«Citation: Dalkin, Sonia (2014) The Realist Evaluation of a Palliative Integrated Care Pathway in Primary Care: What Works, For Whom and in What ...»
Undoubtedly realist evaluation is time consuming and can be frustrating to use. Yet these negatives are overcome by the satisfaction derived from creating refined and robust programme theories that provide credible explanations for observed outcomes. Those outcomes can collectively then inform practice more generally, for example in terms of shared decision making. Now I have become experienced with realist evaluation I feel that it offers a wealth of knowledge to the research field alongside a refreshing change from successive causality. It moves on from using a pure science technique with complex humans. Rather than trying to fit square boxes in to round holes it embraces human volition and understands that it is an inherent feature in the process of complex social programmes.
Limitations An issue that is beyond the control of the study is the variability in recording of GP practice data (MIQUEST and Death Audit). Health care professionals could be using ICP tools such as palliative care registrations, preference discussions and advance care planning and not recording this use. However, the qualitative results seem to reflect health care professional and patient experiences. For example, most palliative care patients and bereaved relatives recalled preference discussions but few had completed locality advance care plans. Also, health care professionals described their dislike of the locality advance care plan format and the time consuming paper work it generated. In relation to palliative care registrations, the quantitative data identified less non-cancer registrations than cancer registrations, despite increases in non-cancer death registrations from 2011 to 2012. Not only does the literature describe the difficulties in the non-cancer trajectory (Murray, Boyd et al. 2005) but health care professionals in focus groups commented on the difficulty they have when registering non-cancer patients on the palliative care register. Finally, the embeddedness chapter identified that there is an underlying context and mechanism to explain why some practices have embedded the ICP better than others, explaining that outcomes could not just be linked to a recording artefact. These examples and supporting evidence indicate that the recordings that health care professionals do on the data templates are representative of what actually happens in practice, although potential variability needs to be acknowledged.
Another limitation is the self-selected or health care professional selected recruitment. Focus groups included motivated health care professionals prepared to give up their time to participate in the study. However, these individuals were very honest, for example, stating their dislike and neglect of the locality advance care plan. Patients and bereaved relatives participating in interviews, consultation recordings and the QDDM were selected by health care professionals. However, there was no other means of access to these potentially vulnerable people. All participants were very candid in their discussions and often would comment on negative issues that they had encountered, thus providing a balanced view of the service.
A potential criticism of the study could be that the focus of the research has been on health care professionals’ decisions - decisions to use tools, register patients, and facilitate home deaths. However, focusing solely on patient experiences or outcomes would have happened if more traditional quantitative or qualitative research paradigms were used, assuming the causal relationship between intervention and outcomes. The focus on health care professionals in this thesis is due to the use of realist logic; the search for underpinning mechanisms. Health care professionals embody these mechanisms and thus they were the primary focus of the study, as the aim was to discover if the ICP worked, how and in what circumstances. This is why some lines of enquiry were not pursed, such as bereaved relatives having feelings of allegiance to their loved ones wishes and not accepting end-of-life care in a hospice for example. The purpose of using interview data was to highlight their experience of the ICP and provide practical information to health care professionals concerning how it works and how it could potentially be used more efficiently. For example, quotes from bereaved relatives about guilt and allegiance to their loved ones highlighted that a bed in the palliative care unit had been offered when end-of-life was approaching. By this time, participants felt that they had lived through the worst of their loved one’s worsening health and had refused to use the service. Interviews with palliative care patients and bereaved relatives thus helped to understand the implementation of the pathway (for example, if the palliative care unit is offered late in the patient’s illness it is not accepted due to feelings of allegiance and guilt). It could be that people’s relationships and closeness will always form a context for services implementation, and it could be that allegiance and guilt are particularly salient in palliative and end-of-life care.
However, exploring further the explanatory potential of such feelings in interviews would have caused additional emotional strain and been unethical. The interview of people in highly emotionally charged situations, such as a bereavement, is an acknowledged challenge in qualitative research and can only be conducted under very strict ethical guidance. Thus, this study highlights that guilt and allegiances may be overarching inhibitive contexts that health care professionals need to operate within in order to trigger shared decision making, and that early implementation may minimise the impact of these feelings on poor uptake, but this could not ethically be tested with greater certainty.
The ICP is adaptive, which is to its credit. The recognition of the locality advance care plan’s flaws from various sources resulted in the change to the advance statement to facilitate the documentation of patient preferences. This makes this section of the ICP evaluation less relevant to the locality but still generates important findings in terms of underlying principles. The practicality of tool use in a time constrained primary health care environment needs consideration, as does the assessment of mental capacity in ACP.
The data set in this study was complex and management of it was difficult. There would have been benefits in collecting and analysing data iteratively (for example, complete interview 1 and follow interview 1 issues up in interview 2 to confirm or dispute interview 1 claims). However, NHS ethics procedures meant that a semistructured interview schedule was required that could not be deviated from, apart from when this was initiated by participants. However, I always had previous interviews in mind throughout data collection and informal refinement occurred, so that for example, if a participant of their own accord discussed an issue that I had understood from a previous interview to have explanatory potential, I would follow that line of enquiry. The focus groups on the other hand were used explicitly to refine programme theories and all focus groups were used to make sense of data form different sources. For example, the focus groups provided explanation as to why some patients with a predictable illness trajectory may not have had ACP; false hope had prevented this. Additionally the breadth of data collection meant that all data triangulated to refine understanding. For example, early analysis of GP practice data highlighted that preference discussions were significantly increasing, but the use of locality advance care plans was not. All patients and bereaved relatives interviewed reported not having locality advance care plans but all had engaged in preference discussions. In order to explain this, the focus groups highlighted that health care professionals did not like to use the locality advance care plan due to capacity assessment and time constraints. Consultation recordings also highlighted that sometimes professionals found it difficult to engage some patients in ACP altogether. The explanation (theory) that emerged from all these data strands was that often once a patient is ready to engage it is too late to use the locality advance care plan (due to time constraints and capacity issues) and this was further confirmed in the focus group. Thus programme theories were refined and tested through different data strands. There was also strength in exploring the explanatory potential of one idea (programme theory) in one data strand and then seeking explanation of this in other data strands, with the ICP founder or in PCQV. This process, at different stages of maturation, was then presented to health care professionals as collective sense makers during the focus groups to further refine theory development. This method of data collection and analysis did mean that at times decisions had to be made between different segments of evidence. Data was always chosen that had the strongest explanatory potential and was corroborated by alternative data sources, the ICPs founder, or the PCQVs.
Finally, the interviews and focus groups provided data on interesting aspects of care which have not been presented here. Although interesting, this data did not offer sufficient explanatory potential to be presented in this thesis. For example, one patient had concerns regarding social care and her dislike of the lack of continuity in the carers that came to her house every day. Another example was the fear of endof-life paraphernalia including the syringe driver, which was referred to by a bereaved relative and one of the nurses in a focus group. This data did not possess enough explanatory potential because it could not be related to a component part of the CMOC. For example, the fear of end-of-life paraphernalia was never linked to emergency admissions or increased likelihood to call an ambulance. Furthermore, the social taboo of death and dying is an important overarching context, which participants commented on. They felt that this taboo needed to be ‘broken down’ for progression in palliative care to occur. Although the social taboo of death and dying did not bear direct influence on the mechanisms detailed in this thesis, it is nevertheless an important context to all social programmes implemented in palliative care, including the ICP. Furthermore, some CMOC include contexts that are inherently underpinned by this taboo (for example, death being hidden in CMOC1, Chapter 8, p.263; litigious society in CMOC3, Chapter 8, p.285).
Future research Future research should focus on the role of active implementation strategies for new initiatives. This thesis supports Greenhalgh, Robert et al. (2004) in that active implementation may impact on the success of an intervention. Related to this, understanding how to identify and use champions’ energies in providing active implementation of interventions is pivotal, as the literature is currently sparse in this area (Greenhalgh, Robert et al. 2004). Further research about champions and peer opinions would generate a deeper understanding of how champions work and if they can be developed.
ACP could be reconceptualised and therefore require further research. The traditional focus of ACP on capacity has meant that other important aspects of it have been ignored. This thesis makes a case for ACP to be a tool to implement shared decision making, allowing the patient to face death and strengthen the relationships with their family. Furthermore, shared decision making needs to be investigated as a multi component communication strategy, as opposed to just the meeting of two experts. The role of shared decision making between health care professionals should be explored as a facilitator to MDT working. Following from this, the role of ACP should be researched. It could have a much broader role for all patients regardless of whether their capacity is likely to diminish. It could help patients face death, plan for it, discuss it with their family and generally take control of their care. This therefore supports the need to investigate shared decision making as a multi component communication strategy that also includes the patient’s family. Therefore the role of ACP in practice and its effect on patients requires further investigation.
Barriers to proactive care were identified throughout this thesis; all of which could impact on participation and shared decision making (Burt, Rick et al. 2013). This could have a negative effect on ACP. The barriers to proactive care (and thus the barriers to ACP) need to be explored and ways to overcome them need to be identified. The insights developed in this thesis could be used to develop more understanding of how, why and in what circumstances ACP and care planning in general lead to successful outcomes.
Conclusion This thesis occupies a currently important political space. The LCP has been subject to great media attention and review (BBC News 2013, Neuberger, Aaronovitch et al. 2013) and recent news articles have suggested that palliative care registers and advance care planning may be subject to the same scrutiny (Daily Mail 2013). A recent article in The Daily Mail (2013) suggests that palliative care registers and advance care planning are carried out to reduce health costs rather than plan for a good death. Unarguably, the NHS does need to save money, and deaths at home may be a piece in this economic jigsaw, especially as they are the preference of many patients. In the article, Dr. Anthony Cole, acting chairman of the Medical Ethics Alliance, acknowledged the need to save money, but also highlighted a danger of uninformed patient choice leading to lack of wanted medical support in their final days of life. Although uninformed patients can lead to uninformed decisions about end-of-life care, the patients interviewed in this study were very aware of their preferences and decisions. They appreciated the conversations they had had about end-of-life care with health care professionals and valued their clinical advice. Unfortunately, referring to palliative care registers as ‘death lists’ and advance care planning as ‘end-of-life advanced care plans to reduce cost’ will only result in a poor public understanding of these tools, when this research demonstrates that they are actually helping people to die in their preferred place.
Furthermore, some participants were acutely aware of this wider political and media context and expressed the need to open up a societal dialogue about death and dying.