Co-construction of the family-focused support conversation: a participatory learning and action research study to implement support for family members whose relatives are being discharged for end-of-life care at home or in a nursing home

The study was informed by Normalization Process Theory (NPT), a structured and theoretically robust approach to understanding the factors that promote and inhibit implementation [33, 34]. NPT proposes that implementation is a dynamic interactive process, influenced by the social actions of those involved. This process is described by four constructs: coherence, sense making work; cognitive participation, relational work; collective action, operational work; and reflexive monitoring, appraising work.

NPT was integrated with a Participatory Learning and Action (PLA) Research approach, a combination previously demonstrated to positively influence the quality of intervention design and implementation, by ensuring inclusion of diverse sources of knowledge and expertise [35]. In this paper we describe three PLA cycles concerned with intervention development and pilot implementation. These cycles broadly followed an intervention development process described by Hawkins and colleagues [36] consisting of: evidence review; co-production of a conceptual framework and interventional structure and process; and pilot implementation.

The evidence review focused on research reporting interventions to provide family caregiver psychosocial support during palliative and end-of-life care. Psychosocial support was defined as support concerned with the emotional and relational wellbeing of family members [37].

There are a growing number of systematic reviews assessing psychosocial interventions for family caregivers during palliative and end-of-life care. However, systematic reviews rarely provide enough detail about interventions to allow implementation [38], but they do provide rigorous assessment of the methodological quality of studies. We therefore identified studies to review from these sources. Systematic reviews were identified from the meta-review by Thomas et al. [9], and from a search for systematic reviews published subsequently. All studies included by the review authors were assessed for relevance (Table 1). They were included if focused on caregiver support during palliative and end-of-life care, if graded by the systematic review authors as being of good methodological quality and if the intervention was amenable to delivery by hospital-based registered practitioners. Identified studies were analysed to identify the theoretical framework(s) underpinning the reported interventions and the interventional mechanisms identified by authors responsible for the reported therapeutic outcomes.

The outcomes of the evidence review were synthesised into a conceptual framework from which the structure and process of the intervention was co-constructed with PLA participants. There is very little guidance about how to undertake this process in healthcare [39, 40] but we previously found participatory theatre techniques [41, 42] valuable in drawing out participants’ cultural, clinical and social knowledge to inform the translation and synthesis of evidence into healthcare processes [43]. We integrated these approaches in the following steps:

Following ethical and local site research governance approvals (REC ref.: 16/SC/0330) pilot implementation was undertaken over a three-month period by seven clinical co-researchers (described above), most of whom were involved in intervention development. In preparation, clinical co-researchers attended a half-day workshop where the intervention was reviewed, discussed and rehearsed. Given one of the purposes of pilot implementation was to assess potential disruption caused by implementation, each clinical co-researcher was asked to limit intervention provision to five interventions.

Data collection was informed by NPT and collected via reflective records of intervention delivery, detailed records of telephone conversations between the research team and clinical co-researchers discussing implementation progress and detailed records of evaluation meetings.

Data analysis was undertaken by three members of the research team (NC, SL, SD), following a Framework Analysis approach [45]. Data were initially coded, using the broad constructs of NPT, followed by a more detailed analysis, using NPT-generated questions designed by Murray et al. [46] as an analytical framework. A workshop to discuss interpretation of data analysis was held with PLA participants.

In addition to research ethics and site governance approvals required for implementation (REC ref.: 16/SC/0330), PLA raises ethical concerns about the location of ‘power’ in researcher and participant relationships and how this is managed to achieve the collaborative relationship intended [35]. Detail of how we approached these concerns is provided in supplementary information 1.

Full details of the evidence review is provided in supplementary information 2. The process is summarised in Fig. 1.

A total of 20 systematic reviews were identified; 18 reviews graded as good or medium quality by Thomas et al. [9], focused on end-of-life care or cancer care [8, 10‐26] and 2 systematic reviews published subsequently [27, 28]. Nine reviews were excluded because they focused on evaluation of services [10, 11], bereavement interventions [12], communication [13], or interventional approaches not amenable to hospital provision, such as art therapy [14], physical activity [15], web-based delivery [18] and mindfulness [27, 28]. The remaining 11 systematic reviews [8, 17‐26] comprised 103 studies of which 96 were excluded.

The seven papers identified for analysis, related to 6 studies focused on psychosocial support for family members, during end-of-life care [47‐53]. Two of these studies related to the FOCUS [49, 50] and COPE [51, 52] interventions and therefore manuals for these interventions were obtained from the authors.

Four studies explicitly stated the theoretical framework(s) adopted [47‐50], in one it was implicit [50, 51] and in another not stated [53]. Five studies were underpinned by stress and coping frameworks derived from the work of Lazarus and Folkman and aimed to mitigate uncertainty, distress and carer burden, enhance family communication and increase access to resources [47‐52]. One study also drew on Horowitz’s work [47], which emphasised the importance of meaning-making in adaptation to difficult life events (Table 2). Thus, collectively, the reviewed studies were theoretically modelled on stress, coping and meaning making, and focused on quality of life.

The interventional processes identified from the review included the provision of information, sense-making, problem-solving, resource networking and self-care. All interventions were delivered over successive consultations, for example FOCUS and COPE were provided in three structured visits of at least 30 min each. There was insufficient detail in the reviewed studies to assess the effectiveness of proposed therapeutic mechanisms on outcomes. Therefore, in consultation with PLA social care experts and clinical co-researchers, we considered it prudent to adopt recommendations from Northouse and colleagues [17] to focus intervention processes on mitigating uncertainty and from Candy and colleagues [8] to at ‘the very least healthcare practitioners should enquire about the concerns of family and friends caring for a loved one’ (p23) and incorporate information and problem-solving coaching processes, to buffer psychological distress.

PLA participants stressed the importance of the intervention being equally applicable to any family member, irrespective of their role in future care, and deliverable as the opportunity arose. Use of a pre-existing intervention such as FOCUS or COPE was therefore considered impractical and inappropriate. Therefore, to ensure the intervention could be delivered flexibly, depending on family and clinical circumstances, we used a structured conversation design, a brief intervention design considered clinically feasible for provision of evidence-based psychosocial support [54].

In addition, PLA participants emphasised the importance of involving family members in end-of-life care decisions, recognising their knowledge of the ill person and ‘how they do things as a family’. Therefore, the intervention was theoretically modelled on Family Sense of Coherence [55]. Family Sense of Coherence is theoretically congruent with the stress, coping and adaptation theories underpinning the reviewed studies, but rather than focusing on ill health (distress) and quality of life, it focuses on family strengths, resilience (coherence) and salutogenesis (wellness). Coherence is influenced by whether life events are considered comprehensible, manageable and meaningful and congruent with ‘how we do things as a family’ [55]. Family Sense of Coherence was also theoretically consistent with the interventional processes of meaning-making, problem-solving and harnessing resources, identified in the evidence review.

Importantly, the meaningfulness component of FSC is considered key to family coherence [55]. Meaning focused approaches are particularly helpful in situations characterised by uncertainty [56] by helping to realign priorities and create or renew a sense of purpose [57]. This approach therefore suited end-of-life care transitions where family priorities move towards the care of a dying member, when there is limited time to make decisions about place of care and where there is uncertainty about care provision and the future. Thus, addressing uncertainty by focusing on meaningfulness is likely to strengthen family coherence. Family members with strong family coherence are likely to have positive caring experiences [58‐60], and confidence in end-of-life care provision [61].

Consequently, the internal logic of the resulting conceptual framework proposed that family members’ uncertainties about end-of-life care could be reduced by identifying and addressing their concerns, by providing information and coaching problem-solving. As a result, family members would make informed decisions about their role in care and harness family and community resources. Thus, the intervention would foster family coherence (resilience) (Fig. 2).

The structure and process of the intervention, resulting from the participatory theatre workshops, are outlined in Table 3 and Fig. 3. This description includes a revision made by clinical co-researchers during implementation training, to use an empathetic phrase at the beginning of the intervention to orientate participants to the family focus of the conversation, followed by a pause. This combination of empathetic statement and pause, increases the likelihood that concerns are raised and addressed in clinical consultations [62].

Data consisted of n = 22 reflective records of intervention delivery (site 1: n = 5; site 2: n = 10; site 3: n = 7), n = 3 records of discussion of implementation progress and n = 2 records of evaluation meetings.

Table 4 provides a summary of analysis, constructed from data coded against NPT-generated questions for intervention evaluation [46]. The table provides the source of data quoted in italics below. All quotes were provided by clinical co-researchers - specialist nurses and occupational therapists in palliative care..

All clinical co-researchers implemented the intervention. Some needed to see the intervention delivered or to deliver it themselves, ‘to get it’. Most needed to practice how to phrase the interventional prompts so that the intervention ‘flowed’ and was consistent with their communication style and patient population. Some approached implementation by delivering the intervention whenever they spoke to a family member about discharge care plans, irrespective of circumstances. Others initially selected opportunities to use the intervention, starting with more straightforward discharge situations. As co-researchers gained confidence, the intervention was incorporated into practice, irrespective of the complexity of discharge.

Co-researchers described how the intervention disrupted and ‘pulled apart their practice’ by ‘flipping conversations to focus on the family’. They described a shift in their understanding of family support as something professionally focused and ‘done to families’ to practice centred on ‘the meaning for the family and their concerns’. They also discriminated the intervention from ‘discharge itself’ or ‘bits of kit’ or ‘a checklist’. They reported a deeper understanding and respect for family care and increased satisfaction with discharge work and provided examples of how the intervention enhanced family care (see Table 4).

Implementation was also influenced by the local working arrangements for discharge between specialist and ward teams. When discharge was considered a ward team responsibility, clinical co-researchers were worried the intervention would disrupt these arrangements and create ambiguity about their role. Some clinical co-researchers therefore restricted use of the intervention to occasions when discharge conversations had been commenced with family members by ward staff. Others used the intervention to clarify concerns family members had previously raised with ward staff or by delivering the intervention in collaboration with ward staff.