«Citation: Dalkin, Sonia (2014) The Realist Evaluation of a Palliative Integrated Care Pathway in Primary Care: What Works, For Whom and in What ...»
A further barrier to health care professionals working proactively and broaching ACP is the false hope some patients have about their prognosis. Health care professionals then do not want to diminish these patients’ hopes. Thus, false hope can create an inhibitive context for the shared decision making mechanism to take place. The literature suggests that information can be given to patients without negativity or instilling false hope using phrases such as “there is little chance of A or B but a real chance of C” (Penson 2000). It is thought that this approach puts emphasis on what the patient can hope for (Penson 2000). However, findings from this study (Chapter 6, pg. 199 and Chapter 7, pg. 231) highlight that it is not easy to engage patients who have false hope in ACP, thus suggesting that the technique suggested by Penson (2000) may not be effective. Thus although false hope has been explored in the literature, some of the techniques to overcome it may not be viable in practice. Furthermore, the root of false hope about prognosis, which these findings highlighted as secondary care, has not been previously investigated as a barrier to ACP in primary care. In order for health care professionals to be able to broach ACP and work proactively, secondary care must also acknowledge that the patient is appropriate for palliative as opposed to curative care.
Another barrier to ACP was highlighted in relation to non-cancer patients in these findings; health care professionals have a fear of repercussions from inaccurately assessing mental capacity which is required for ACP. The literature states that if a non-cancer illness such as dementia is empathetically discussed at an early stage it would allow the patient to express their wishes whilst they had capacity (Robinson, Hughes et al. 2005), which may increase the use of advance statements to deliver more proactive and patient centred care (Hughes and Robinson 2004). However, although the premise of ACP is meant to be to document preferences before capacity is unstable, in practice this is not always viable. This thesis highlighted that often preference discussions occur in the amber phase, when patients are declining and thus can have capacity issues, especially in those with non-cancer diagnoses.
Since the premise of ACP is that it happens prior to a time when an assessment for capacity is needed there has been little research into the assessment of capacity as a barrier. The findings from this study are grounded in practice - they suggest that ACP can occur late and that in this case capacity will need assessment, which health care professionals have low confidence in doing and a fear of repercussions from an inaccurate assessment. However, the overall programme theory mechanism can still fire in these circumstances, in the form of preference discussions. Yet, formal documentation of these preferences, which is a more robust method, may not be possible. In order for health care professionals to work proactively they must have an understanding that even when patients are identified in the amber phase, efforts to document preferences are still valuable.
Study findings in relation to the current literature on ICPs The limited literature on palliative care ICPs suggests that they result in positive outcomes for patients (increased home deaths, improved quality of life and mood, optimal symptom control) (Bakitas, Lyons et al. 2009, Bower, Roderick et al. 2010, Temel, Greer et al. 2010) and health care professionals (increased job satisfaction, ability to provide palliative care, communication and coordination) (Hall, Goddard et al. 2011, Reymond, Israel et al. 2011). However, they give limited understanding of the underlying reasons behind the success, no explanation of mechanism or context. The findings from this study, although drawn from one practice example, highlight underlying mechanisms with more translational potential (implementation, shared decision making, patient centred care, and proactivity). They show the importance of identifying and understanding underlying mechanisms (resources and reasoning), alongside conducive contexts. Inhibitive contexts and mechanisms are also important to identify, as they highlight what could be changed to achieve positive outcomes in a specific situation. Collectively, the CMOCs identified in this thesis uncover general themes and needs (described above) to provide high quality palliative and end-of-life care in primary care. This highlights the potential of realist evaluation to contribute to the knowledge translation agenda of research findings into practice.
Implications for practice The study findings highlight that new social programmes require active implementation strategies that encourage the innovation to become embedded into practice. They also indicate the importance of context and how this can affect the implementation and effectiveness of a social programme.
The findings highlighted that shared decision making in palliative care should be a lot more inclusive in practice, involving more significant individuals than just the patient and the health care professional. This may allow for more realistic plans and less undesired outcomes (for example, emergency admissions). Building upon this, the role of ACP should be reconsidered as a psychosocial tool in practice to aid in shared decision making, rather than a tool with the main goal of capitalising on capacity in order to provide patient centred care to those who lose autonomy for health care decisions. This reconsideration of ACP may make it easier for health care professionals (and potentially care home staff) to engage with and aid discussions surrounding death and dying which have previously been identified as difficult. Thus, as the End of Life Care Strategy (Department of Health 2008) states, palliative care is appropriate for all, as is ACP; it is not only for those who are likely to lose capacity. Furthermore there is scope for shared decision making to play a more explicit role between health care professionals forming the best care possible for a patient. It could be said that this already occurs in MDT meetings but conceptualising it as shared decision making in practice may facilitate communications and aid health care professionals in their interactions. Barriers to proactive care in palliative care still remain and these are areas that health care professionals need to focus on in order to provide higher quality palliative care for all. If these barriers are not addressed, issues related to ethical justice will continue to occur in terms of provision of palliative care to those with non-cancer diagnoses, especially those who lose capacity (Koffman 2012). Furthermore, in order to provide proactive care and avoid solely reactive care, health care professionals must be given time efficient tools. These implications are relevant to many individuals involved with the ICP - health care professionals, care home staff, care home managers, patients, family members, carers, OOH staff, and commissioners due to the broad scope of individuals involved with the ICP.
Challenges Greenhalgh, Robert et al. (2004) highlight that drawing realist conclusions about the generative causality of particular programme theories can only be achieved though much negotiation and contestation. Developing and refining CMOCs in this study has been a time consuming and demanding process, and although it is presented in a logical and neat format in the thesis, the process was somewhat disordered. This was due to the methodology (realist evaluation) and several forms of data collection and analysis that were used collectively and concurrently to form many CMOCs.
However, the continuous refinement of CMOCs has resulted in findings that reflect the complexities of health care, primary care and palliative and end-of-life care. For example, the embeddedness CMOC4 (I) and CMOC4 (II) were developed over a long period of time throughout which they were constantly refined. Initially, I considered the MIQUEST and Death Audit data and what the numbers meant; use of more interventions and more ‘home’ deaths were undoubtedly positive outcomes.
Following this my attendance at PCQV, the Palliative Care Partnership and integration into the field identified a difference between practices. There were those that were very involved, attended Palliative Care Partnership meetings, engaged in PCQV discussions enthusiastically, discussed issues as a team, and those that did not. This prompted me to wonder whether those GP practices that behaved in this positive way used more interventions and/or had more ‘home’ deaths. Looking at the MIQUEST and death audit data some practices did have higher numbers than others, but this was a very subjective observation. The cluster analysis transformed my observations from the field and of the data into a robust outcome – there were three high performing practices out of the fourteen under study. From here I needed to establish the common essential context and mechanism that led to increased embeddedness in these three practices. The shared nursing team on site was identified first. This was initially thought of as a mechanism. However, further consideration refined the shared nursing team on site into a context that facilitated MDT meetings. The mechanism was then identified using my own knowledge of the locality, conversations with the founder and confirmed in FG3. I knew of some key individuals who attended the Palliative Care Partnership regularly, took interest in the study and could be considered as a champion or an opinion lead.
Conversations with the founder and within FG3 confirmed (without prompt) that these people were key figures in pushing the palliative agenda in their practice.
Hence the mechanism was formed, having a champion or opinion lead (resource) who pushes the palliative agenda within the team and highlights its relative advantage (reasoning). Although this seems fairly straightforward to explain it required a lot of thinking, sketching (using the CMOC diagrams) and collaborative working with the founder, the health care professionals in FG3 and my supervisors who offered methodological support. It also required adjudication between competing theories, i.e. was it the champions and opinion leads in the practices or was it that these teams generally had enhanced communication skills? This could have been the case but FG3 participants highlighted the champion and opinion leads in the three high performing practices as pivotal.
Facilitating the refinement of CMOC was the increased understanding of contexts, mechanisms and outcomes as the project progressed. At first I found it difficult to decipher between contexts and mechanisms especially, a difficulty reported by other realists (Jagosh, Pluye et al. 2013). Deciding what constitutes a context and a mechanism requires intense thought and debate - the ICP offers a great variety of resources, used in an array of contexts impacting at a range of levels. However, I found as my time working with realist evaluation progressed, so did my knowledge of the ICP and I feel this facilitated my ability to formulate CMOCs with increasing explanatory potential. I also found that drawing several versions of a CMOC (using the CMOC figure framework used throughout this thesis) often helped me to decipher between context and mechanism and the exact formulation of each.
Finally, understanding mechanisms in terms of resource and reasoning was extremely helpful, a technique suggested by Dr. Gill Westhorp on the Realist And Meta-narrative Evidence Syntheses: Evolving Standards (RAMESES) Discussion list (RAMESES 2014). Using resources and reasoning allowed me to understand fully what tool was being used (resource), and whose decision making process this was changing (reasoning). This method was also helpful when discussing mechanisms with health care professionals in FG3. Identifying these two factors often made the mechanism more obvious, which in turn highlighted the context.
Furthermore, throughout my time using realist evaluation I developed an understanding that the concept of mechanisms ‘firing’ in context did not always apply. This is because, at the individual level especially, reasoning is often on a continuum. People have enhanced patient centred practice or increase their confidence due to a resource. The resource in itself sometimes does not produce the reasoning but can instead enhance it, thus mechanisms often have increasing degrees of activation. However, at the system or practice level, it is sometimes more definitive; for example if the ICP is not present, it is not embedded into practice – the mechanism cannot fire. I feel that use of resources and reasoning highlights and explains the complexity of CMOCs; human volition is not often comprised of complete oppositions, and often there are spectrums. For example, different GPs will be situated at different points on the continuum of patient centred practice.
Mechanisms can have degrees of activation as opposed to being categorical (‘firing’ or not ‘firing’).
Not only was the understanding and refinement of CMOCs difficult but also the change from a positivist standpoint to realist logic of analysis. Sporadically I would feel a positivist mind frame creep back in and would have to pull myself back in to the realist evaluation world. Facilitating working within a realist logic and confirmation of its value was made apparent when I read non-realist literature. Time and again, I was left wondering what made an intervention work, even when it was reported as successful.