidal ideation, or medication issues; (b) identified and managed the patient’s risk factors for depressive symptoms and other chronic conditions in accordance with evidence-based guidelines [30, 31]; (c) provided system navigation support, coordination, and follow up for participants and caregivers [32]; (d) conducted medication reviews and management in collaboration with participants, primary care providers and pharmacists using evidence-based best practice guidelines [30, 31]; (e) conducted problem-solving therapy with participants and caregivers using an established guide [33]; (f) implemented social and behavioural activation by assisting and encouraging participants to participate in regular physical activity programs tailored to social and behavioural needs; and (g) provided education to patients and their caregiver (e.g., regarding their chronic conditions and self-management strategies). CAST was designed to complement other supports and services, including home care. As a result, some older adults received services from both home care and the CAST CTCs. As this was part of a pragmatic trial, the CAST intervention could be tailored to the needs and preferences of the clients receiving the CAST intervention. Clients could decide which elements, and how much of the intervention, they wanted.

Overview of the pragmatic trial

Study participants were recruited within local community hospitals by trained recruiters. The recruiters used a standardized script to approach potentially eligible patients and determine their interest in learning about the study. When patients agreed to be screened for study eligibility, the recruiter completed an eligibility questionnaire, after which individuals were informed if they were eligible to participate. To be eligible, individuals had to: (a) be aged 65 years or older; (b) be planned for discharge from hospital to the community; (c) have at least two self-reported chronic conditions; (d) have depressive symptoms as measured by the two-item Patient Health Questionnaire (PHQ-2) [34]; (e) reside in one of the three participating regions with no plans to move out of the region during the trial; (f) be competent to provide informed consent, as determined by a score of ≥ 5 on the Short Portable Mental Status Questionnaire (SPMSQ) [35] or have a substitute decision-maker who could provide consent on their behalf; and (g) be competent in English (or French in one study site) or have an available interpreter.

This trial was conducted in three sites (Hamilton, Burlington, Sudbury) within the catchment area of two regional health authorities (Hamilton Niagara Haldimand Brant and North East Local Health Integration Networks) in Ontario, Canada. These communities were diverse as related to geography (two sites were suburban/rural, one site urban), socioeconomic, language, and ethno-cultural characteristics. Each site had one full-time equivalent CTC administering the intervention. In one site, the CTC role was split between two part-time Registered Nurses.

This trial aimed to recruit 216 participants across all three sites, however enrollment was below target (N = 127) [28]. More details about this clinical trial and the communities in which it was conducted, as well as the findings from the effectiveness study, are provided elsewhere [28].

Collective impact

Given the complex nature of transitional care and the need for various partners and sectors to work together for hospital to home transitions to be successful, the Collaborative Intervention Planning Framework by Cabassa et al. [36] was used to enhance the potential for collective impact [37]. Collective impact involves a variety of key stakeholders coming together to address a complex issue by: establishing a shared vision of the issue; developing an accepted intervention to respond to the issue; agreeing on how the intervention will be evaluated; and establishing mechanisms for ongoing communication [37].

Prior to the initiation of the trial, the research team worked with key stakeholders in each community to organize and host a Community Forum. The Community Forums aimed to better understand the needs of each community in terms of transitional care, learn about existing transitional care initiatives, and gain insight into how the CAST intervention should be adapted to each community context. The research team worked collaboratively with the community stakeholders to determine who to invite to each community forum to reflect diverse perspectives. Attendees included older adults, family/friend caregivers, and representatives from hospitals, home care organizations, community support services, municipal governments, regional health care authorities, and other local organizations involved in care transitions.

Information obtained from the Community Forums was used to establish Community Advisory Boards (CABs) in each participating community. CAB members included: older adults, family/friend caregivers and members of the public (referred to as patient and public research partners); representatives from hospitals, home care organizations and community support services involved in hospital to home transitions; and members of the research team. The role of the CABs was to guide the local implementation of CAST in each community. This involved determining how CAST may need to be adapted to suit local conditions, keeping the research team abreast of relevant activities happening in the community, and problem-solving around implementation challenges experienced by the study (e.g., challenges with recruitment) and the CTCs (e.g., support to identify local resources that the CTC could refer to address specific client needs). The CABs met six times during the course of the study. Some CAB members also contributed to the interpretation of study findings and developing key messages [38].

Study design

We used a qualitative descriptive study design [39] to identify facilitators and barriers to the implementation of the CAST intervention. We used the Standards for Reporting Qualitative Research (SRQR) to report study methods and results [40]. Data collection and analysis were guided by the Consolidated Framework for Implementation Research (CFIR) [41]. CFIR is a commonly used framework for examining health service implementation using evidence-based factors associated with successful implementation of interventions across a wide range of settings [42]. This framework consists of 39 constructs organized into five main domains known to affect implementation of interventions: (a) intervention characteristics, (e.g., adaptability, complexity, and design quality); (b) outer setting, which includes external factors (e.g., political, social, and economic); (c) inner setting (e.g., organizational factors such as structures, communication, and readiness for implementation); (d) characteristics of individuals involved in the intervention including their knowledge and beliefs about the intervention, stage of change, and personal attributes; and, (e) implementation process including planning, engagement of key players, and how the intervention is executed [41]. Use of the CFIR constructs can inform an understanding of why the implementation of an intervention was successful or unsuccessful [43].

Participants

This study focuses on the implementation of CAST with older adults who received the intervention. Study participants were members of the three local CABs (a total of 35 individuals) and four CTCs. All CAB members and all CTCs were invited to participate in the study. There were no exclusion criteria. Client and caregiver perspectives on the implementation of CAST were also sought through post-intervention interviews. However, very few clients and caregivers agreed to take part in these interviews. As a result, these data were not included in the analysis.

Data collection

Data analysis

All anonymized interview transcripts and meeting minutes were uploaded to NVivo 12 [45] for data management and analysis. We used qualitative content analysis, consistent with a qualitative descriptive approach using deductive coding with the CFIR framework and inductive analysis for sub-codes within the CFIR domains [39]. This analytical approach allowed us to develop an understanding of the factors that served as facilitators and barriers to the implementation of the CAST intervention.

An RA, not involved with data collection, independently analyzed all interview transcripts line-by-line applying a priori codes of the CFIR domains and constructs. In addition, lower level nodes were developed inductively from the data and then integrated into the CFIR domains framework, as appropriate. Interview responses were coded across all CFIR domains. In addition to the RA, one principal investigator (CM) independently coded 20% of the interview transcripts to ensure inter-rater reliability in coding and then reviewed all subsequent coding with the RA to further categorize or collapse second level codes related to the CFIR constructs. Other authors (MMR, RG) also reviewed the coding within each of the CFIR domains. The research team met to review the coding to identify relationships between the codes and, where relevant, further combine or collapse codes, discuss emerging ideas, and identify discrepancies; adjustments to coding were made through consensus. The RA and lead author (CM) met to review all codes to identify patterns among first, second and third level codes [46], which were then merged into categories representing implementation factors. Themes and subthemes were then identified and reviewed by the research team, who agreed on data interpretation by consensus. The RA reviewed the data to select exemplar quotes; modifications and explanations within quotes are presented within square brackets to increase comprehensibility. Each quote has been identified as coming from CTC interviews or CAB focus groups in square brackets following the quote.

A similar approach for the analysis of study documents was used. Study documents were reviewed and analyzed by three authors (CM, MMR, RG). Data from the document analysis were triangulated with findings from the interview analysis.

We ensured study rigour with an audit trail of all procedures and decisions related to recruitment, data collection, and analysis. Interview transcripts and documents were reviewed independently to reduce potential for selection bias regarding positive responses and increase attention to negative responses. All authors were involved in providing feedback on the analysis and interpretation of findings.

Ethical considerations

We obtained institutional ethics approval for this study from the Hamilton Integrated Research Ethics Board, McMaster University (REB#2586), the Office of Research Ethics at the University of Waterloo (#40867), the Laurentian University Research Ethics Board (REB#6009840), the Research Ethics Boards from Health Sciences North (REB # 17–007), and Joseph Brant Hospital (REB #000-039-17). All interview and focus group participants provided written informed consent prior to data collection.

Participants characteristics

We conducted individual interviews with four CTCs across the three sites. These interviews were on average 35 minutes in length. All the CTCs were female, and were Registered Nurses with baccalaureate degrees; one CTC had a master’s degree. The CTCs had experience working in both acute care and community settings, with 2–20 years of clinical experience between them.

Two CAB focus group interviews were conducted at each site. At Time 1, 13 CAB members participated across the three sites; at Time 2, 12 members participated across the three sites. All but 1 focus group participant was female. In both Sites A and C, a CAB member was unable to attend one of the focus group interviews, but was interviewed individually at a subsequent time. The focus groups and interviews ranged from 35 to 87 minutes in length, with an average of 53 minutes. In Site A, 5 CAB members participated in the focus group/interview at Time 1 (2 community partners and 3 patient and public research partners) and 4 participated at Time 2 (2 community partners and 2 patient and public research partners). In Site B a total of 7 CAB members participated in the focus group at Time 1 (4 community partners and 3 patient and public research partners) and 3 participated at Time 2 (1 community partner and 2 patient and public research partners). In Site C, 1 CAB member (a community provider) participated in an interview at Time 1 and 5 (4 community partners and 1 patient and public research partner) participated in the focus group at Time 2 (S4 Appendix).

Facilitating factors

The qualitative analysis of interview transcripts and program documents generated five themes that describe factors which facilitated the implementation of the CAST intervention. The themes were: (1) engaging the community to gain buy-in and adapt CAST to the local community context; (2) planning, training and research meetings; (3) engagement, relationship building and collaboration with local partners; (4) availability and provision of support, information and resources to address CTC needs; and (5) tailoring the intervention to individual client needs and preferences. The CFIR domains and constructs that contributed to each theme are described.

Introducing CAST and obtaining local buy-in

The document analysis revealed the sub-theme–introducing CAST and obtaining local buy-in. Within each community, attempts were made to understand the local context in terms of existing services, the priority of the intervention, and alignment of CAST within participating hospitals’ strategic plans, which supported community buy-in. Understanding the local context began with holding a Community Forum in each site prior to study initiation.

The forums enabled researchers to learn more about community resources, including existing programs and services that supported the transition of older adults from hospital to home. The Community Forums also provided the community with an opportunity to learn about the CAST intervention implementation and evaluation plans. Input from the Community Forums helped to identify members for the local CABs that were subsequently established to support local implementation of CAST. Together, the Community Forums and CABs provided the research team with information on the unique characteristics and needs of each community. Meeting notes also highlighted the variability of services available in the different study communities, with one community having fewer community services available to clients following hospital discharge than were available in the other communities. Meeting notes reflected sensitivity in understanding how the CAST intervention, as a new service in the region, implemented by an external research group, would be perceived.

Adapting the implementation of the intervention to the local context ensured that CAST complemented existing services

Understanding the diverse characteristics and needs of the local context and adapting the implementation of the intervention, helped to ensure that the intervention complemented and did not duplicate or interfere with existing services. A CAB member described the importance of adapting the implementation to the local context to help with service efficiency and reducing duplication:

“Groups will be able to share all of the processes, and because of that, it will improve the efficiency of the work that has to be done, because the more resources, the more that can be shared. And, the first thing they’ll have to do is look at where there’s duplication of services, and take out that duplication and focus on the priority things that need attention, through all of the resources that each of the groups will have.” [CAB Site C]

Given that other services in each community also help to support older adults transitioning from hospital, there was a potential that services could be seen as competing. Adapting the implementation for each community helped to prevent this. One CAB member highlighted the significant needs of this population and that there were no feelings of competition in their community as a result of working together.

“We can divide and conquer and support people because there’s a lot of vulnerable people in our city, and it takes a village, right? So, I don’t feel any territorial things at all.” [CAB Site C]

Meeting notes also revealed that there was discussion about adapting the CAST intervention to meet local needs within each of the three communities. Fewer adaptations of the intervention were made in one community which did not have a vast array of available community services. One concern raised during the Community Forums was about the transition back to usual care following receipt of the CAST intervention. To address this, CAST was adapted to enhance the discharge process at the end of the 6-month intervention period. A clear plan was developed and documented, outlining which organization(s) participants would be referred to (where applicable) after completing the intervention.

Training of the CTCs

Notes from research team meetings described the development of a comprehensive, standardized training manual. The manual included: (a) information on the study population; (b) resources to support the CTCs in implementing each component of the intervention (e.g., clinical assessment tools, medication reconciliation tools, problem-solving therapy); (c) research articles and presentation notes to support both the rationale for, and implementation of the intervention components; (d) documents to use when communicating with other providers (e.g., family physicians); and (e) tools to document CTC activities. To further support the CTCs, the CTC who was the first to begin working in the role supplemented the training manual with additional resources that she found or created to assist her in working with clients and caregivers.

Training was provided over a 2-day period, led by the study principal investigators and supported by the study’s research coordinator and a CTC (except when the first CTC was trained). As part of the training, CTCs engaged in discussions about the intervention rationale and components, watched training videos, and had opportunities for role play. Training notes revealed that CTCs valued time spent discussing their nursing role, specifically related to their role in case management, care coordination and system navigation support, and their experience developing interorganizational relationships.

Ongoing and timely access to the research team by the CTCs

Meeting notes indicated that the regular research team meetings with the CTCs while they implemented their role served as an opportunity to: (a) address questions of a clinical nature (e.g., clinical supervision of CTCs, management of clients that became emotionally distressed during visits, management of suspected elder abuse); (b) reinforce and address questions related to implementation of the study (e.g., participant recruitment, documentation, ethics/confidentiality); and (c) share success stories related to positive clinical outcomes and effective collaborative efforts with community partners. These meetings also served to discuss strategies to address clinical issues within the context of local resources and supports for clients available within each site. (e.g., how to meet client needs for specific supports, such as physiotherapy or exercise programs, given limitations in availability or eligibility criteria for these services). The CTCs valued prompt access to the principal investigators via phone, email and in-person to address implementation related questions. The following two quotes highlight the support provided to the CTCs:

“So we’ve had regular CTC meetings with the study [principal investigators] over the course of the intervention to touch base on what’s working well and what may be not working so well, as well as study [principal investigators] and research coordinators have been able to be reached by phone or email or in-person at the office as needed for any further direction.” [CTC Site C]

“Well I think, having the study [principal investigators] present [for training] was positive for sure because I had the opportunity to digest the information, come back the second day and ask some follow up questions. Since then things have been evolving a bit and there’s been some, ongoing questions and not retraining but follow-up for sure. … And having the [principal investigator] in the office available in person or by phone for specific concerns with intervention visits, I have to bounce off someone and clinical guidance for any unexpected situations so that’s been good to have that.” [CTC Site B]

Local champions supported and promoted the implementation of the intervention

In the process of developing and implementing CAST, some individuals representing hospitals and community services (e.g., senior nursing leaders) with whom the CTCs initially met, became effective champions of the intervention. These champions supported the intervention within their organizations, which served to legitimize the work of the hospital-based recruiters as well as the CTCs. Meeting notes reflected that, within hospitals, intervention champions promoted the study, helped identify eligible clients, and facilitated recruitment by increasing access to various hospital units. One CAB member described the value of having the support of a vice president at the hospital to the implementation of the intervention:

“And I think [Hospital Vice President] too, having someone who’s high-up in the hospital, she was also a really strong advocate for this… to me that was something that was very beneficial from my perspective.” [CAB Site C]

Similarly, in the community, organizational leaders who were champions of the CAST intervention provided staff members in their organization with information about the CAST intervention and advocated for the importance of assessing referrals from the CTC. An example of a community-based champion was highlighted by a CAB member:

“Having [the partner] go in and talk to people and say “This is not just another box for you to check every day, another job for you to do”, this is a really important thing for the community.” [CAB Site C]

Document analysis revealed that some champions also helped identify and connect the CTCs with leadership from other services to further support implementation.

Established relationships with health and social services facilitated information sharing and collaboration for client care

Document analysis revealed that early in the development of the intervention, CTCs scheduled meetings with program leaders (such as managers from home care) to explain the intervention and explore synergies in service delivery and opportunities for collaboration. These meetings resulted in establishing working relationships that optimized the CTCs’ clinical care of clients through enhanced information sharing across services, expedited referrals to community services, and enabled collaborative problem-solving to match services to specific client needs. One CTC provided an example of how she worked collaboratively with another provider to meet the goals of a client:

“I’ve made a few calls to care coordinators, saying you know there was this concern in the home but this is their goal, how can we—can you think of anything I can do in the home to help them meet this goal safely and please let me know. And I receive a call back saying let’s try to modify it this way and I said okay great, I’ll provide that same encouragement so everyone’s on the same page.” [CTC Sites C]

Having strong relationships with other providers also helped the CTCs understand the work realities of other providers. This CTC acknowledged the limited time available to some home care providers and described the appreciation that home care providers expressed in being able to work together to meet the client’s needs:

“So [the care coordinators from Home and Community Care are] spread very thin. So going in to reassess, they don’t have the opportunity to do that as often in some cases and … perhaps [see] other needs that are … changing over time … the care co-ordinators that I’ve spoken to, they appreciate [me] at least connecting with them and sending my notes, providing these updates … because of their caseloads and their time limitations…” [CTC Site B]

Another CTC explained working with a physician to involve another health care professional to support the care needs of a client:

“We worked with a physician to get a social worker involved with my case. That we were trying to get [client] into a complex continuing care facility…. And the nurse that was working with me and the social worker that was working with me were all on board to get her there.” [CTC Site A]

Provision of resources based on CTC needs for information

The review of meeting documents revealed that CTC need for information and supports were regularly assessed in meetings and resources were developed or existing resources were tailored to address these needs. These resources included consent forms to allow for communication within the circle of care, information on problem-solving therapy, behavioural change support, system navigation, behavioural and discharge planning to facilitate linkages and transitions across services at the end of CAST program, and creating hard copies of client and caregiver resources for education. The CTCs also described the importance of having access to information on available community services in terms of what services were provided, eligibility criteria, and referral forms and processes for access. Two CTCs explained the value of this information:

“So the community resources were available to me. All the different community resources. All the forms were available to me. All the supporting data on how to do it was available to me.” [CTC Site A]

“Getting information on day programs too was, was important just to understand how those work..” [CTC Site B]

Development of a CTC community of practice

To provide additional support to the CTCs, the research team established a CTC CoP, that included the CTCs from each of the sites and the study research coordinator. The principal investigators attended the initial meetings of the CoP and later attended when invited by the CTCs. Document analysis revealed that through the CoP, CTCs discussed questions and issues related to intervention implementation, and shared effective strategies for managing these issues, and brainstormed solutions for common problems. When new CTCs were recruited, established CTCs offered to provide training and mentorship support, for example, attending home visits together (job shadowing) and offering key insights gained through implementation of their role. One CTC described the value of this peer-to-peer support:

“My colleague [CTC in another community] is available to help discuss cases, work through how to implement components of the intervention. We talk about clinical questions with each other… So the most supportive thing that works for me is to be able to discuss individual patients at an individual level with [CTC].” [CTC Site C]

CTCs engaging with providers in the client’s circle of care

Care coordination was challenged by the fact that the CTCs lacked information about who was in the clients’ circle of care. Many clients, while aware that services were provided in their home, did not always know who these care providers were or what agency they were from. A CTC described this challenge:

“The other issue that comes up often is determining the circle of care; and then what, who is the most important to communicate with in the community, so that you can appreciate a lot of people don’t fully appreciate who’s in their circle of care or who is coming or where they are from; so they sometimes mix up organizations so that’s been a challenge too.” [CTC Site B]

CTCs reported spending a lot of time identifying service providers involved in clients’ care and experienced difficulty connecting with them. The CTCs did not have access to the clients’ electronic medical record so they were unable to identify community providers involved in their care through this administrative mechanism. Further, there was no shared information system for all of the providers and agencies involved, which further challenged care coordination. A CTC explained the multiple steps involved in trying to identify the circle of care, and the added complication of not having access to a shared information system to assist with this work:

“I think it’s quite complicated … trying to connect with the circle of care at times. Figuring out who the LHIN care coordinator is, connecting with the team, faxing them, calling them back to make sure that they did get the fax, and the other notes. So unfortunately there was no streamline[d] set of communication process, not having access to clinical connect [a web-based portal for sharing health information within the health care system] [CTC Site C]

Engaging family physicians

CTCs noted that they had minimal interactions with clients’ family physicians despite attempts to connect with them via phone and fax. Communications were described as mostly one-way, with CTCs sharing their concerns and observations about clients, but receiving few responses from family physicians. A CTC explained her experience:

“I didn’t do a lot of connections [with family physicians]. It seemed mostly by fax. And I have done follow-up calls … mostly alerting … maybe some medication concerns here or … their goal for maybe a referral to a program, alerting them of any depressive symptoms. We did get some notes back and communications back but it was mostly one way.” [CTC Site C]

It was also documented in the meeting notes that CTCs reported difficulty getting information from family physicians on the client’s medical history to support their work (e.g., information on diagnoses and current medications). It was suggested that this may be the case, in part, because family physicians did not know whether they could trust the CTC as they did not have established working relationships or because clients have so many home care services that physicians were not sure who the CTCs were or with which service they were associated.

CTCs described the importance of being able to engage with family physicians and the need to have them actively involved. For example:

“Well we have to get more physicians, the family physicians, to be on board. And be open to receiving that feedback that we’re seeing in the community. The primary care providers of patients…They were not receptive to feedback.” [CTC Site A]

Strengths and limitations

A strength of the study was use of the CFIR framework to guide the development of data collection tools as well as the analysis. Using CFIR to guide the study enabled us to build on the existing knowledge about implementation science to explore determinants or factors at multiple levels of influence on implementation of a transitional care intervention for older adults with multimorbidity and depressive symptoms. However, one challenge with using CFIR was that it was developed to examine the implementation of interventions by individual organizations. As CAST was a community-based intervention, the data collection tools needed to be revised to reflect the community-level focus of the intervention. Another strength of the study was that it was conducted in three diverse geographic regions, enabling the opportunity to explore factors that helped or hindered implementation of the intervention in multiple settings. In terms of limitations, while data collected from the CABs was a source of data in the study, we did not obtain information from a broader range of community members regarding the implementation of CAST. As well, since CAB members were ‘partners’ in the project, we did not collect information on their demographic characteristics, limiting our ability to describe these groups. Finally, while attempts were made to gather information from clients and caregivers regarding the implementation of CAST, because of the complex health conditions of the clients (and the impact of this on caregivers) very few agreed to take part in this component of the study. As a result, this study is missing this important perspective.