The PECAN pilot trial
The full pilot trial details are reported elsewhere [22]. In summary, the PECAN pilot trial was planned as a multi-centre pragmatic trial with a two-armed, parallel group design. Ethical approval was obtained from the responsible ethics committees. Residents were included if they were aged 65 years or older and affected by at least one joint contracture diagnosed by a physician, therapist or nurse. Residents at the terminal stage of a disease were excluded. Seven nursing homes (i.e. the clusters) with a total of 129 residents were recruited from a convenience sample in two German regions. Prior to the start of the study, all the residents (and/or the legal guardians) were asked for a written informed consent by the research team. Structured face-to-face interviews by blinded assessors were used to collect residents’ data at baseline, then after three and six months. The primary outcome was defined as the residents’ participation and measured with the PaArticular Scales [23]. The secondary outcomes were defined as residents’ activities, instrumental activities of living, health-related quality of life, as well as falls and fall-related consequences to ensure the safety of the intervention. After baseline assessment four nursing homes with 64 participating residents were randomised to the intervention group (PECAN) and three nursing homes with 65 residents were randomised to the control group (optimised standard care i.e., standard care including an information session addressing general aspects of care for residents with joint contractures).
Study design of the process evaluation
A mixed-methods process evaluation was employed with data collection in conjunction with the PECAN pilot c-RCT. As recommended for process evaluation studies, we applied quantitative methods to assess whether the key processes of the implementation followed the study protocol and qualitative methods to determine enablers and barriers during the implementation [20]. Quantitative and qualitative data were given equal consideration, as they complement each other in a deeper interpretation of the findings [24].
We applied the MRC guidance for the evaluation of complex interventions by Moore et al. [20] along with the framework proposed by Grant et al. for the design and reporting of process evaluations for c-RCTs [25]. Grant et al. differentiate in their framework between processes involving clusters, processes involving individuals and their interaction with the context in which the trial is embedded [25]. Since the PECAN intervention is delivered first to the nursing homes and not directly to the residents, this process evaluation focuses on processes involving the nursing homes (i.e. the clusters) in order to improve the implementation strategy for the main trial. We used the Standards for Reporting Implementation Studies (StaRI) Statement [26] for reporting our implementation and the Template for Intervention Description and Replication (TIDieR) checklist [27] for reporting our intervention.
The PECAN intervention
The core idea of PECAN is to facilitate a participation-oriented understanding of care in nursing homes, based on the biopsychosocial model of the ICF [8] to allow improved analysis of the residents’ situation and to guide the nursing home staff in their decision-making. The mechanisms of the expected changes in the nurses’ professional behaviour to improve the residents’ participation are based on the principles of the Theory of Planned Behaviour (TPB) [28]. This theory describes attitudes, subjective norms, and the felt degree of control over behaviour as the essential determinants of behavioural change and states that a change in behaviour can be initiated through appropriate interventions [28]. The approach of the TPB is proven to predict or explain the behaviour of healthcare professionals [29, 30]. Intermediate intervention goals to change the behaviour of the nursing home staff are presented in the logic model of the PECAN intervention in Additional file 1.
The PECAN intervention focuses on the dynamic interaction between an individual’s health condition and existing personal and environmental factors that can act as facilitators or barriers in performing activities and participation [17, 8]. The policy of the PECAN intervention is to improve the residents’ participation by adapting environmental factors, taking personal factors into consideration, and supporting mobility. Therefore, an individually tailored approach is used, including both the individual (i.e., resident) and the organisational (i.e., nursing home) levels to achieve the intervention goals [17].
On the individual level, the residents’ activities and participation are addressed by defining individual participation goals. Accordingly, the residents’ care plans and daily routines are reviewed and adapted. Measures to meet the participation goals on the individual level contain, for example, the use of a biographical approach to identify the residents’ potential motivation for activities and participation, the optimisation of the provision of medical or technical aids and the involvement of additional persons in the daily care, such as external therapists or relatives.
On the organisational level, the review and change process to integrate the perspective of the ICF is guided by using a checklist with predefined criteria. For example, changes on the organisational level can comprise the adaption of the care concept, as well as the redistribution of tasks involving the nursing home management, the nursing team and the interprofessional team (i.e., social workers, physicians and therapists) [17].
Implementation strategy
The PECAN implementation strategy used a facilitation approach [21]. Facilitation is the active part of the implementation, carried out by trained facilitators, who guide individuals or organisations through a challenging change process [31, 21]. As change agents, facilitators were responsible for guiding the implementation and for offering education and counselling to their colleagues. The implementation of PECAN proceeds in multiple steps: In the first step, the intervention is introduced to skilled nurses, who are trained as facilitators. The research team guided the delivery of the intervention across all nursing homes. In the second step, the facilitators are responsible for the integration of the PECAN intervention into the daily practice by involving the nursing team, physicians, therapists, social workers and relatives. During this process of the implementation the facilitators were supported by experienced peer-mentors, who were members of the research team. The targeted changes in the nurses’ professional behaviour was expected to improve the residents’ participation [17].
An overview of the PECAN implementation strategy is presented in Figure 1.
Kick-off meeting with the head nurse/nursing home director
In the kick-off meeting, the intervention was introduced to the head nurse and/or the director of the nursing home, who signed a declaration ensuring their commitment.
Facilitators’ workshop
The key component of the implementation was a one-day facilitators’ workshop to prepare nominated skilled nurses who have received a degree for their role as facilitator following at least three years of formal vocational education. Based on predefined qualification criteria (e.g., formal vocational education) the facilitators were selected by the head nurse. During the workshop, the intervention was explained, including comprehensive information on the phenomenon of joint contractures, a training session on how to implement residents’ participation goals in individual care planning using the biopsychosocial model of the ICF, and a training session on peer counselling methods to involve all team members in the implementation process and to improve interprofessional collaborations.
Information session
A single information session lasting 40 minutes was held by a member of the research team in each nursing home for the residents, relatives, nursing staff and other interested health-care professionals (regardless of their participation in the study). In the intervention group the aim of the session was to introduce the PECAN intervention, the facilitators and their tasks, and to provide ideas about how everybody could support the implementation. In the control group the aim of the session was to inform about risks and consequences of joint contractures, to introduce the study, and provide contact to the research team.
Peer-mentoring
The facilitators were supported via a mentoring approach, where they received counselling by a trained mentor (a nurse in the research team). Starting with the peer-mentor visit in each nursing home, a mentor and an external peer expert gave the facilitators counselling and support in evaluating and adapting implementation measures tailored for their institution. Using structured assessment tools, the facilitators reviewed the residents’ individual care plans and the organisational procedures (in collaboration with the head nurse) in order to identify barriers and enablers for the residents’ participation. Following the visit, peer-mentoring was conducted via phone calls from their mentors every second week in the first two months and later once a month. The peer-mentors were free to offer fixed and regular counselling appointments or to provide counselling only if required. The peer-mentors at both study centres shared their experiences in regular telephone meetings and discussed with a third member of the research team any problems that arose during peer-mentoring.
Supportive materials
Posters and other written materials were provided to inform and remind nursing home staff and residents. Outpatient therapists, physicians and relatives were addressed by leaflets with customised information about the intervention and contact details of the facilitators.
Standard care – the context
In Germany, nursing homes are financed by the German statutory long-term care insurance and additional payment from the residents. On a legal basis, 50% of the nursing staff had to be skilled nurses with at least three years of vocational training. Nursing home residents are frequently affected by age-related disorders and multimorbidity. Social activities are usually planned by in-house social care assistants and social workers. Physicians and therapists typically do home visits to the nursing homes. Medical and technical aids as well as physical therapy, occupational therapy and speech and language therapy need to be prescribed by a physician and are financed by the German statutory long-term care insurance with a co-payment from the residents.
Study population of the process evaluation
The study population of this process evaluation included all persons who were closely engaged in the implementation of PECAN and provided the perspective:
- of the facilitators, responsible for the implementation of PECAN.
- of the nurses, who were introduced to the intervention by the facilitators.
- of additional persons, who were closely engaged in the care of residents with joint contractures, i.e., therapists, social workers and relatives.
- of the research team, especially the trained peer-mentors, who were responsible for support of the facilitators during implementation.
The nursing team included skilled nurses, nursing assistants, nursing students and social care assistants, since they represent the nursing team in each nursing home ward. Therapists were defined as physical or occupational therapists employed by the nursing home or by an outpatient practice. Social workers were employed by the nursing home and were responsible for supporting residents in independent living and social participation, e.g., organisation and coordination of individual and group offers. Relatives were defined as a family member or a legal guardian of a participating resident and were randomly selected by the research team based on the participants’ list of the residents. The residents had already been involved in the feasibility testing of the study procedures and were asked to participate in structured face-to-face interviews. We decided to exclude residents from the process evaluation of the interventions’ implementation to keep the burden of questioning as low as possible for the residents in this pilot trial [22].
Data collection
Data were collected prior to, during and post- intervention to illustrate changes over time [20]. Figure 2 displays the flow of the process evaluation. During data collection we focus on the component’s “delivery to clusters” (i.e., process where the research team delivers intervention content to the nursing home), “response of clusters” (i.e., process where the nursing home adopt intervention content into daily nursing care), and “the context” (i.e., anything external to the intervention) which might be an interacting component [25]. An overview of the components and data collection methods of the process evaluation for the PECAN intervention adapted from Grant et al. [25] is presented in Table 1.
Characteristic of nursing homes – the context
Characteristics of the included nursing homes were collected at baseline via structured interviews with the head nurse or the director of the nursing home.
Delivery to and response of clusters
Process of implementation
The facilitators’ workshop and the information session were evaluated by their participants with standardised questionnaires to assess content-related (e.g., relevance for professional development, practical relevance) and educational aspects (e.g., structure, comprehensibility, quality of training materials). As overall feedback, the participants rated the events on a scale ranging from 1=“excellent” to 6=“inadequate”. The predefined qualification for the role of facilitators was reviewed in detail as part of the survey (e.g., formal vocational education). The participants in the information session were asked whether they were nurses, relatives, residents, or members of other groups.
Standardised documentation forms were used by the research team to review the implementation process according to protocol. We assessed the attendance in the information session (number and group affiliation of participants), the fidelity of the peer-mentor visit (number of participants, procedure according to protocol), the fidelity of the counselling interviews during peer-mentoring by telephone (content, number of interviews per facilitator, interview duration), and amount and type of supportive materials used (e.g., leaflets, poster). To gain insight into the “what” of the intervention at the nursing home level, the facilitators’ activities during the implementation process were summarised in the facilitators’ diary.
Attitude and behaviour of nurses
A standardised questionnaire was used for a survey on the nurses’ professional attitude and behaviour in order to reach the target 20% subgroup of nursing staff in a short time. The questionnaires were distributed by the head nurse in the intervention group and control group at baseline and at the 6-month follow-up (convenience sample). Participants were randomly selected based on their presence (staff roster) during the data collection period. Nurses were asked to rate six statements about the care of residents with joint contractures to verify to what extent the PECAN intervention is associated with a professional change in behaviour. Three additional statements regarding the reach of the intervention were rated exclusively in the intervention group at the 6-month follow-up. All statements were rated on a 5-point Likert scale (1=“strongly agree” to 5=“strongly disagree”; with “don’t know” as a sixth option).
Enablers and barriers of implementation
After the intervention period a detailed insight into the experiences of all stakeholders was needed. Therefore, all the facilitators were invited to join a group discussion in their respective study centre. Facilitators who could not join in were asked to participate in a problem-centred interview. Relatives, therapists, social workers, and the trained peer-mentors were also invited to take part in problem-centred interviews.
Both the problem-centred interviews and the group discussion followed semi-structured interview guides. To identify key enablers and barriers of a successful implementation, questions were asked regarding how the intervention was delivered, who was reached, how every single implementation component was experienced, and which outside factors were influencing the implementation.
The group discussion was moderated by one researcher (HK) and a study assistant at the study centre. The problem-centred interviews were conducted by single researchers (HK, JH, KB) at the participants’ workplace or at home via telephone. All the interviewers were trained by the research team in methods of leading group discussions [32] and problem-centred interviews [33]. The interviews and the group discussion were audio recorded. Field notes were taken and summarised in a post-script.
Data analysis
Quantitative data were analysed by descriptive statistics using SAS Version 9.4 [34].
Qualitative data from the problem-centred interviews and group discussions were analysed using a mixed deductive-inductive approach based on the structured approach of directed content analysis [35]. Audio records of the group discussion and the interviews were “abridged transcribed” [32] with priority given to relevant contents by members of the research team (HK, JH, KB). Meaningful examples of quotations from the participants were transcribed verbatim. For quality assurance reasons, the participants were offered the opportunity to review and modify the transcripts.
Two researchers (HK, KB) developed a coding guideline based on one transcript from each group of participants. To finalise the coding guideline, categories were cross-compared and discussed until a consensus was reached [36]. The final coding guideline was reviewed by two senior researchers (MM, SuS). Any data that could not be categorised with the initial coding guideline were assigned to a new sub-category. Where reasonable, the description of the categories was based on the categories of the ICF, which was the conceptual model used to design the intervention [37, 8]. The data analysis was supported by MAXQDA Version 12 [38]. The results were classified into enablers and barriers.
Qualitative data from documentation forms or minutes and field notes were classified inductively into categories, based on the content of the given answers.