«Citation: Dalkin, Sonia (2014) The Realist Evaluation of a Palliative Integrated Care Pathway in Primary Care: What Works, For Whom and in What ...»
Epistemology considers the nature of the relationship between the knower or the wouldbe knower and what can be known (Guba and Lincoln 1994). The answer to the epistemological question is always constrained by the answer to the ontological question. What can be known about any social programme is not definitive. For example, the heart failure service described by Smith (2012) may not achieve the same results in a different locality and is thus not a definitive finding. This is due to the complexity of social reality meaning that people have different understanding, values and meanings and thus a social programme is never exactly replicated. However a decision can be made between opposing explanatory theories (the heart failure service may have worked due to improved communication between the clinical team, as Smith (2012) suggest, however they do not unpack explicitly how improved communication has led to the outcomes achieved). Realist evaluation begins with a theory of causal explanation known as a CMOC, which is based on the idea of the generative mechanism exposed further here (Pawson and Tilley 1997). The researcher searches for causal powers within objects or agents or structures under investigation and expresses them in terms of CMOC (Pawson 2006). It requires complex and systematic understanding of causal powers which takes into account the underlying constructs that connect two events, and the context in which that relationship occurs (Pawson 2006).
Put simply, what causes something to happen has nothing to do with the number of times it is observed happening (Sayer 2000). Therefore gathering data on regularities is misguided, however these may suggest where to look for causal mechanisms (Sayer 2000), one of the integral concepts in realist evaluation.
Individuals may be aware of patterns of regularities into which their lives are shaped, the choices that channel their activities and the wider social forces that limit their opportunities (Pawson and Tilley 1997). This can result in individuals wishing to change the pattern. This change may or may not happen as the individual may or may not have the resources to do so, or their efforts may be overcome by an opposing group who have more resources. Furthermore, individuals have incomplete knowledge of the contextual conditions in which they reside and these contextual conditions may limit their actions, and the proposed change mechanism itself may have unanticipated consequences (Pawson and Tilley 1997). For example, recently medication availability has been affected by locality in the UK. Patients may wish to have a course of medication but cannot due to this contextual factor. Social programmes or interventions are an attempt to change the current regularity in a domain through generative mechanisms.
Methodology considers how the would-be knower can go about finding whatever he or she thinks can be known (Guba and Lincoln 1994). This question is constrained by the answer given to the first two questions. Realist evaluation (Pawson and Tilley 1997) has been created for researchers to investigate the world from a realist perspective and focuses on the development and refinement of CMOCs. Realism and realist evaluation have further explanatory potential in relation to death and dying and palliative care ICPs. They can offer an understanding that is grounded in the locality but provides more macro knowledge about death and dying and palliative care ICPs. It opens up the ‘black box’ and links outcomes to explanations that are mindful of contextual factors and the individuals who implement and receive social programmes. Previously the management of death has been considered in a very causal way; improved communication will lead to more home deaths (Smith 2012). Realist evaluation allows for a deeper understanding that embraces human volition and the complex social systems in which people reside.
Realist evaluation of this palliative care ICP will therefore provide novel insights.
Theoretical framework: realist evaluation The conceptual approach in this study is realist evaluation (Dalkin, Jones et al. 2012) (Appendix 2), as it enables in-depth analysis of interventions through the means of realist programme theories, embracing both qualitative and quantitative research (Pawson, 2013). Programme theories have been developed and iteratively refined for each section of analysis. Analysis will highlight what Contexts (C) need to be present so that the relevant mechanisms (M) are likely to be triggered to cause observed outcomes (O). The purpose of this is to generate understanding of how resources provided by the ICP interact with contexts to trigger the necessary mechanisms to produce positive outcomes. Social programmes, realist logic, and CMOCs are explained in detail below.
Social Programmes To understand realist evaluation, an understanding of social programmes or interventions must be developed. Social programmes are active, they do not operate in laboratories, they are affected by contexts which are changeable and thus although two social programmes may have the same name, they will never behave in exactly the same way. The ICP can and will thus be considered as social programmes in this thesis.
‘Realistic Evaluation’ (Pawson and Tilley 1997) is the main text for realist evaluation.
The subject matter of a realist evaluator is a social programme (intervention), otherwise understood as social systems. They consist of the complex interactions between individuals and institutions and of micro and macro social processes (Pawson and Tilley 1997). A realist approach states that social programmes are theories incarnate; this is plural as one social programme is likely to have several theories. Social programmes are delivered under the hypothesis that if the programme is delivered in a certain way it will improve outcomes (Pawson 2006). This means that whenever a social programme is implemented, it has an underlying theory about what might cause change, which is being tested. However, this theory is not always explicit. It is the role of the realist evaluator to make these theories explicit, and ensure that the right questions are asked of the data. For example, the underlying theory in ACP may be that planning for a good death results in more patients dying in their location of choice.
Interventions are implemented into existing social systems that are believed to account for a particular problem, such as difficulty talking about death or issues identifying palliative care patients. The fresh input the social programme gives to the system is expected to improve patterns of behaviour, events or conditions via changing and rebalancing the system (Pawson 2006). The underlying theories of the ICP are discussed later in this chapter (Research questions section, p.88). In order to describe a social system or programme, realists use three predominant concepts: context, mechanism and outcome. These three concepts produce a generative mechanism to explain causality.
In realist evaluation, mechanisms work in partnership with context to lead to outcomes in a causal way (Pawson 2006). In other words, the relationship between causal mechanisms and their effects (outcomes) is not fixed, but contingent on context (Sayer 1984). Thus realist approaches make explicit use of broader insights in order to explain the impacts of interventions in context, as will be done in the analysis in this thesis. All social programmes/interventions are introduced into pre-existing social contexts, therefore there is sometimes a struggle for them to prevail in these contextual conditions, hence, they are contingent. It is impossible to establish a straightforward relationship between intervention and outcome without identifying pre-existing contexts. Each social programme implemented has a great number of different contextual constraints and facilitators and the interrelationships, institutions and structures in which it is embedded all affect its success (Pawson 2006). Context refers to not just the physical, but to the culture and drivers (professional cultures, power dynamics within GP practices, cost effectiveness, disease specific clinical reasoning), institutional features (patient list sizes in GP practices, shared nursing teams, staffing levels in care homes) and ethical issues (equality of care, capacity to make a decision).
Context works by constraining the choices of stakeholders in a social programme.
Stakeholders in the ICP can be programme leaders, programme policy makers, clinical staff, social care staff, voluntary organisations, palliative care patients and bereaved relatives. The subjects of a social programme are always faced with a choice about whether to participate (Pawson 2006). Subjects have different pre-developed or pregiven characteristics that leave some well prepared and some badly prepared for the programme theory, resulting in varying success for individuals, whether this is those implementing the social programme or those receiving it. They also have different preexisting relationships, which means that some are well placed and some are ill placed to use the opportunities provided by the social programme (Pawson 2006). For example, the ICP has a focus on being patient centred. However, the policy context encouraging proactivity and patient centredness may compete with pre-existing systems.
Alternatively, from the patients’ point of view, a patient who is less familiar with their GP may be less likely to engage with the ICP (through, for example, ACP), than a patient who has built a relationship with their GP over their life-time. Additionally, on a more macro level, some GP practices may have more flexible systems that make explicit allowances for patient involvement. On an even higher level, societal taboos about death and dying can be a prominent context, meaning that death is often not discussed and is shrouded in mystery. Currently there is an effort to change this taboo through the National End of Life Care Strategy (Department of Health 2008) and documentation such as Deciding Right (NHS North East 2012). Although these documents are tailored for, and aimed at, health care professionals, they encourage early discussions about death and dying with patients. This social change in the way GPs regard and discuss death and dying is a social change that will have effects on any social programme in palliative and end-of-life care. Despite this, it could be that some GPs and patients may be reluctant to talk about death and dying due to this societal taboo. This would make the resources of the social programme, such as ACP, very difficult to use.
The context thus has clear implications for the successes and failures resulting from a social programme.
It is not expected that massive contextual change will occur during a programme or as a result of a programme; a social programme does not aspire to cause the downfall of existing cultural and social order (Pawson and Tilley 1997). However, an accommodating context is crucial for implementation of a successful social programme that aims to change behaviour. If a context is inhospitable, the programme mechanisms are unlikely to be activated and therefore will not combat or neutralise the original problem mechanisms that were sustaining the bad outcome pattern. A physical analogy of this could refer to gunpowder; a spark causes an explosion. However, there will be no explosion if the conditions are not right – damp, insufficient gunpowder, inadequately compact, no oxygen present, or no heat applied (Pawson and Tilley 1997). Thus in research it is necessary to identify the social and cultural conditions necessary for change mechanisms to operate.