Interpretive Medicine

Abstract

Patient-centredness is a core value of general practice; it is defined as the interpersonal processes that support the holistic care of individuals. To date, efforts to demonstrate their relationship to patient outcomes have been disappointing, whilst some studies suggest values may be more rhetoric than reality. Contextual issues influence the quality of patient-centred consultations, impacting on outcomes. The legitimate use of knowledge, or evidence, is a defining aspect of modern practice, and has implications for patient-centredness.

Based on a critical review of the literature, on my own empirical research, and on reflections from my clinical practice, I critique current models of the use of knowledge in supporting individualised care. Evidence-Based Medicine (EBM), and its implementation within health policy as Scientific Bureaucratic Medicine (SBM), define best evidence in terms of an epistemological emphasis on scientific knowledge over clinical experience. It provides objective knowledge of disease, including quantitative estimates of the certainty of that knowledge. Whilst arguably appropriate for secondary care, involving episodic care of selected populations referred in for specialist diagnosis and treatment of disease, application to general practice can be questioned given the complex, dynamic and uncertain nature of much of the illness that is treated.

I propose that general practice is better described by a model of Interpretive Medicine (IM): the critical, thoughtful, professional use of an appropriate range of knowledges in the dynamic, shared exploration and interpretation of individual illness experience, in order to support the creative capacity of individuals in maintaining their daily lives. Whilst the generation of interpreted knowledge is an essential part of daily general practice, the profession does not have an adequate framework by which this activity can be externally judged to have been done well. Drawing on theory related to the recognition of quality in interpretation and knowledge generation within the qualitative research field, I propose a framework by which to evaluate the quality of knowledge generated within generalist, interpretive clinical practice. I describe three priorities for research in developing this model further, which will strengthen and preserve core elements of the discipline of general practice, and thus promote and support the health needs of the public.

Preface

There have been many seminal papers in general practice over the years, written by well-known practitioners such as Balint, Fry, Pendleton and Neighbour – to mention but a few. This paper by Joanne Reeve will become one of these. It will provide an opportunity to reflect on how general practice is changing as a result of the influence of evidence-based medicine, with research mainly grounded in secondary care, guidelines and protocols.

To quote Reeve:

I acknowledge the power of positivist science to describe and predict our world. However, I also ask that its limitations be recognised, and that general practitioners be not only ‘allowed’ but also supported/valued for their capacity to use a range of knowledge in the individualised assessment and management of their patients.

We need to consider the application of protocols of care on the shared decision-making process with individuals based on their illness experience, and the role of the personal primary care physician who has gained the privilege to learn more and understand partially the often difficult issues. This paper helps us to identify the essential role of GPs in modern primary care, which goes beyond protocols and so differentiates their role from nurses and other allied healthcare professionals.

Patients are more than disease, and GPs can make decisions – for this is their skill – in complex situations where the boundaries relating to possible diseases are blurred. This is important when trying to establish and manage the resultant illness being experienced. Imparting an ongoing narrative in terms of the patient's physical, psychological, social and spiritual circumstances raises the age-old debate of the nature of the ‘self’. I commend this paper, and ask you to read and re-read, reflect on and recommend it to your colleagues and trainees.

Dr Rodger Charlton, Editor, RCGP Publications, October 2009

Introduction

The discipline of general practice values the provision of personal, holistic, lifelong, generalist care to individuals within the community it serves.^1,2,3 Models of practice describe personal attributes needed to deliver this care, focusing on the nature of the consultation and interpersonal skills.^4,5,6 There have long been concerns that values may be more rhetoric than reality;^7,8,9,10 however, there is now evidence that wider social changes related to defining the best use of knowledge in practice may be undermining this approach to care. ^{8,11,12,13,14} In this paper I will argue that what is missing from current accounts of general practice is a theoretical (rather than value-based) account of why personal care is of worth and why it should therefore be protected and preserved: an account of the nature of the ‘person’ that general practice is seeking to support, and therefore of the knowledge needed to do so.

I start by critiquing current models of understanding knowledge in practice when applied to the discipline of general practice. I identify a theoretical account of the person based on critical accounts of illness in the literature, and highlight its relevance to general practice. I thus propose a new model of Interpretive Medicine (IM) as a defining purpose of general practice, outlining the core components necessary to strengthen the continued development and application of this way of working, and hence contribute to the sustainable delivery of primary medical care within local communities. I conclude by identifying the clinical and research priorities to operationalise this model of care within policy and practice.

The nature of general practice

The legitimate use of knowledge, or evidence, is a defining aspect of modern clinical practice. Current accounts of the best use of knowledge in clinical practice, including the Evidence-Based Medicine (EBM) ¹⁵ movement and the subsequent development of Scientific Bureaucratic Medicine (SBM)^16,17 (see later), derive from secondary care models of clinical practice. Indeed, much of the evidence they promote derives from studies of secondary care patients. In order to understand the limitations of these accounts for general practice, the differences between primary and secondary care approaches to understanding and treating illness must first be considered. Essentially, secondary care delivers a disease-centred, specialist model of care; primary care an illness-focused, generalist model of care.^2,3

Patients access secondary care when their primary care clinician suspects a pathological cause for the person's illness experience that cannot be diagnosed or treated in primary care alone. The patient is referred to a specialist, who will treat identified disease. Once the patient is cured or stabilised, he or she is discharged. If no disease status is identified, the patient may be referred on to another specialist, but is again ultimately discharged back to his or her GP. The pathological, disease model assumes illness results from a biological deviation from ‘normal’, and that correction of the anomaly will restore health.¹⁸ The model lends itself to the development of objective measures of disease status, and hence quantification of the effectiveness of intervention: a pattern of knowledge use described within the current models of EBM and SBM.

In contrast, general practice treats a broader range of illness experience, which is often harder to categorise or define. In the ‘swampy lowlands’ of general practice.^19,.²⁰ patients present with undifferentiated problems² that they believe to be health related,²¹ where disease models are often inadequate to explain individual illness experience,^18,22,23,24 and where illness is often complex, dynamic and uncertain. ^{25,26,27,28,29}

GPs manage illness even when no certain pathology is identified. Primary care recognises that, just as illness can cause social disruption,²⁹ so the disruptive impact of living in the social world can cause illness.^27,30 Much of GPs' time is spent dealing with ‘indistinct’ illness – stress and distress, tiredness, pain – where the disease model is inadequate^22,23,31 since no clear pathology or causal chain can be identified, although social factors are often implicated. Patients cannot be discharged, so GPs must find ways to manage these problems.

Some patients will have identifiable disease, but also describe needing help with the wider personal and social consequences of the pathology. The need for these supportive aspects of care is increasingly recognised across a health service that seeks to be more patient focused. It is a core part of the GP's role because GPs provide ongoing, rather than episodic, care, see patients more regularly, and do not discharge them.^1,2 Yet these aspects of care are less clearly defined within the world of objective, certain and measurable knowledge.^8,9,11

At other times the challenge for GPs is not a lack of pathology, but dealing with the complexity of multiple pathology, or dynamic, changing complex illness.^{25,26,27,32,33} As generalists,^2,3 GPs manage all aspects of a person's illness experience; it may not be possible or desirable to manage each disease as would be done for a single pathology. In addition to dealing with the epistemological uncertainty that comes in seeking to apply ‘certain’ knowledge derived from the study of populations to understand this individual,³⁴ GPs face the challenge of integrating different bodies of knowledge to understand the complexity of illness.³⁵

During the interaction between practitioner and patient, there is potential for at least three illness stories to emerge: the patient's own experiences, perceptions and analysis of his or her illness; a pathological account of disease as described in the biomedical scientific literature; and the practitioner's account of the patient's illness based on an understanding of pathology and normal function, professional knowledge and values, and experience of illness in previous patients. Salmon³⁶ noted that ‘practitioners commonly fashion explanations which integrate patient and practitioner conceptual accounts of presented health problems’. How each of those accounts is heard and managed has long been a focus of concern for the profession,^37,38,39,40 but has more recently become a focus of interest at a wider organisational level, related to efforts to define and measure quality in the publicly funded primary care sector. The interpretive role of GPs in integrating biological and biographical accounts of illness³ to develop individualised explanations, assessments and plans for intervention is ‘not greatly valued amongst practitioners or medical curricula’,³⁶ or indeed by health policy and management.

Current models of knowledge and practice

The failure to value this interpretive aspect of practice is fuelled by social changes over the last 20–30 years in what counts as legitimate knowledge within health and health care.^16,17 These changes reflect the wider reorganisation of the public sector associated with the introduction of what has been described as ‘New Public Management’.

New Public Management (NPM) is a term used to describe a wave of public sector reforms, first seen in the UK in the 1980s, which have sought to introduce ‘market-type values’ into publicly funded services.^41,42,43 Changes aimed to introduce principles of management and accountability derived from business models into publicly funded organisations, making the public sector ‘less distinctive as a unit from the private sector’.⁴² It was seen as a necessary response to a number of social changes, including: the economic crises of the 1970s; the rise in radical consumerism and need to control public expectations; heightened perceptions and decreased tolerance of uncertainty and risk; concerns over inequalities in health and health care; and erosion of public confidence in the individual professional's ability to define and practise his or her own work.^41,43 NPM brought a new emphasis on quality, efficiency and accountability to publicly funded services, including health care.⁴¹ It also saw a shift in responsibility for defining and controlling services, away from the ‘opinions’ of professionals, to the ‘objective’ assessment of organisational systems. ^16,17,41,44

Harrison^16,17 described the impact of these changes on the use of knowledge in medical practice, being an observable shift in the dominant model of medical care away from reflective practice based on professional consensus about the effectiveness of medical care to the greater use of externally generated, objective research evidence. This essentially describes the introduction of EBM,¹⁵ where ‘best’ evidence is defined by an epistemological emphasis on statistical analysis of empirical observations of defined disease within populations. Evidence is derived from the scientific method rather than clinical experience, providing objective knowledge including quantitative estimates of the ‘certainty’ of that knowledge.⁴⁵ A hierarchy of knowledge values this objective account over the ‘opinion’ of a clinician, or indeed a patient. Professional knowledge and skills are still needed in the ‘conscientious, explicit and judicious application of best evidence in making decisions about the care of individual patients’,¹⁵ namely the interpretation of the applicability of external evidence to individual patients. However, no theoretical framework was described to support this interpretation, a gap that perhaps contributes to the development of what Harrison has described as Scientific Bureaucratic Medicine (SBM).^16,17

SBM^16,17 was an extension of these ideas about quality within the context of NPM. Like EBM, SBM regards the primary source of ‘valid knowledge about the effectiveness of medicine’ as that based on objective research, conducted by experts and judged by its proximity to ‘truth’, rather than the opinion of professionals. However, it is in the use of this knowledge that EBM and SBM differ. EBM acknowledges the legitimacy of professional judgement in the interpretation and application of this knowledge, in deciding when and where to apply research findings to guide decision making in the assessment and management of patients. In contrast, SBM hands the responsibility for interpreting and judging this externally derived knowledge to ‘external knowledge management experts’. Contending that working clinicians ‘don't have the time to discover and collate the information themselves’, ^16,17 SBM creates a role for external experts to distil findings for professionals into ‘algorithms for action’ in the form of diagnostic tools, decision aids and clinical guidelines. Management systems can then be designed to support consistent implementation of tools. SBM is a model that supports organisational needs for consistency and predictability in health care (over the variability/uncertainty of professional practice). Responsibility for decision making shifts from the individual professional to systems or protocols; trust in individual professionals is replaced by confidence in (quality) systems.¹⁷

Both EBM and SBM define best evidence in terms of an epistemological emphasis on statistical analysis of empirical observations of defined disease within populations. They are undertaken by the scientific method rather than being based on clinical experience, and provide objective knowledge including quantitative estimates of the ‘certainty’ of that knowledge.⁴⁶ Whilst arguably appropriate for secondary care, involving episodic care of selected populations referred in for specialist diagnosis and treatment of disease, application to general practice can be questioned given the complex, dynamic and uncertain nature of the discipline.

Critiquing SB/EBM: May's Normalisation Process Model

Further insights into the strengths and limitations of EBM and SBM when applied to general practice may be gained by the application of May's Normalisation Process Model 47 to understanding changes in the use of knowledge in practice.

The work of May et al. seeks to understand how new technologies become embedded into routine health care. The Normalisation Process Model explores how innovation becomes ‘taken for granted’ in everyday work. It provides a framework by which we can evaluate the process of change, and also judge between different complex interventions. In particular, the work recognises the changes resulting from the increased organisation and regulation of health care: the processes that take place in a clinical consultation no longer being governed solely by the clinician, but subject to the control of ‘corporate impulses [emphasising] effective throughput and outcomes’.⁴⁸ May et al. define complex interventions as any deliberate attempt to introduce new, or modify existing, patterns of collective action in health care.⁴⁷ The use of knowledge in practice, and specifically the introduction of EBM and SBM, can therefore be viewed as one such complex intervention.

May et al.⁴⁷ argued that, for a complex intervention to be normalised, or become integrated, into everyday working practice, it must meet four propositions (see Box 1). We can use this framework to critically explore the described changes in the use of knowledge within general practice. Applying each proposition to the use of Scientific Bureaucratic or EBM (hereafter denoted as SB/EBM) within general practice raises significant doubts as to the appropriateness and viability of these concepts as the basis for generalist primary care.

Box 1: May's four propositions underpinning normalisation⁴⁷

• Interactional workability: Does it help patient and practitioner accomplish congruence (of purpose and action) and shared outcomes?

• Relational integration: Does it improve professional accountability and patient confidence?

• Skill set workability: Does it affect allocation and performance of tasks within existing resources and structures?

• Contextual integration: Are resources available for its implementation and can it be realised?

Interactional workability

Interactional workability relates to what is an essential part of general practice, namely the identification of a shared understanding about the nature of the problem, possible actions, and the goals of intervention.^2,3,4,6 Interactional workability may be threatened by the assumption of a dominant body of knowledge within the SB/EBM approaches, especially in the context of illness that is uncertain, not adequately described by pathology, or complex. Specifically, the rigid application of SB/EBM may contribute to iatrogenic harm through failing to support delivery of shared outcomes.

At a societal level, the failure of pathological models to explain individual illness may have contributed to the medicalisation of human illness experience. In order to fit current models of health care, and thus gain ‘legitimacy’ and treatment, ‘altered states’ such as tiredness, stress, or distress are re-conceptualised as ‘diseases’ (e.g. chronic fatigue syndrome or mental health problems). Illness experience, or need, is reconceptualised to fit the current model of care, rather than ensure care fits need. Yet this raises concerns about extending the medical gaze, with power shifted from the sufferer to the ‘state’. More recently, there have been concerns about the ‘widening [of] the boundaries of treatable illness in order to expand markets for those who sell and deliver treatments’; what Moynihan et al. describe as ‘disease-mongering’. ⁴⁹ The commercialisation of health care may be an unforeseen consequence of the SB/EBM approach.

For the individual, there is potential for harm as a result of both undertreatment and iatrogenic escalation of illness. The rigid application of the disease model as epitomised within SB/EBM may lead to undertreatment of those who do not meet diagnostic criteria but still have need.^22,23 Furthermore, if no ‘true’ disease is identified, the evidence-based practitioner may legitimately seek to ‘normalise’ patients' experience: demonstrating the absence of pathology with the implications that the individual should seek an alternative solution to his or her problems, perhaps outside the health service (for example looking at lifestyle or social circumstances). Yet Dowrick et al.⁵⁰ identified the potential for harm in this approach in patients with so-called medically unexplained symptoms (persistent illness experience in which no physical pathology can be demonstrated). In seeking to reassure patients through demonstrating an absence of pathology, clinicians directly contributed to an escalation of reported symptoms and debility, this being the patients' response to not having their needs met.

A ‘correct’ diagnosis of true disease (within a biomedical framework) may also cause harm. It has long been recognised that diagnoses may be resisted, especially when associated with stigma.^51,52,53 My recent work looking at distress in people with terminal illness highlighted the need to recognise that professional accounts of illness may both support and undermine an individual's sense of him or herself and the world, and hence people's efforts to continue living their everyday lives.⁵⁴ A disease diagnosis that is incongruent with patients' perceptions of themselves living their lives may undermine self-capacity and thus self-management⁵⁵ of daily life. I highlighted the need to understand the consultation not as a technical process of identifying a ‘true’ diagnosis, but as a social interaction with potential consequences for the individual beyond the identification and selection of an evidence-based intervention.⁵⁴

A pathological account, especially in conjunction with the assumptions of a hierarchy of knowledge, may also lead to harm through overtreatment. In a modern world where ageing populations and altered lifestyles make multiple pathology a common occurrence, the biggest risk in rigid application of diagnostic and treatment protocols for each separate disease is in multiple prescribing. In addition to the costs incurred and risk of adverse drug reactions,²² multiple prescribing in elderly patients creates risks of falls, with consequent risk of loss of mobility, confidence, independence, and even loss of life. Application of a disease account may be ‘correct’, but may not support shared needs in that it may not best support quality of life for the individual patient. There is also risk in applying epidemiological/population-level assessments of risk to individuals, with the potential for inappropriate anxiety and hyper-surveillance, as highlighted within the medicalisation literature.^49,56,57,

Yet it must be acknowledged that the disease model can have demonstrable utility in identifying need, predicting prognosis,^58,59 and thus supporting both patient and practitioner in identifying and addressing need. Much of the illness presented to general practice can (eventually) be described and understood using the disease model. However, GPs also recognise that lay and professional accounts of illness differ.⁶⁰ Professional models of practice value both accounts: seeking to integrate perspectives to achieve congruence and shared outcomes,^2,3,36 and recognising the limitations of the disease model. Whilst EBM allows for some professional judgement, SBM is less flexible.

Individual practitioners may seek to use professional judgement to establish shared priorities and aims with individual patients, but the failure to recognise this process within current quality criteria, and thus priorities for care, means this role is neither valued nor adequately supported. The very existence of a hierarchy of knowledge, together with the epistemological emphasis within it, may inhibit establishment of a shared sense of purpose and outcomes, and thus limit interactional workability. General practice needs a more flexible framework for understanding and valuing knowledge in practice.

Relational integration

Relational integration looks at issues of professional accountability and patient confidence, both important concepts in modern health care. The shift to SB/EBM could be expected to contribute to relational integration through improving professional accountability. There is an ongoing requirement for clinicians to demonstrate, and be held accountable for, consideration of the strengths and limitations of the latest external evidence in interpreting individual illness experience. ⁶¹ Within cultures that value scientific knowledge, the emphasis on SB/EBM could also be expected to improve patient confidence. Yet there are a number of problems that can be identified.

Accounts of scientific evidence that value truth above all other markers of quality may contribute to an expectation of ‘certainty’. This creates potential for harm in the face of uncertainty: patient dissatisfaction when there is no evidence that can answer his or her health problem;³¹ and even illness escalation when the evidence that is presented ‘doesn't apply to me’.⁵⁰ SB/EBM places greater power (or at least maintains power) in the hands of the clinician (over the patient) since the clinician is the arbiter of the scientific account and thus ‘best knowledge’. Yet both also create greater uncertainty for the clinician over the use of tacit, professional knowledge in the interpretation of individual illness; especially if the individualised care promoted by the application of professional values is in conflict with the protocol-driven care. Rather than aiding clinical practice, they potentially make it more difficult. There are already suggestions that, in the UK, guidelines from NICE (the external body responsible for distilling and disseminating ‘best’ evidence with regard to clinical practice) may form a new ‘normative framework’.⁶² At present, guidelines are offered as a summary and evaluation of evidence, providing suggestions for course of action. Individual clinicians may still choose to adopt an alternative approach if they believe it is more appropriate for their individual patient. However, there is concern that in the future, for example for the purposes of litigation, it may be increasingly difficult for a doctor to defend practice that does not adhere to NICE guidelines since a court may refuse to regard the deviation as ‘logically defensible’.⁶² With greater pressure on clinicians to ‘adhere to protocols’, patients may also experience greater difficulty in exerting their own wishes and accounts.

Clinicians have a wealth of experience and expertise in dealing with uncertainty,^34,63 but there are risks that over-emphasising certainty may undermine clinicians in this important aspect of their job. If the interpretive skills of clinicians in dealing with situations where the evidence does not fit are not valued, then the clinician is left working in a vacuum of professional accountability. SBM places responsibility for knowledge generation on the external experts who have developed the protocol, the clinicians' role being to deliver it. This may disempower clinicians and, paradoxically, not improve accountability. The consultation is the cornerstone of general practice.^4,6 Whilst there has always been a ‘third voice’ within the consultation (namely the pathological account), the higher status granted to that third voice within the SB/EBM model may, paradoxically, have a negative impact on relational workability.

Skill set workability

The final two questions refer to organisational issues. Skill set workability looks at whether existing resources, including organisational structures, are adequate to deliver the new ‘intervention’. The move to SB/EBM has implications for the allocation and performance of tasks within general practice.

SBM removes power from both the clinician and patient in terms of decision making in deciding priorities for care (e.g. in diverting focus away from the care of individuals to managing disease-focused systems of care within the incentivised approach of the new GP contract). There is evidence of a negative effect on existing professional morale.^41,60 The expanding number of protocols of care within SBM, together with the administrative need to manage delivery and demonstration of quality care, has created a need for more staff. This has contributed to the creation of new professional roles, for example nurse practitioners to meet the costs and resource demands of these models of care, and a new breed of GPs managing complex ‘businesses’. The development of new roles and structures indicates that existing resources were unable to deliver the changes (at least at the required speed). Whether the current changes to adapt to the new model are sustainable remains to be seen.

Contextual integration

This leads in to the final question of contextual integration: Can we see the (sustained) availability of resources to realise the changes of these new models of knowledge and practice? EBM fits well in to the disease model of secondary care. However, there are concerns about how well it fits a primary care, generalist, interpretive model that deals with undifferentiated illness. SBM was intended to improve rationing and predictability of the NHS, but there are now concerns that ripple effects are leading to spiralling costs, with a greater proportion of the population destined for medical intervention.^64,65 The emphasis on certain knowledge and measurable change has promoted a focus on disease, and a ‘downstream’ emphasis on treatment. Risk factors have been re-conceptualised as ‘disease’ (e.g. hypertension and even smoking)⁴⁹ in order that they can be identified and treated within the SB/EBM approach. However, with ever greater numbers of people on medication to reduce risk – bringing new complications such as the need for drug monitoring, the impact of polypharmacy and drug interactions, and the opportunity costs of diverting attention/perceived need to address lifestyle and social factors – there are spiralling costs. This raises questions about the need to revert back to a more upstream approach to promote a sustainable model of delivery. SBM neglects to ‘treat’ the less easily defined, wider social causes of illness and disease, an ongoing concern for public health practitioners.^66,67

Implications for general practice: models of the consultation

Both EBM and SBM define a hierarchy of knowledge, valuing knowledge derived from a positivist epistemological perspective over others.^45,46 By implication, the consultation or clinical interview becomes viewed as a process in which the clinician seeks to uncover ‘true facts’ about the individual's illness. Good clinical communication skills support establishment of a therapeutic relationship, which is valued by both the profession and patients,^68,69,70 but also within SB/EBM as necessary to elicit the required information. External assessment of the ‘quality’ of the outcome of the consultation is defined by reference to an external account, namely the ‘correct’ identification of the ‘true’ presence of a disease state. There is much overlap with Kvale's⁷¹ description of the ‘mining’ metaphor within ualitative research interviewing, where the quality of the researcher's efforts is defined in terms of his or her ability to uncover the true story. The clinician/researcher is the active agent; the patient/interviewee is a largely ‘passive vessel’⁷¹ containing the information to be discovered (and, in the case of clinical practice, subsequently confirmed or validated through ‘objective’ laboratory testing). The logical extension of this model has been the development of a myriad of decision-making/diagnostic tools to maximise the accuracy of this process, with protocols for care providing certain accounts of appropriate intervention arising from a particular diagnosis.^16,17 Quality can be defined in terms of the ‘accuracy’ of the diagnostic process against a gold standard – the ‘true’ status of the presence/absence of pathology.

But in discussing the application of May's Normalisation Process Model⁴⁷ I have identified two major concerns with the application of a mining metaphor, and therefore the SB/EBM view of knowledge, to clinical practice. First, I have described problems with the assumption of a ‘certain’ account of disease, or illness, that can be mined for. Second, I have particular concern about the passive status of the patient described within the mining metaphor and implicit in the SBM approach and the EBM hierarchy of knowledge.

Social changes have challenged traditional professional roles, seeking to alter what have been regarded as ‘paternalistic’ models of clinical practice to give the patient more say in his or her own health care. Much of the current drive to recognise the patient in health care translates into a consumerist understanding of patient empowerment, offering greater ‘choice’ in aspects such as place and timing of care. Yet there is clear evidence of patients being more than consumers of care, of rather being proactive and sophisticated users of care and, by implication, knowledge.^72,73,74 There is evidence within the complementary and alternative medicine (CAM) literature of individuals undertaking sophisticated and multidimensional assessments of health status and need.⁷⁴ For some, explanations of ‘dis-ease’ within CAM frameworks make more sense of individual illness experiences than a pathological account.^73,74 CAM users are not simply dissatisfied with conventional medicine,⁷⁵ but make sophisticated assessments of their health needs and the appropriate therapeutic approach needed to address them,⁷⁴ drawing on a range of knowledges to support this process.

I would therefore argue that the biggest barrier to the normalisation of SB/EBM into general practice lies in SB/EBM's understanding of an external account of legitimate/best knowledge, creating a passive role for the patient. There is a stated desire to move away from paternalistic decision making by doctors, giving greater power to patients. Yet SBM with its model of the application of externally validated decision-making tools creates a new paternalism, with power removed from doctor and patient alike, and shifted to the ‘organisation’. We need a model of knowledge for practice that supports an active role of patients in the co-understanding and management of their own health.

The literature on research interviewing once again offers us an alternative model for understanding interactions. The metaphor switches from one of ‘mining’ to ‘travelling’, with researcher and interviewee viewed as being on an interactive journey where new meaning is created through the interaction.⁷¹ This travelling metaphor may also be applied to the clinical interview. Clinicians use their knowledge to support patients in developing an ‘exploratory map’⁷⁶ rather than an ‘explanatory model’ in creating meaning and identifying options.^25,76 Patient and practitioner contribute (to a greater or lesser degree) to the integration of personal, professional and scientific accounts, producing an individualised account of an illness experience and plan for further action. This proactive role of patients in determining and deciding patterns of care is recognised in professional descriptions of patient-centred, holistic care, and in some models of the consultation.^40,77,78 Yet some research has questioned the extent to which these professional models reflect rhetoric or reality.^10,11 Analysis of consultation data reveals evidence of the existence of this ‘travelling’, exploratory approach within daily practice.

With colleagues in the UK, USA and the Netherlands, I am currently involved in a study looking at GP consultations with patients who are depressed. Our original research question asked whether the degree of overlap and discordance in the doctor's and patient's conceptual models of depression predicted the outcome of a consultation, and particularly the emergence of guideline-concordant care. The work is still in its analytical stages, but findings have already led us to question our original hypothesis. A key observation has been the lack of a single conceptual account of illness in many patient or practitioner narratives. Our assumption that either or both parties bring a single (or dominant) conceptual model to the consultation may be inappropriate. Some transcripts do reveal a dominant pathological account with, for example, the doctor ‘mining’ for evidence of depression. Most are more complex, with either or both parties exploring multiple interpretations of the presented illness experience (for example situational versus biological causes). The resulting discussion reveals a co-constructed account of illness and an identified ‘plan of action’ for that individual.

General practice is about interpretation of illness, not identification of disease; knowledge is not uncovered (‘mined’) but constructed as the clinician and patient ‘travel’⁷¹ together, creating a joint account of illness that meets the needs of both. This recognises an interpretive role for practitioner and patient, with knowledge being generated by the interaction between them. It also views the consultation not as a technical process of identifying a ‘true’ diagnosis, but as a social interaction with potential consequences for the individual beyond the identification and selection of an evidence-based intervention.⁵⁴

This describes a shift in emphasis with regard to the nature of the consultation, as illustrated in Figure 1. The ‘mining’ view of the consultation within the SB/EBM approach places the focus of attention on the consultation itself: the processes occurring within the consultation, and the outcome (successful identification of ‘true’ disease status). The travelling/exploratory view of the consultation sees it as one social interaction within the context of an individual biography, and efforts to continue living life. The consultation may impact on the subsequent direction and nature of that biography. The focus of attention shifts from the consultation itself to the individual and his or her capacity to live his or her life. The quality of the consultation is understood in terms of its impact on that process.

Figure 1.

Models of the consultation

Application of the Normalisation Process Model predicts that SB/EBM cannot be normalised, or fully integrated, within generalist primary care.³ Both have arguably been adopted; that is, taken up but not routinely embedded.⁴⁷ The language of evidence-based policy and practice is evident in: the contractual arrangements dictating GP services; the appraisal and revalidation processes by which GPs are allowed to practice; and even in the lay health literature.⁷⁹ Observational studies have shown that evidence, including the assessment offered by evidence-based diagnostic tools, is used within clinical practice.^80,81 However, whilst evidence is accepted when it supports clinical judgement, adding weight to professional decision making, it can be rejected when it clashes with professional opinion.^80,81 Thus SB/EBM has not yet been fully integrated into practice.

However, a shift in the nature of general practice to a more specialist, disease-focused model could accommodate the normalisation of SB/EBM. Moreover, there is evidence of just such a shift in practice away from the person-centred model towards the disease model promoted by SB/EBM.^11,12,13,14 If personal care is something that needs preserving, then we need a theoretical framework that supports an interpretive, integrative approach to use of knowledge within practice: a general practice-based alternative to the models of SB/EBM.

Introducing Interpretive Medicine

I have argued that neither EBM nor SBM support a model of general practice; their dominance in Western healthcare systems threatens the generalist, patient-centred approach. EBM and SBM seek to create apparent certainty through defining best knowledge; however, uncertainty in clinical practice is inevitable, being a consequence of the nature of scientific enquiry.^32,34 The question not addressed by either EBM or SBM but often keenly felt by practising clinicians is that of epistemological uncertainty:³⁴ how to apply the knowledge derived from observation of populations to an understanding of the individual patient. Yet this is just one aspect of the bigger challenge raised by complexity theorists:³⁵ How do we integrate knowledge derived from different perspectives with diverse positions on the nature and creation of knowledge (for example genetics and molecular biology, physiology and pharmacology, sociology, even ‘non-scientific’ lay and tacit professional knowledge) to understand a coherent account of our world, and indeed an individual patient?

The complexities and uncertainties described in delivering generalist, person-centred care is part of the daily experience of working GPs.^3,32,63 Griffiths et al. ³² have described some of the strategies doctors use in practice to deal with uncertainty, describing a dynamic model of practice that recognises the changing nature of illness, uses provisional diagnoses and review, and specifically seeks to avoid contributing to a myth of medical certainty. Professional accounts of the core elements of general practice have defined the interpersonal skills needed to deliver person-centred, generalist care. But we are lacking a framework by which we can judge the quality, and thus demonstrate the value, of these interpretive elements. We recognise the limitations of SB/EBM but have no alternative account to offer in their place. We cannot afford to ‘ignore’ the public and government demand for more explicit accountability with regard to quality.^4,16,17,41 We therefore need a theoretical account of quality of knowledge use, of these integrative and interpretive aspects, in order to support the ongoing delivery and development of generalist, holistic and person-centred primary care.^1,2,3

Based on a critical review of my own, and others', research, I therefore propose an account of IM, being the critical, thoughtful, professional use of an appropriate range of knowledges in the dynamic, shared exploration⁷⁶ and interpretation³ of individual illness experience, in order to support the creative capacity⁸² of individuals in maintaining their daily lives.²⁷

Box 2: Defining Interpretive Medicine

IM is the critical, thoughtful, professional use of an appropriate range of knowledges in the dynamic, shared exploration⁷⁶ and interpretation³ of individual illness experience, in order to support the creative capacity⁸² of individuals in maintaining their daily lives.²⁷

I have outlined the exploratory, interpretive nature of the interaction that is the GP consultation. Professional accounts of quality of practice focus on the consultation processes that support holistic, patient-centred care.^2,4 However, the efforts to demonstrate an impact of this model of care on outcomes have proven disappointing.⁹ But I have also highlighted the need to look beyond the consultation (Figure 1), thus revealing a gap in the current theory of general practice, namely in describing what GPs seek to do for the individual. We can describe how we should practise, but not what the goal should be. Cassell⁸³ described the role of medicine as the relief of suffering, but Carel⁸² considers whether we can still be happy even if we are ill and, by implication, ‘suffering’. So what suffering are we trying to relieve? What should be the priorities for care? The Alma-Ata Declaration⁸⁴ called for ‘Health for All’, but what does that mean for the individual?

Drawing on the exploratory model described, and illustrated in Figure 1, I propose that the aim of general practice should be to support patients in living their lives. A core outcome of general practice is to support individuals in their efforts to maintain continuity of daily life.

Defining the ‘self’ in personal care: creative capacity

My discussion thus far reveals what I believe is the biggest conceptual gap in academic primary care theory. Professional accounts refer frequently to person-centred care and support of the individual. But there is no account of what the ‘person’ is. The nature of the self has been the basis of debate amongst philosophers for millennia. I do not propose the need to answer, or even debate, questions about the ‘true’ nature of the self. However, I do argue that there is a need for a conceptual model of the individual (person) that is the focus of care in general practice.

The limitations of a biological notion of the self (emphasising disease as biological dysfunction)⁸⁵ have long been recognised within primary care, and are addressed, for example, in the narrative-medicine literature.^76,86 Here, the emphasis is on a ‘narrative’ view of the self, with understanding of illness and clinical care defined in terms of the need to maintain continuity of meaning. The sociological literature describes biographical accounts of the self,^29,87,88 characterising illness in terms of its disruptive impact on the personal assumptions and social structures of daily life. Each perspective views illness as a deviant state and focuses on the need to correct the negative state.

Each account has been criticised. Biological accounts fail to recognise the social context of illness, but biographical accounts may not provide adequate account of potentially ‘treatable’ physical problems. Narrative accounts have been criticised for an overly cognitive view of the self,^89,90 neglecting the felt experience. Williams and Bendelow⁹¹ proposed that an emotional account of the self embraces aspects of the mind and body, as well as our interactions with the external world, thus acting as a ‘bridge’ between biological and sociological accounts of illness. Each describes an important element of the person who presents to his or her GP for help, but downplays the importance of other elements. General practice needs a theory of the self that integrates these perspectives and thus provides a frame of reference for its activity. People present to GPs because a perceived health-related problem is interfering with their experiences of daily living. The account of the self in general practice should therefore include these individual aspects of the self (psychological, physical, social, emotional) within a dynamic account of living. Carel's⁸² concept of the self as the creative capacity of embodied consciousness provides just such an account.

Like Williams and Bendelow,⁹¹ Carel⁸² recognises embodiment as a ‘bridge’ between our concepts of mind, body and the social world, viewing the lived body as our ‘being within the world’ and not simply a ‘physical container for the mind’. ‘Being’ is viewed as more than existence: a dynamic process of living in the world underpinned by self-agency (‘the ability to assert yourself and perform actions’). He thus describes a dynamic account of the self as ‘embodied consciousness’, linking self-agency and the embodied self. The potentially disruptive effect of illness on both the embodied self and self-agency is acknowledged. However, rather than viewing illness as necessarily negative, Carel⁸² proposes that its disruptive impact can promote adaptability. Adaptability (at physical, psychological, social and temporal levels) is viewed as a highly personal and creative response to disruption of the embodied self; ‘adversity is the source of creative response’, and thus development.

Whilst biological and psychosocial notions of the self have proven useful in developing models of health and health care, Carel's⁸² dynamic account of living with, and adapting to, illness in the context of ongoing lived experience is of particular interest in developing an account of the self for general practice. It has long been recognised that symptoms alone do not prompt people to seek professional help; the decision to seek help is linked to an individual's personal and social circumstances.^21,92 Patients that I see come to get help for the relief of suffering⁸³ that is affecting their ability to continue living their daily lives. Sometimes focusing solely on the biological aspects may be enough, particularly in acute illness. However, Carel⁸² emphasised the need to acknowledge the creative capacity of individuals in living their lives, criticising medical accounts of illness for failing to recognise, and thus support and utilise, concepts of individual agency. This is particularly relevant when dealing with chronic or recurrent illness, and thus much of the illness experience presented to GPs. Productive, catalytic or transformative aspects of illness have been recognised in a variety of illness settings.^{25,26,93,94,95} Carel's work⁸² provides a framework by which to understand and evaluate the impact of illness and care on this creative capacity.

Whilst practising GPs will recognise the dynamic account of the ill self from their daily contact with patients, it is also described in a body of empirical data. In my own work with people living with terminal cancer, people's account of living with illness focused not on the cancer itself but on the disruptive and supportive aspects of illness (and other) experiences in continuing to live their lives. The work highlighted the need to understand illness as an individual experience: one aspect of the challenges to people's creative capacity^82,96 in the ‘self-management’⁵⁴ of their own daily lives. ²⁷ People clearly described the challenges of dealing with threatening or distressing aspects of their lives (including, but not exclusively, the effects of the cancer). There was evidence of distress as a stimulus for change, and its productive effects. Some revealed reaching a threshold beyond which they were no longer able to maintain continuity; exhaustion precipitated a period of profound disturbance and upset that required external (medical) intervention to support restoration of continuity.²⁶

Interactions with healthcare professionals were identified as both a source of support and threat to individual capacity to maintain continuity of daily life. Noxious impacts of diagnoses were seen when these were incongruent with people's own sense of who they were and what mattered most to them, and thus their dynamic attempts to continue living their daily lives. Even ‘correct’ diagnoses may undermine people's self-agency, creating the need for an alternative perspective on ‘appropriateness’ or ‘quality’ of use of knowledge within the consultation. Instead of focusing on a technical process of identifying a ‘true’ diagnosis, the consultation becomes a social interaction with potential consequences for the individual beyond the identification and selection of an evidence-based intervention.⁵⁴ ‘Best’ knowledge is that which adequately supports patient and doctor in constructing an account of illness that supports, and certainly does not undermine, the individual's own creative capacity to live his or her life.

Biddle et al.'s ⁹⁷ work of (non-) help-seeking behaviour in young people with distress describes a similar account of people seeking to continue to live their lives, managing distress until a threshold is reached whereby they can no longer cope on their own, and so seek help. The researchers' concern was to highlight the potential harmful impacts of the delay in help-seeking caused by the initial attempts to manage, not least because these efforts involved ignoring or accommodating distress rather than resolving it. Fear of having a ‘real’ illness rather than ‘normal’ distress, associated with concerns about the stigma of mental illness, inhibited help-seeking. Their model provides empirical evidence of Carel's⁸² dynamic account of the self as embodied consciousness, living life; however, it also highlights the need for a dynamic approach to understanding help-seeking, need, and clinical care of young people experiencing distress.

In Beyond Depression, Dowrick²⁵ builds on philosophical, theoretical and empirical data to construct an account of the self as an active, creative agent. His account understands the personal experiences sometimes labelled as ‘depression’ not as a pathology to be ‘treated’, but as an important and even productive aspect of living that can be used to shape and promote that dynamic process. Beyond Depression describes the therapeutic benefits of a model of the self based on creative capacity, including a framework to address the need highlighted in Biddle's study. ⁹⁷

Griffiths et al. 's ²⁸ account of living with diabetes offers a similar account of an active self, but this time within the context of physical illness. Although primarily intended as a study to explore research methodology, a key analytical finding was the need to understand the dynamic nature of living with diabetes, together with the creative aspects of individuals in the ‘testing of possibilities’ to identify more or less successful strategies of ‘living’. The need to understand individuals as ‘complex living organisms’ – with constant interaction, construction and re-construction of ideas and experience over time needed to support functioning – is highlighted by Griffiths's work. This once again mirrors Carel's⁸² account of the self as the dynamic, creative capacity of an ‘embodied consciousness’.

Further examples of dynamic accounts of living with physical illness include Casiday et al. 's ^60,98 study of patients living with irritable bowel syndrome, and Exley's⁹⁶ account of people with terminal cancer. Both give insights into the significant efforts of people to carry on with their daily lives, and not let their condition ‘take over’.^60,98 Casiday et al.^60,98 highlighted gaps between patient and professional expectations of approaches to managing illness, and between current research-based guidelines for practice and the needs of patients and GPs alike. They highlighted a need for ‘better translation of research into practice’, this being the need to address incongruence between experimental bodies of knowledge and personal experience. ^98,99

Supporting creative capacity as a defining aspect of general practice

Carel's⁸² account of the self, and the empirical findings presented, address the primary care, generalist model of clinical practice that seeks to provide holistic, person-centred care: aiming to relieve suffering by addressing illness rather than focusing solely on disease. ^3,18 By providing a theoretical framework for the ‘self’ in personal care, it describes a reference point by which practitioners, individually or collectively, can judge the quality and outcomes of their practice. But by supporting and promoting self-agency, general practice also contributes to sustainable community development and enhancement of social capital through supporting creative and capable citizens. ⁸⁴ It directly challenges a notion of primary care as the community-based delivery of secondary disease-centred care within local polyclinics.¹⁰⁰ Stipulating support of creative capacity as a defining aspect of the GP role supports individual self-agency in the dynamic process of living life. It also values practitioners in their role as independent generalists, rather than technicians delivering those aspects of specialist, secondary care that can be addressed in the community. Finally, it promotes communities through supporting the development of a sustainable, involved approach to maintaining health rather than a consumerist one.

Interpretive Medicine in practice

Thus I have described an account of general practice, explicitly based on interpretive skills supporting individual creative capacity. Adopting a model of IM, rather than SB/EBM, could result in a very different consultation.

For a person presenting with emotional or psychological distress, an SB/EBM model would promote the use of screening and/or diagnostic tools to establish a diagnosis. Decisions on care would be guided by current protocols; the onus is on the professional to justify any ‘non-compliance’ with protocols of care. Good communication skills are used to promote the doctor–patient relationship, engendering trust and empathy. But ultimately the quality of the consultation would be defined primarily by the adherence to protocol.

Clinical practice informed by a model of supporting creative capacity may still include an assessment of ‘risk’, drawing on knowledge of a pathological account, but would also explore the patient's own explanatory accounts and efforts, and the tacit knowledge of the practitioner. All would be used to co-construct a personalised, integrated, interpretive account of illness experience that could be used to identify processes which support and even enhance creative capacity and daily living. This may still include the use of antidepressants or psychotherapy, but only where this supports the maintenance and development of creative capacity. The onus on the professional is now to justify the use of an external/disease model of care.

For physical illness, an emphasis on creative capacity may support and help decision making. Currently, the onus is once again on the professional to defend deviation from defined protocols of management of disease such as diabetes, coronary heart disease and risk factors such as hypertension and hyperlipidaemia. Current notions of quality of care often result in polypharmacy and difficulty in decisions about when to stop medication,¹⁰¹ creating a problem for patient and practitioner alike. A model of care emphasising creative capacity assesses benefit and risk of treatment against the implications for dynamic continuity of embodied consciousness, rather than risk of death or morbidity.¹⁰²

Decision making therefore explicitly involves the identification and integration of different accounts of illness and the self, discussed between practitioner and patient to identify an account of illness and a ‘management plan’ which best supports the creative capacity of that individual. With evidence that interactions with healthcare professionals, and the use of knowledge, can both support and undermine creative capacity,^50,54 we need to shift under standing of quality in primary care practice away from a focus on what knowledge is used, to how it is used. Thus I have described an account of IM as the critical, thoughtful, professional use of an appropriate range of knowledges in the dynamic, shared exploration⁷⁶ and interpretation of individual illness experience, in order to support the creative capacity⁸² of individuals in maintaining their daily lives.²⁷

Delivering Interpretive Medicine: skills for interpretation

The processes for delivering IM are described within professional accounts of education and practice, and within the literature on consultation processes and reflexive practice. Fundamental to the interpretive process is the use of both of the intellectual virtues described by Aristotle, namely sophia and phronesis.¹⁰³ Sophia refers to possessing and utilising knowledge of the external world derived from scholarly research, including biological and biographical accounts of human experience.³ Sophia is valued by SB/EBM approaches. Phronesis, usually translated as ‘practical wisdom’, refers to the use of intellect and knowledge of the world to effect change. In the context of this paper, I propose that this refers to how, rather than what, knowledge is used: the ability to critically integrate and interpret knowledges, both internal and external, in the construction of an individualised assessment of need that supports the dynamic needs of the patient in continuing with his or her daily life. It refers to the ‘judgement’ in clinical practice that Cassell⁸³ describes as necessary to ‘bridge the gap between the two cultures of medicine’, namely illness and disease: applying knowledge about normal function and disease to this individual's illness.⁸³ Within medical education, it is the skill of reflective or professional judgement, personal knowledge, or capability,¹⁰⁴ which educators seek to enhance through reflection, performance feedback, and flexibility in goals.¹⁰⁵

Exploration and interpretation

Phronesis also encompasses the interpretive aspects of clinical practice: integrating different perspectives to support construction of an individualised account. This intellectual capacity is one that derives from professional knowledge and experience. It is developed over decades of experience, passed on through professional training, and found within the intuitive or tacit knowledge of working professionals. This tacit knowledge¹⁰⁶ is more advanced than the external research efforts to describe it, reflecting both the different processes governing knowledge generation and ‘validation’ within clinical practice versus science or ‘academia’, as well as the length of time each has cumulatively devoted to the endeavour. Rejection of this aspect of clinical practice within SBM represents a fundamental flaw in the model in terms of delivering care that supports individual creative capacity.

Judging quality of practice

However, whilst the generation of interpreted knowledge is an essential and daily aspect of working general practice, what the profession does not have is an adequate framework by which it can be externally judged to have been done well. Judgement of interpretation is viewed as a skill and task for the individual practitioner; it is part of reflective practice, and assessed within, for example, appraisal processes. But in an era of NPM this is not adequate.⁴ Turning again to an analogy with the qualitative research field, just as researchers have had to define, describe and defend their interpretive approaches and knowledge production to an external world, so must the profession of general practice.

SBM and EBM both share a quality standard for knowledge and practice, being ‘objectivity’ as a marker for ‘truth’, valuing consistency, repeatability and predictability as evidence of quality. Yet an interpretive approach valuing individual generation and application of knowledge is associated with variability and uncertainty, rendering truth and objectivity an inappropriate solitary marker of quality. But if we cannot use truth as the basis for judging the quality and appropriateness of knowledge and knowledge use, what are the implications? Being able to judge between competing bodies of knowledge has a powerful social function through promoting shared understanding and response to situations.^57,107,108 We therefore need an alternative framework for judging between competing knowledges.

Such questions are not commonly asked by clinicians, but have been the subject of debate amongst philosophers and scholars for centuries. Just as academic debates about the nature of knowledge in the interview offered us the ‘mining’ versus ‘travelling’ metaphor, so we can look to academic accounts of quality of knowledge to develop this framework.

Theoretical (ontological) perspectives on the nature of reality can be seen along a continuum between those who assert (realist) or deny (antirealist) the real existence of ‘some kind of thing … or state of affairs’ that exists independently of human thought or experience of it, not being ‘an artefact of our minds, language or conceptual scheme’.¹⁰⁹ The natural sciences operate within a realist perspective; they use a (logical) positivist framework to reveal, through empirical observation, the ‘universal laws’ that govern the world. That which cannot be directly observed (and therefore measured) falls outside of the remit of scientific study. Medical science and the study of disease traditionally operate within this framework. Knowledge, and therefore scientific accounts of disease, is valued by virtue of its proximity to ‘truth’ or reality.

By contrast, antirealists believe that no external reality exists, only that which is constructed by human endeavour. Science does not ‘discover objective reality’, but creates and tests theories, all within the assumptions (constraints) of the current worldview.^110,111 These assumptions shape what scientists choose to observe, and the way they interpret their observations; they are influenced by religious, economic, political and other social factors.¹¹⁰ All knowledge is therefore ‘knowledge from a certain position or perspective’;¹¹² there is no single version of truth to be valued above another.

This view of knowledge informs the postmodern account of general practice argued by Mathers and Rowland.¹¹³ Postmodernism is a term used to refer to the social, political and economic changes associated with the move away from a focus on industrial, urbanised societies, to a period of globalisation, mass media and ‘technological revolution’.¹¹⁴ Many theoretical accounts of postmodernism exist, but all share an antirealist view; they are sceptical about the existence of a fixed or certain reality, and hence exhibit a distrust in science, or indeed any single theoretical view of the world, as ‘the truth’.^107,114 Postmodernists argue that biomedicine alone cannot provide a single, ‘best’ account of illness, but rather offers just one (of many) accounts.^113,114 Recognising the complex, dynamic and uncertain nature of general practice, Mathers and Rowland¹¹³ argued that knowledge derived from a realist emphasis on objectivity and truth had ‘very limited application in our day-to-day work [as GPs, and] … can create very real difficulties for both doctors and students as they struggle to apply such a model that does not very often or necessarily fit’. They proposed that a postmodern approach to evidence and knowledge has the potential to support the flexibility needed for practice.¹¹³ However, postmodernism, and therefore Mathers and Rowland's account,¹¹³ has failed to offer a framework by which to judge between different knowledges.

Others have described an, arguably pragmatic, ‘third way’: that of subtle realism.¹¹⁵ Subtle realists neither assert nor deny the existence of an external reality. Instead they propose that, even if such a reality exists, we can never know it with certainty. They acknowledge the need to be able to judge between different accounts, to judge the ‘quality’ of knowledge, but propose that we need a marker other than ‘truth’ to establish ‘trustworthiness’.¹¹⁷ Knowledge should therefore be judged not only by reviewing the method by which it is produced,^116,123,124 but also by its ‘plausibility’ or ‘credibility’.¹²⁵ Subtle realists explicitly recognise a legitimate role for judgement in the assessment of knowledge.¹²⁴ Maxwell¹¹⁶ described this as evaluative validity, while Kvale as pragmatic validity:⁷¹

whatever assists us to make actions that produce the desired results. … The effectiveness of our knowledge beliefs is demonstrated by the effectiveness of our action.

Maxwell¹¹⁶ developed these ideas to identify five criteria for judging the merits of the knowledge generated from research within a subtle realist perspective. These include the collection of data (descriptive validity), its interpretation (interpretive and theoretical validity), and the use to which the knowledge could be put (generalisability and evaluative validity) (see Table 1). Applying this model to the process of IM offers us a framework by which to judge the knowledge generated within clinical practice (see Table 1).

Table 1.

Definition of quality of interpretation within research and clinical settings, drawing on the subtle realist perspective

Criteria for judging knowledge in subtle realist perspective (Maxwell)116	Application to research setting (drawing on Maxwell116 and Lincoln & Guba;117 described further in Reeve118)	Application to clinical setting, as described within IM
*Descriptive validity: quality of processes underpinning data collection	Refers to processes that ensure accurate recording of what happened/was observed	Refers to the processes that ensure accurate collection of ‘data’ about individual's current illness within the context of his or her lived experience
	Includes methods of recording data, time spent in the field, reflexivity and peer examination	Includes interpersonal and person-centred communication skills described within professional accounts, for example as summarised by Stewart119
	Demonstrated through a transparent, critical account of the research process	Demonstrated through measures of person-centred care, for example empathy70
*Interpretive validity: quality of processes used to interpret observations and identify meaning	Refers to processes that translate ‘data’ into ‘meaning’ or ‘knowledge’	Refers to processes that support interpretation of the presented illness account into a meaningful understanding of this illness (in this context at this time … )
	Includes the critical, iterative processes described within the hermeneutic canons of Radnitsky120	Includes the iterative, shared exploratory process76 by which doctor and patient identify and test possibilities, underpinned by an understanding of the nature and limitations of different knowledges121
	Demonstrated through a transparent account of the process of interpretation supported by data	Demonstrated through reflexive case studies; the narrative clinical record; Cape et al. 's 122 measure of complexity of understanding as a proxy measure
*Theoretical validity: assess how well the outcome of analysis serves as an interpretation/explanation of the phenomenon	Refers to the ability to infer from this research study and comment on the wider phenomenon under study	Refers to the extent to which the interpretation (generated knowledge) from this (these) consultation(s) offers an understanding of the individual patient's illness experience (rather than disease)18 and supports patient and doctor in constructing and updating a management plan
	Assumes work is grounded in existing theory, demonstrates structural coherence and referential adequacy	Assumes the consultation adopts an ‘exploratory’ rather than ‘mining’ approach; seeking shared understanding and opportunity for reflection on/learning from outcomes
	Demonstrated through structural coherence and critical reference of new findings to existing knowledge (referential adequacy)	Demonstrated through, for example, knowledge audit in the appraisal process (reviewing knowledge used, opportunities missed); (future) research into exploratory model of consultation; outcomes in supporting creative capacity
Generalisability: the ability to infer from the account	Refers to the ability to draw inferences about other populations/situations from the result of this research study	Refers to the ability to draw inferences about the illness experience and needs of other patients from learning derived from working with this individual
	Assumes processes of sampling, data collection and analysis are adequate/ well-described so reader can judge generalisability	Assumes processes of data collection and analysis (i.e. the consultation) are adequate and well-described
	Demonstrated through adequate demonstration of the processes of theory generation and critical evaluation by others. Recognises an iterative/developmental process where application of knowledge (theory) in another setting may contribute to further development of a theory. Knowledge is not fixed and certain, but evolving and adapting	Demonstrated through new work to evaluate the processes of knowledge generation and critical evaluation by others, for example consultation analysis. Recognises an iterative, developmental process of lifelong learning and reflection, raising issues about the ‘validity’ and utility of tacit knowledge106 in supporting others
Evaluative validity: the utility of the knowledge produced	Refers to the application of knowledge generated in answering the question ‘So what? ’	Refers to the ‘So what? ’ of IM; the extent to which this interpretation supports (promotes) the creative capacity of both individual patients and the communities we work with, in their efforts to continue living their lives
	Assumes knowledge cannot be judged in isolation (as a ‘thing’ in itself), but be understood in the context in which it is used	Assumes knowledge cannot be judged in isolation; a diagnostic score alone is meaningless but must be interpreted in terms of the context in which it is used (namely in supporting creative capacity and self-agency)
	Demonstrated by application of the knowledge, for example in further research	Demonstrated through narrative accounts that demonstrate maintenance of daily living;27 through (new) work to assess creative capacity and through community development outcomes

Note * Maxwell acknowledges that these three are interlinked and often hard to distinguish.

Source after Maxwell.¹¹⁶

Within the research setting, descriptive, interpretive and theoretical validity relate to the processes and outcomes of the generation of knowledge ‘at the front line’, namely the fieldwork and generation of knowledge (ideas) within the context of the single study. Researchers interpret their data from a particular theoretical perspective in order to draw conclusions about the observed phenomenon. The final two criteria, generalisability and evaluative validity, refer to the application of findings: to the ability to use the knowledge generated from the individual study to infer something about the rest of the world.

The clinical equivalent of ‘front line’ data collection and knowledge generation occurs within the interaction(s) between clinician and patient that take place in the consultation(s). The ‘theoretical perspective’ used by GPs to inform interpretation becomes that of the creative capacity of the individual;⁸² the interaction generates theory (or new knowledge) about this illness within the context of this individual's creative capacity. The utility of the knowledge can be judged in terms of how well it supports (or at least does not undermine) creative capacity. Generalisability relates to how understanding of this individual can be applied to others.

Drawing on Maxwell's¹¹⁶ account I therefore propose a structure for judging knowledge produced from an IM process (Table 1).

Operationalising the account

Table 1 outlines a framework for demonstrating quality of knowledge production within IM. Some of the processes are described within existing models of professional practice: within patient-centred medicine and reflexive practice.^1,2,3,104 There are a number of measures of quality within the consultation^5,69,70,122 that support the assessment of the ‘data collection’ elements of knowledge production, and even the interpretation of illness. Existing education and appraisal processes seek to promote and identify reflective practice as a key component in the demonstration of interpretive practice. Yet SBM and EBM are the dominant models of knowledge use within Western healthcare systems. This reflects their ability to measure and demonstrate their implementation and apparent effectiveness, an aspect of practice and policy that is highly valued in Western society. With evidence that current theories of knowledge and practice are having a negative impact on the model of general practice,^{8,11,12,13,14} we urgently need new work to demonstrate the utility and effectiveness of the interpretive approach.

Specifically, there are three areas for development: markers of creative capacity to demonstrate the evaluative validity of interpretive general practice at the level of the individual; understanding the evaluative validity of interpretive general practice in supporting community development; and methods to support the explicit recognition and value of tacit knowledge and thus the generalisability of interpretive general practice.

Evaluative validity: individual measure of creative capacity

The utility of IM and general practice lies in its potential to support the creative capacity of individuals to continue living their lives. This can be captured in narrative accounts of illness,^27,77,85,126 which will be known to many within the profession. However, we now need a ‘short-hand’ measure by which we can demonstrate this to those who fund the health service and direct health policy.

Narrative studies offer insights into the identification of narrative patterns, or markers, which predict outcomes.^27,127 Research within palliative care populations suggests that measures of exhaustion may function as markers of impending loss of capacity.⁵⁴ Consultation-based studies into the impact of use of knowledges within practice on creative capacity and patient outcomes could support identification of new markers. Ethnographic approaches could be used to look outside of the consultation and explore the wider impact of knowledge about illness on creative capacity and maintaining daily life.¹²⁸

Evaluative validity: supporting community development

IM promotes individual's creative capacity, but there is a danger in emphasising individual autonomy over and above broader issues of social justice.¹²⁹ Individual empowerment must not come at the expense of what has been described as ‘community capacity’: ‘the set of assets or strengths that residents individually and collectively bring to the cause of improving local quality of life’ [my emphasis].¹³⁰ The Community Development for Health movement argues for a ‘bottom up’ approach to health, whereby communities identify health issues, plan and act on strategies for change, and thus gain increased self-reliance and decision-making powers as a result.^129,131 Empowerment of the individual, for example through support and promotion of creative capacity, is an important component in empowerment of communities. IM is vital if general practice is to contribute to the goals of primary care outlined in Alma-Ata,^3,84 and more recently in the WHO's update Now More than Ever.¹³² Joint research between general practice and public health, looking at the impact of the use of knowledge in practice on community capacity and ‘social capital’, will be an important part of establishing the wider evaluative validity of IM.

Generalisability: establishing ‘trustworthiness’ of tacit knowledge

The question of generalisability brings to the fore issues about professional learning and how we can ‘legitimately’ use knowledge derived from one patient to support the care of future patients. As with the generation of knowledge in qualitative research, this can only ever be a ‘theoretical generalisation’. Previous insights offer us a possible explanation, or a hypothesis, which may be relevant to this patient. However, it must be explored critically, and rejected if it fails to support an adequate explanation of the new case. This is the tacit knowledge described by Polanyi,¹⁰⁶ knowledge that EBM accepts but views as ‘inferior’ to external scientific knowledge, and knowledge that is ‘rejected’ by SBM. A way needs to be found to legitimise this body of knowledge. As with qualitative research, this requires it to be externalised and subjected to public scrutiny. Development of a model of Practice-Based Evidence must be a priority for academic general practice. There is a clear role for academic primary care to highlight to policy-makers and politicians how and when the biomedical, disease model is inadequate to understand and support primary care, with research that demonstrates and illuminates the complexity of daily practice, and not solely the development of pathways to deliver uniform care. This needs resources in the form of time, expertise, funding and priority (recognition of value); however, this also raises methodological questions about how to evaluate knowledge generated from practice. Action research methods¹³³ and analytic autoethnography¹²⁸ may offer insights here.

Critiquing the Interpretive Medicine account

I have proposed a model of IM as a defining aspect of general practice, central to a sustainable system of health care that promotes individual capacity and hence community development. I started by suggesting that the application of May's Normalisation Process Model⁴⁷ reveals that SB/EBM cannot adequately support general practice. But does IM fare any better?

Given that the explicit aim of IM is to support interpretive assessment of illness experience in order to support the creative capacity of the individual, IM should, in theory, support interactional workability (see Box 1) with congruence and shared accounts of illness. This might be expected to improve relational integration in the form of patient confidence in the consultation process. Professional accountability is, paradoxically, increased since the responsibility to both identify data and interpret findings is placed back on the professional, rather than an ‘anonymous’, impersonal diagnostic tool or protocol. The problems perhaps arise within the organisational areas, namely skill set workability and contextual integration. The collection of ‘data’, the shared, reflexive process of interpretation, and the expectation of an iterative process of review and revision takes time, energy and commitment. There are resource implications in the allocation and performance of tasks. Thus establishment of the evaluative validity of the approach is a priority.

Harrison^16,17 argued that NPM, and hence SB/EBM, was a reaction to the need for organisational (over individual professional) control of the publicly funded healthcare system. IM would arguably fail to address this fundamental driver of the witnessed changes. Yet it could be argued that SBM is not working as a model to control costs. We are witnessing (unsustainable) escalation in demands for health care, rising costs of clinical care that increasingly focus on medical control of risk factors, and further extension of the ‘medical gaze’⁵⁷ within the phenomenon Moynihan et al. describe as ‘disease mongering’.⁴⁹ Work in the field of the expert patient suggests a moral and financial argument to focus on self-agency.^134,135 Critical management theorists have questioned models that fail to acknowledge complexity and address the need for flexibility within dynamic systems.¹³⁶ If IM can be demonstrated to promote self-agency, and promote an upstream rather than downstream approach to health care, then it may be in a position to challenge the NPM approach.

Perhaps addressing disease is the only legitimate role for the GP. The mining metaphor is appropriate: if no disease is revealed, the individual should be signposted elsewhere. Any other model of care risks inappropriately extending the medical gaze,⁵⁷ (over) medicalising human experience,⁴⁹ and using expensive resources (GPs) for problems that would be better (more cheaply) dealt with by others. Yet illness is more than disease; it can significantly threaten people's capacity to continue living their lives. Interactions with doctors are perceived to play an important part in supporting self-capacity.⁵⁴ GPs' extensive tacit knowledge, and their ability to integrate knowledge to ‘travel’ with their patients in dealing with presented problems, plays an important part in this therapeutic role. GPs are currently well placed to undertake this role.

Conclusion

General practice fought to establish itself as a unique discipline: a desirable career for the brightest medical students and not a ‘second class’ career for ‘failed’ hospital consultants. Current changes threaten these efforts, seeking to move disease-focused secondary care models into the community, with primary care clinicians delivering the ‘easier’ aspects of care that do not need specialist hospital intervention.

Instead, I have proposed that general practice needs to be understood as an intellectual discipline, defined by its interpretive skills. We need to recruit and retain the brightest medical students. GPs need to have a working knowledge of the philosophy of science, the generation of knowledge, its critical application, and the creative capacity to support individual patients in living their lives.

External quality standards and measures appear to be here to stay for the foreseeable future. I have argued that they are inadequate to support ‘best’ practice in primary care, being unsustainable and even harmful. Notions of legitimate knowledge shape ideas about quality in practice. I have argued that practical wisdom, and specifically the ability to integrate knowledges to provide individualised care, should be developed as a marker of quality in general practice. The profession should be defined not by the body of knowledge it owns/uses but by the way it uses knowledge. We therefore need to shift the gaze from easier-to-measure but limited accounts of practice based on the application of certain knowledge to a more appropriate assessment of knowledge use. Such change is necessary to both strengthen and preserve core elements of the discipline, and promote and support the health needs of the public.

Abbreviations

CAM

complementary and/or alternative medicine

EBM

Evidence-Based Medicine

GP

general practitioner

IM

Interpretive Medicine

NPM

New Public Management

SBM

Scientific Bureaucratic Medicine

Charter

The Royal College of General Practitioners was founded in 1952, with this object:

To encourage, foster and maintain the highest possible standards in general medical practice and for that purpose to take or join with others in taking steps consistent with the charitable nature of that object which may assist towards the same.

Among its responsibilities under its Royal Charter the College is entitled to:

Encourage the publication by general medical practitioners of research into medical or scientific subjects with a view to the improvement of general medical practice in any field and to undertake or assist others in undertaking such research.

Diffuse information on all matters affecting general medical practice and establish, print, publish, issue and circulate such papers, journals, magazines, books, periodicals and publications, and hold such meetings, conferences, seminars and instructional courses as may assist the object of the College.

College Headquarters

The Headquarters of the Royal College is at 14 Princes Gate, Hyde Park, London SW7 1PU (telephone: 020 7581 3232). Publishing enquiries should be addressed to the Publications Manager and general enquiries to the Chief Executive Office.

Occasional Papers

The Occasional Paper series was established in 1976 to offer authors the opportunity to publish manuscripts too long for publication as articles and too short for books. They are assessed for academic acceptability by the Editor, College Publications, and other professionals with special interests. Readers should note that they express the views of the authors, and not the College, unless otherwise stated.

Occasional papers are included in Index Medicus.