Abstract

It is more than 30 years since the World Health Organization (WHO) called for a paradigm-shift to a model of primary healthcare: a vision of creating ‘Health for All’ through ‘putting people at the centre of health care’. 1 Since then, we have seen significant advances in health and healthcare. People are ‘healthier, wealthier and liver longer today than 30 years ago’. 2 Economic growth, increased resources, and rapid expansion of scientific knowledge and technological advancement have all contributed. 2 But we see growing concerns that the primary healthcare vision is being lost, with a worrying impact on our current systems of healthcare. 2 The burden of care on individual patients is increasing. 3 An excessive focus on disease, fragmentation of care and unregulated commercialization 2 has been linked to inefficiency, ineffectiveness and inequity. 4 All against a background of changing health needs, notably a rise in the burden of chronic, complex illness. 2,5 WHO calls for a revival and strengthening of the primary healthcare vision to meet modern needs, refocusing health services around people. 2 Generalism describes an approach to care which is person-, not disease-focused; continuous, not episodic; integrates biotechnical and biographical understanding of illness; and promotes health as a resource for living, and not an end in itself. 6 So can generalism help revive the primary healthcare vision?
Generalism in an era of evidence-based medicine
In the 30 years since the WHO declaration, 1 professional decision-making has come under increasing scrutiny. 7–9 Particular attention has focused on the way that practitioners use knowledge, or evidence, to inform the decisions that they make about patient care. Evidence-based medicine was developed as an educational tool supporting continuing professional development, 10 promoting the necessary skills among health practitioners to ensure that decisions were informed by the best scientific evidence. 10 Few reject a principle of evidence-informed decision-making. 11 And although authors have acknowledged the evidence base for evidence-based medicine is still limited, 12,13 widespread adoption of the principles has contributed to rising standards of quality, at least in specialist disease-focused care. 10,14
But concerns have been raised about limitations in the evidence-based medicine model, which may in turn contribute to problems with a person-centred generalist approach to decision-making. Many have criticized the hierarchy of ‘best evidence’ within evidence-based medicine; 11,12,15 being based on a set of assumptions about the nature of our world, and the most appropriate way to study it to reveal a ‘true’ answer, which are not universally shared. 8,12 Goldenberg 15 argues that the hierarchical model of evidence-based medicine is incompatible with person-centred care since it is unable to accommodate a subjective element.
But such criticisms can be countered by the requirement within evidence-based medicine for the use of professional judgement in the application of evidence to individual decision-making. 10,11 This emphasis on clinical judgement explicitly allows professionals to draw on other forms of evidence, integrating them with the patient's own account to make person-centred decisions. 17 Threats to person-centred care come not from evidence-based medicine per se, but from the way it has been implemented in modern healthcare settings.
From a hierarchy of evidence to a hierarchy of decisions: the mountaineering view of clinical practice
Problems arose when policymakers identified evidence-based medicine as a tool to address variation and quality in care. 18 The ideals of evidence-based medicine were adapted into what Harrison describes as scientific bureaucratic medicine: emphasizing the use of externally validated scientific research as the primary source of knowledge underpinning decisions about health and healthcare. 18 External knowledge experts following defined rules for interpretation of scientific research create definitive guidelines for clinical decision-making. 18 Performance incentives enhance adherence to guideline-defined care.
Scientific bureaucratic medicine translates the assumptions of positivist science directly into the applied setting. Just as the goal for the scientist is to reveal the true answer through rigorous study, scientific bureaucratic medicine expects the frontline practitioner to find the ‘true’ answer or make the ‘correct’ decision. The result is a hierarchy of decision-making, with the ‘best’ decisions being those that meet the ‘evidence’ from the ‘best’ studies (at the top of the hierarchy). Decision-making is seen as a technical process which can be assessed in terms of whether we have made the right or wrong decision; with correct decisions equated to ‘facts’ – only meaningful if they can be verified by science. 15 Scientific bureaucratic medicine has changed the way we understand and describe quality of practice, becoming an assumed standard of ‘best’ 8 or even ‘reasonable’ 15 practice.
In this model, we liken the practitioner to a mountaineer. Laden with a rucksack of ‘evidence-based medicine tools’, and supported by climbing aids such as communication skills and empathy, s/he struggles towards the summit and the ‘evidence beacon’ that is an (externally defined) ‘right’ answer. Quality of practice is predominantly defined in terms of ‘reaching the summit’: the capacity to correctly apply knowledge to find the true explanation and make the right disease-focused decision(s). Governance tools such as clinical audit assess how close we get to the summit.
The result is a normative framework dominated by a powerful beacon emanating from the summit; with ‘scientific evidence’ (or at least a limited part of it) dictating what we do and the decisions we make. The authors' conversations with frontline primary care practitioners reveal a sense of being ‘caught in the beam’, fearful to step outside the prescribed model of care even when aware of its limitations. Scientific bureaucratic medicine may have contributed to observable improvements in the quality of disease-focused care; 14 but has in turn undermined the delivery of person-centred, primary care. 6,7
Interpretive medicine: moving from mountaineering to exploration
The evidence-based medicine framework reflects the needs and values of the disease-focused secondary care setting in which it was developed. However, the generalist approach underpinning the primary healthcare vision defines the purpose and nature of healthcare differently. Illness is understood within the context of an individual's daily life, being a disruption to ‘health as a resource for living’. 19 Disease models offer one way to interpret and respond to this disruption; having demonstrable utility, but also limitations. 20 In the ‘swampy lowlands’ of general practice, patients present with undifferentiated problems that they believe to be health-related, where disease models are often insufficient or unable to explain individual illness experience. 7 The skill of the generalist practitioner is in supporting an interaction which explores a range of explanatory accounts and potential solutions; drawing on the patient's account, the professional's tacit experiential knowledge, and external scientific evidence about disease and illness. 7,21 The generalist practitioner must decide if and when it is useful to use a biomedical account of disease to understand and intervene with this individuals presented problem(s). Based on a critical analysis of the application of evidence-based medicine to generalist practice, Reeve proposed that generalist care is better described by an account of Interpretive Medicine: 7 the critical thoughtful professional use of a range of knowledge in the dynamic shared exploration and interpretation of individual illness experience, in order to support the inherent creative capacity of individuals in maintaining health as a resource for daily living.
Gabbay and le May's 21 ethnographic study of generalist decision-making describes this interpretive process ‘in action’. Decisions are informed by scientific evidence, but shaped by experience, interaction and tacit knowledge. 21 They refer to the complex act by which general practitioners integrate a range of knowledge in the interpretation of individual illness experience as the use of ‘knowledge-in-practice-in-context’, or ‘contextual adroitness’. The metaphor changes from mountaineering to exploration. Rather than an evidence hierarchy, interpretive generalist practice uses an ‘exploratory decision map’ (Figure 1). The exploratory decision map clarifies ‘person-centredness’ in the generalism model, being the ‘lens’ through which evidence is interpreted and decisions are evaluated. A generalist practitioner considers how evidence and a resulting decision will impact on this person living their life, rather than on this disease process.

The Exploratory Decision Map – five factors to consider in judging the exploratory decision-making*
Comparing scientific bureaucratic medicine and generalism: implications for understanding quality
There are important differences between scientific bureaucratic medicine and an interpretive generalist account (Table 1); the significance of which become apparent when we consider how we currently understand quality of care, and hence the implications for generalist practice.
Comparing and contrasting scientific bureaucratic medicine and generalist practice
Since ‘quality’ refers to the ‘degree of excellence of something’, 22 it is viewed as inherently desirable. Yet theoretical concepts of quality remain controversial; 23 while the task of distinguishing quality has driven some to breaking point. 24 In healthcare, quality was traditionally defined by health professionals by virtue of their control and ownership of specialist knowledge. 18,25 The introduction of new public management in the 1980s saw a shift away from professional control; with quality now defined in terms of quality of organizational systems. 25 The definition and delivery of quality healthcare became a management process; managed healthcare as synonymous with quality now a ‘self-evident truth’. 25
The result was a model which equates certainty and uniformity to quality. 26 Yet generalism is about uncertainty, complexity and difference. Reconciling these ideas challenges us to reconsider our understanding of quality and what it is we value in healthcare. For this we need to understand the ‘benefit’ and ‘risks’ of uncertainty and a dynamic response, in order to support a judgement of value.
An evidence base for generalist practice
There are both theoretical, 3,4,7 and some empirical reasons, 27 to think that a generalist approach could support a primary healthcare vision, and thus offer value to modern healthcare approaches. But generalism refers to a (professional) philosophy of practice; 28 it lacks an evidence base supporting it. Generalism focuses on individual decision-making; the currently valued scientific bureaucratic medicine approach considers effectiveness and efficiency of healthcare at a population level. We need to evaluate whether generalist care can show a benefit at the population level; to investigate the impact of an interpretive, generalist approach on the primary healthcare vision.
To date, primary care impact has been evaluated through comparisons of national health outcomes in countries recognized as being more or less primary care-oriented. 29 As ‘boundaries’ blur between primary and secondary care models, with more disease-focused ‘secondary care’ delivered in the primary care setting, 5,29 such comparisons become increasingly complex to interpret.
New methodological approaches to evaluate generalist care are needed: 5 which focus not only on individual impact, but also on population effects. We propose the use of the complex interventions framework to consider the impact of generalist healthcare. This approach requires us to consider how to recognize care when it is happening, and also the ‘outcome’ of care (how we know when it has been done well).
To date, outcome measures for general practice have focused on the quality of consultation processes; assessing the impact of interpersonal skills and empathy on immediate outcomes such as a sense of empowerment. 5 Health as a resource for living, and the primary healthcare vision, consider impact beyond a single interaction but rather the impact of a healthcare approach on a way of being. We are currently exploring creative capacity 7 as a potential new measure of ‘health as a resource for living’; and therefore a tool to evaluate the impact of generalist care on the primary healthcare vision.
Our work to develop tools which can recognize, and thus distinguish generalist and non-generalist care, draws on Gabbay and le May's 21 existing work in this area. We propose to start with observational studies to describe the essential elements of generalist care: what they are, when they are used, and how we can distinguish them from non-generalist care. In this way we seek to translate their observational account into a framework which will support the evaluation of care. Comparing the impact of generalist and non-generalist care on a range of outcomes, including health as a resource for living, offers data on the value of generalist care (for example, impact on the primary healthcare vision). Examining both the added value and risks of generalist care addresses questions about who needs generalist care, and when.
Practice-based generation of evidence 11
The generation of interpreted knowledge is part of the daily work of generalist decision-making. In Table 2, we describe the use of the exploratory decision map in starting to articulate and evaluate the process and impact of generalist decision-making. Capturing this practice-based evidence of usual interpretive care 11 through action learning in context offers a potentially valuable new source of evidence about process and impact. Indeed, the King's Fund propose that such ‘peer scrutiny’ is central to quality improvement in general practice. 5
The exploratory decision map in practice
But we also highlight the difficulties of a more dynamic and uncertain model of decision-making: notably the resources, including time, 5 needed to support the reflective, learning process inherent in the exploratory decision map. We suggest a role for academic primary care in supporting this reflective process, and thus enhancing a bi-directional flow of knowledge, evidence and learning between research and applied settings.
Conclusions
Generalism resonates with, and may help reinvigorate, the primary healthcare vision. But first we must address the gaps in the generalist account: ‘demonstrat[ing] it is doing things differently and better’. 5 We have identified key challenges: in addressing the evidence gaps which limit applicability of the generalist approach in the managed quality healthcare context; and in providing adequate support for evaluation and learning from complex decision-making by practitioners at the coal face. We call for closer working between academics and clinicians to make evidence part of practice, as well as practice part of evidence. 11 Our proposals resonate with the King's Fund 5 appeal for new work to understand and promote quality in general practice. We seek to evolve generalism from a professional philosophy of practice 28 to an evidence-based model of care. We suggest that it is work to establish the value and utility of the generalist approach which offers most scope to revive the primary healthcare vision.
DECLARATIONS
Competing interests
None declared
Funding
JR is funded by an NIHR Clinician Scientist award supporting development of a body of work looking at generalist solutions for complex problems; CFD is her mentor for the NIHR award; GI has an NIHR Doctoral Research Fellowship
Ethical approval
Not applicable
Guarantor
JR
Contributorship
JR led the writing of the paper; with support from NIHR funding and mentorship from CFD, she has developed and published a conceptual account of Interpretive Medicine underpinning generalist practice; this paper represents work to translate the model into clinical and academic contexts and has been developed through discussion between all authors; GI contributed to the workshop described in Table 2; JR wrote the first draft of the paper based on discussions with both co-authors; all three authors contributed to revising and refining the account and have approved the final version
Acknowledgements
The authors thank Roger Harrison, University of Manchester, and John Gabbay and Andrée le May, University of Southampton, for helpful conversations related to these topics
