Participant characteristics
In total, exceeding the team’s anticipated target, 75 people participated in the study, 42 in one-to-one interviews and 33 in focus group interviews. Participants included EOLAS co-ordinators (n=16), EOLAS facilitators (clinical n=12; peer n=25), programme participants (n=16) and other key stakeholders (n=6). The stakeholders consisted of EOLAS Steering Committee members (n=3) and project workers (n=3) who had been employed to support the development and roll out of the programme. A breakdown of the profile of interviewees is given in table 2.
Insert Table 2: Overview of interviewees by method of data collection and role in EOLAS
Enablers and Barriers of Implementation based on CFIR
Participants described a highly complex range of factors that mapped to the five CFIR domains. While some groups of participants spoke in greater depth to a specific domain, all groups identified issues across all five domains, with the exception of the programme participant group who mainly contributed to the ‘intervention characteristics’ domain. In addition, some domains yielded a greater number of coded units (pieces of data) in comparison to others, with the breakdown between barriers and enablers also differing. For example, the implementation process domain yielded 519 code units whereas only 180 pieces of data were coded to the outer setting. Similarly, there was variation in numbers when data were coded as an enabler or a barrier. With regard to the outer setting there were 115 coded references (units) made to barriers and 65 made to facilitators, with many comments referring to the same construct or idea. Figure 1 provides a breakdown of the number of enabler and barrier coded units per domain.
Insert Figure 1: Number of enabler and barrier codes per domain
Analysis of the data identified enablers and barriers across all five CFIR domains, with some of the factors identified as being both an enabler and a barrier depending on context. Figure 2 provides an overview of the findings per domain and Additional File 6 provides data to support findings described in each domain.
Insert Figure 2: Summary of barriers and enablers
Domain A: Outer Setting
The Outer Setting focuses on the socio-cultural and infrastructural context in which an organisation resides [35]. Within this domain four key factors were identified; policy, instability intrinsic to the statutory agency, payment, and infrastructure.
In terms of policy there were mixed views, with some participants expressing the view that EOLAS’ compatibility with the national recovery agenda, and the fact that it was referenced in national mental health documents were an endorsement which gave the programme legitimacy and facilitated its promotion within local services (A1-A2). In addition, some contributors viewed the independence of EOLAS as critical to its success, being developed and co-ordinated by a steering group independent of the Health Service Executive (HSE), and ensuring ongoing fidelity to the EOLAS model (e.g. in training, co-production and co-delivery). This group expressed concern that full assimilation of EOLAS into HSE structures would result in a loss of fidelity to its ethos (A3-A4). On the other hand, other contributors commented on the fact that the external positioning of EOLAS potentially led some stakeholders to consider it to be an ‘addendum’ programme and made the EOLAS Programmes more vulnerable if choices were to be made regarding funding (A5-A8).
The second factor identified as a barrier to implementation related to the structural and personnel instability intrinsic to the HSE who were the funders of the programme. Senior personnel within the HSE who were key to supporting the implementation transitioned out of roles and as a consequence the EOLAS steering group were constantly building and renewing relational ties with funding decision makers and educating them about the unique features and benefits of the programme (A9-A10).
The third factor identified was the payment of peer facilitators. All participants agreed that payment of peers was an important enabler as it valued their contribution and ensured they were not out of pocket; however, they were adamant that the model of payment operated by the health service for peer delivered programmes (including EOLAS) was a significant barrier. In their view it was causing significant distress to some service users and family peer facilitators, due to its unwieldy nature, and as a consequence had adversely affected social welfare payments, (including cuts to social welfare and other entitlements or incorrect and untimely payments) and added to the workload of local co-ordinators (A11-A14).
The fourth factor impacting implementation related to the infrastructural context in which the mental health service operates. The lack of public transport, particularly in rural areas, was identified as a barrier to people accessing EOLAS (A15-A16). Clinical facilitators and coordinators emphasised the importance of taking people’s transport needs and their fears of going out in the evenings into account when arranging the location and timing of the programmes (A17-A19).
Domain B: Inner Setting
Inner Setting is defined as the internal socio-cultural context of the organization in which an innovation is being implemented (e.g. cultural, leadership, values, innovation climate, organisational capacity) [35]. Within this domain the following issues were identified: culture, implementation climate and readiness for implementation.
From a cultural perspective, the degree to which a recovery approach to mental health service provision was embedded within the organisational culture and practice was viewed as a critical factor (B1-B2). As co-production and co-facilitation represented a significant cultural shift, where services were promoting service user participation there was a greater openness and willingness to implement the programme (B3-B4). In contrast, in those services that had not fully integrated recovery principles, and where the medical model was perceived to dominate, the pace of implementation was slower (B5-B8).
In terms of implementation climate, there were mixed views on the degree to which the programme’s alignment and compatibility with the principles and values of recovery facilitated its acceptance and smooth integration into work plans. Some people believed EOLAS was an efficient way to fulfil the organisation’s local commitments to delivery recovery-oriented care (B9-B10). However, others considered that its recovery orientation was a barrier in some settings, such as acute inpatient care. In this context EOLAS was not viewed as a natural fit, as the emphasis within such services was on containment rather than recovery (B11).
In terms of implementation readiness, there was a consensus that organizational leadership and resources were critical. In services where senior nursing and medical personnel proactively promoted the EOLAS programmes, this greatly facilitated implementation (B12-B13). These key stakeholders not only leveraged their managerial position to promote integration of EOLAS into services’ plans and operational processes (B14-B15), but they enabled information about the programmes to be transmitted through services, increasing buy-in at local level (B16). In addition, they adopted a flexible approach to time management, which facilitated clinical staff to manage the demands of their workload, while contributing to EOLAS implementation e.g. time in lieu (B17). In contrast, where there were difficulties in engaging the support of senior clinicians (nursing, medical), programmes were delivered in a more ad hoc manner and not fully embedded within services (B18-B19). Similarly, where there was a consensus among Multidisciplinary team (MDT) members around the importance and value of the programme, support for implementation was achieved across teams (B20-B21). However, implementation was hindered in teams were there was a lack of buy-in among team members (B22), and particularly consultant psychiatrists (B23), or where there was an over-reliance on one discipline (nursing, social work) or individual champions to implement the programme (B24-B25). Where this occurred, there were significant challenges in generating referrals and enlisting MDT members as guest speakers.
Another aspect of implementation readiness is resources. In terms of resources, where few competing programmes existed, the allocation of resources to support the delivery of EOLAS was straightforward (B26). However, as the number of recovery-oriented programmes increased, competition between programmes emerged. This decreased the availability of support, and personnel to EOLAS (including potential programme participants, as many of the same services users/family members were being targeted for participation in other programmes (e.g. Behavioural Family Therapy, WRAP, Early Intervention in Psychosis Program) (B27-B28).
The level of human resources dedicated to implementation was also an issue. In some services a surplus of facilitators existed which enabled services to accommodate unexpected facilitator absences (B29-B30). Other services experienced difficulties because facilitators (clinical and peer) dropped out due to changes in circumstances (B31-B32). In addition, the availability of time was consistently reported to be a key factor influencing implementation. Some were of the view that their co-ordination or facilitation roles didn’t impact greatly on their current workload, either because the programme operated outside of clinical hours or they had a sufficient degree of flexibility within their roles to enable them to manage their schedules, including receiving time in lieu (B33-B34). In contrast, others reported difficulty in finding adequate time for the planning and preparation required by the programme, either due to the lack of protected time or because the role was considered an add-on to an already cumbersome workload (B35-B36). In similar vein, some reported challenges in securing protected time for staff to attend the EOLAS training programme.
Some services provided venues from within their existing room complement, whereas others, who wished to run the programme in a community setting, experienced difficulties securing venues (B37-B38). The final resource issue was an absence of a system of data management, which prevented teams from being able to systematically identify individuals and families who might be eligible to participate in EOLAS (B39-B40).
Domain C: Intervention Characteristics
The intervention domain focuses on aspects of the intervention. Within this domain six key factors were identified: design; evidence strength; relative advantage; trialability; adaptability; and complexity.
In terms of design, participants reported that the ‘ready-made’ manualised programme was an important enabler for a number of reasons. Where participants perceived they had deficits in service provision or had limited resources to develop programmes to respond to the needs of people with severe mental health problems and/or family members they reported finding the ‘ready-made’ nature of the programme appealing (C1-C3). In addition to the handbooks being perceived as well-written (using layman’s language) and user friendly (easy to read and follow) (C4-C5), they were deemed to be a comprehensive source of information for service users and family members (C6-C7). The Facilitators Handbooks were viewed as providing structure, support and guidance on delivering the programme as well as aiding facilitators to open communication with participants (C8-C11). While the handbooks were appraised as effectively bypassing the time and resource challenges participants encountered when trying to establish similar initiatives, there were mixed views around the referral aspect of the design. Participants understood the rationale for referral through the Multi-Disciplinary Team (MDT) e.g. to ensure that only people with the relevant diagnoses were referred and that participants had ready access to support from the MDT if needed (C12-C13). However, some contributors felt that the referral pathway (through Community Mental Health Teams) contravened the ethos of recovery, on the basis that all recovery education needs to be embedded within community facing initiatives, such as recovery colleges (C14-C16). In addition, some perceived the referral pathway as limiting the opportunity to advertise and promote the programme outside of local services, in-turn impeding recruitment of sufficient numbers to sustain the programme on an ongoing basis (C17-C18).
Participants noted that the information within the handbooks was strongly evidence-based, which enhanced their legitimacy and credibility and buy-in within services (C19-C20), with many viewing the piloting and evaluation of the programmes as an enabler. In their view, these ensured that key informants were consulted about the content, structure and delivery of the programme and the information gleaned in turn informed the ongoing development (C21).
The programmes were also perceived to have a number of relative advantages compared to other interventions, including filling an educational void in relation to psychosis and severe mental illness (C22); being suitable for people recently diagnosed and starting their recovery journey (C23); and having the potential to run alongside other interventions (e.g. Behavioural Family Therapy, WRAP), thus providing service users with a stepped pathway to recovery (C24). In terms of adaptability, there were mixed views. While some facilitators were of the view that by its nature, a manualised programme was rigid and lacked a certain amount of flexibility (C25-C26), most were of the view that the programmes offered a fair degree of flexibility which enabled them to be responsive to the needs of programme participants (C27-C28) as well as enabling them to harness different theoretical perspectives (C29). Some were of the view that the duration of the programme acted as a barrier, with the 8 weeks being a considerable commitment for participants (C30-C31).
The extent of work involved in organising programmes acted as a barrier, as participants perceived it as a complex intervention to implement. Co-ordinators recounted the numerous tasks that had to be fulfilled in order to advance implementation, including recruiting and training facilitators, relationship building with and between facilitators, securing venues and guest speakers, promotion and awareness raising amongst their colleagues, prompting colleagues to refer, assisting with payment difficulties, assisting with the research evaluation and reporting progress back to managers (C32-C33). Although each task in isolation was not particularly onerous, cumulatively they were time-consuming in the context of coordinators and facilitators existing workload (C34-C35).
Domain D: Characteristics of Individuals (Provider Level)
The characteristics of individuals or provider domain is defined as ‘Aspects of the individual provider who implements the innovation with a patient or client’. In the context of EOLAS this included both the clinical and peer facilitators. Within this domain three key factors were identified: Beliefs about the Intervention; Self-efficacy; and Other Personal Attributes.
While some participants were of the view that some members of the mental health teams still lacked knowledge about the programme which impeded their ability to promote it (D1-D2), participants themselves highlighted the need for such programmes (D3-D5) and had a belief in their own ability (self-efficacy) to deliver them (D6-D7). Facilitators’ self-efficacy was attributed to the Facilitator training programme, regular practice, and prior facilitation experience or clinical experience (D8-D10). In terms of other personal attributes, facilitators were of the view that having the ability to communicate compassionately and engage group programme participants, manage group dynamics, manage dual identities (as facilitator and as clinician/peer), negotiate potential tension between content delivery and time constraints, and navigate a non-hierarchical co-facilitation relationship, were important for effective facilitation (D11-D14).
While some facilitators felt they successfully expressed these competencies, others recounted challenges. Clinical facilitators recalled struggling with letting go of their ‘expert’ status and hierarchical position (D15), while some peer facilitators reported struggling with managing their dual identity of co-facilitator and service user, and in establishing an equal partnership with clinical co-facilitators (D16-D17). Job flexibility and family support enabled some family facilitators to participate (D18-D19), while conflicting commitments and ill-health limited the availability of some service user facilitators (D20-D21). Finally, participants considered that a key enabler was the personal motivation and commitment of each individual involved. Facilitators who were motivated and committed were deemed to continuously promote the programme, follow-up on recruitment efforts, engage with and support potential programme participants and make themselves available to deliver the programme when needed (D22-D24). However, some participants noted that motivation was limited to a core group of dedicated individuals (D25-D26).
Domain E: Implementation Process
In terms of implementation key factors identified included planning, engaging key stakeholders and champions, and evaluation.
In terms of planning, identifying formal leaders was of key importance. The hiring of payed project workers, to coordinate the overall project and the establishment of local steering groups within services was viewed as critical. The role of the project worker was not only critical in introducing services to EOLAS and encouraging them to adopt it, but they provided ongoing support to co-ordinators and facilitators on day to day issues (E1-E2). Having a local steering group that comprised all stakeholders was also central, as this group mapped the local pathway to rolling out EOLAS and addressed issues and concerns related to resources, recruitment, and promotion (E3-E5). Many services also appointed a coordinator, from within existing resources, to oversee local implementation, which meant the person took on the extra responsibility of coordinating EOLAS. Having a coordinator with status, credibility and who was capable of influencing and persuading key influential people (E6-E7), was vital to getting EOLAS off the ground. While the coordinators’ activities varied from service to service, coordinators who linked with, and supported clinical facilitators was a key enabler (E8-E9), as successful implementation depended on them working together to plan advertising, dates, venues, guest speakers, and secure and follow up on programme participant referrals (E10-E11).
Successful implementation also involved recruiting and engaging key stakeholders, such as consultant psychiatrists. Securing buy-in from consultant psychiatrists was critical, as they were perceived as a powerful group with significant influence on a team’s approach to care (E12-E13). Coordinators and clinical facilitators spoke of using a number of strategies to engage this group, including presenting evidence from evaluations to having family members and service users make presentations about EOLAS at medical fora (E14-E15). When support (beyond verbal tokenism) was not achieved, referrals to the programme were not forthcoming (E16-E17). In addition to consultant psychiatrists, the recruitment of facilitators was considered critical. While recruitment of clinical facilitators was through word of mouth within services, recruitment of peer facilitators was more challenging. Factors that supported recruitment of peer facilitators, included clinicians having well established connections within community settings and knowing which service users and family members might be interested in becoming a facilitator (E18-E19).
From a peer perspective, a key enabler was the credibility and nature of the interaction they had with the clinician or project worker who issued the invitation (E20-E21). Once recruited and trained, part of successful implementation involved co-facilitators (and sometimes coordinators) meeting prior to and after each session, to plan sessions and foster collaborative, non-hierarchical working relationships (E22-E24). Clinical facilitators and coordinators also described the importance of supporting the wellbeing of peer facilitators (E25-E26), as well as ensuring that all trained peers got opportunities to facilitate programmes (E27).
Successful implementation was also attributed to having multiple dedicated and active local champions (coordinators, clinical facilitators, mental health nurses) within teams who constantly kept EOLAS on the agenda by promoting it at meetings, and continually engaging and following up with colleagues to create an awareness and understanding of EOLAS and to increase referrals (E28-E30). In contrast, in services that depended on a single champion, the loss of this person through turnover or role change threatened implementation (E31-E32) and succession planning for staff turnover was felt to be needed (E32-E33).
The formal evaluation and feedback processes that was built into the EOLAS process were also perceived as an enabler, as this enhanced buy-in among clinical and management personnel (E34-E35), and enabled participants to identify ways in which EOLAS implementation could be improved (E36). While the evaluation process was an enabler, the time-consuming nature of producing and updating the programme handbooks was an unanticipated barrier, as delays in the availability of up-to-date handbooks slowed down implementation within sites for a time (E37-E38).