1. Influence of context on personal and professional autonomy

Western philosophy accords considerable importance to the notion of individual moral agency, a facet of which is the personal autonomy underlying a patient’s clinical decision making. Such a view is represented by our fictitious patient in the face of potentially serious illness. Involving complexity as we have described it has several important consequences. Our patients do not arrive at medical interactions in isolation: they have families and contexts within which their decisions are made. Like Mr Hobbs, they are nested in complex systems. Similarly, the decisions made by the doctor in our case have been influenced by the relationships she has within the practice and outside it.

Interactions between clinicians and patients in primary care are thus complex, dynamic and continually evolving. Each brings their moral positions to the clinical encounter, which as we have seen are determined by the systems of which they are part. When these moral positions differ, it can bring challenge to both parties.8 These differences may be influenced by religious or ethnic perspectives and expectations and by the culture of the practice. The temporal relationship between clinician and patient can augment this challenge.

Moreover, we would challenge the notion that in the interaction between patient and doctor, autonomy (personal choice and decision-making ability) is independent of the interaction itself, a concept which underpins normative ethical theories. Autonomy, as seen through the lens of a doctor–patient interaction, is realised through the choices made available to a patient by the doctor and those entertained by the patient. Personal autonomy could in fact be considered illusory in the absence of an interaction that acts as a catalyst for the realisation of choice and the enactment of decision making.

In the dynamic, and continually evolving (sometimes rapidly) interaction between clinician and patient, normative moral theory struggles to take account of the complexity arising from multiple interactions between agents.

2. Power relations

There may be a tension between the desire of the individual clinician to provide patient-centred, context-specific care to his or her patient and the need for the practice as an organisation to be seen to be achieving quality markers and to be adhering to standard practice. The way in which this tension is resolved will depend on how communication takes place within the practice and on the relationship the GP has with their patient. Ring and Launer9 suggest that in some practices, it is possible to talk openly about the effect of power differentials on roles and relationships, whereas in other practices, these factors remain hidden, influencing the type of conversation that can take place. The types of communication which are ‘permitted’ within the practice emerge through historic cycles of interaction (gesture-response cycles), which determine who is ‘allowed’ to speak and who is heard. This includes not just verbal and non-verbal communication between individuals, but also emails, social media and other communications from external bodies. So, in the example of Mr Hobbs, power relations are relevant not only to the interactions of staff within the practice, but also to the way in which the practice relates to external regulatory bodies and how health policy is communicated to individual clinicians. In our example, the history of the communication between these external regulators and the practice has resulted in an implicit acceptance of the current dominant discourse which privileges the extending of life over other considerations, in this case through the early detection of cancer.

3. Group identity and the development of values

Group and professional identity are developed and maintained through a process of repeated interaction between members of the group and with those from outside the group. It can be surmised that norms and rules emerge from the complex interactions of a group, and we would assert that moral rules can be considered similarly. Knowledge creation, narrative and group identity are the product of an ongoing iterative communication between individuals that ultimately define the ethical rules which the group abides by. In our case, the GP responds as she does because she is part of the practice and the decision to refer is based on the rules of medical behaviour which have emerged as a result of the interactions of staff with each other and with the external structures of the National Health Service. Or put differently, it is in the nature of groups to identify ways of being and ways of doing that approximate to normative ethical rules.10 Clinicians in primary care specialise in clinical generalism but this does not extrapolate to assuming an affinity with moral generalism. A rational decision to move away from a collective, principle derived course of action may be justified if the particular situation demands it.11 There are an infinitely variable number of clinical dilemmas in primary care, which are not amenable to rule or principle-based analysis. In this moral particularism, there may lie a similarity to the complexity we describe.

4. Characteristics of primary care

There are certain features of primary care (or its analogues general practice and family medicine) which are unique, and which are relevant when considering the interplay of complexity and normative approaches. Archetypally, primary care is typified by long-term relationships between clinician (with an increasingly multiprofessional representation) and patient, encompassing a wide spread of clinical presentations, which are influenced by families and communities; with clinicians embedded within these.12 Aspects relevant to the present discussion include the degree of uncertainty about diagnosis and outcome,13 and the notion of ‘first contact’ whereby patients consult as their initial engagement with organised healthcare.14 Of significance is the claim of primary care, in some jurisdictions, to universal coverage such that all persons have access to that first contact care, and to specialist input if needed thereafter.15 In addition to the diagnostic uncertainty and multimorbidity encountered in primary care, classification of health status is challenging compared with the more controlled and structured environment of a hospital.

As described above, the nature of the relationship between the patient and his GP is likely to be salient in any treatment decision. If the GP has an established relationship with the patient unlike in our case, it may alter what is privileged as she makes the decision about whether to refer; the role of the doctor as representing the practice may be subordinated to the individual doctor–patient relationship.

In some encounters between a GP and a patient, a degree of persuasion may be involved: clearly persuasion lies on a spectrum between laissez faire and enforcement, which impacts on or respects the autonomous decision making of patients. Such persuasion may be mediated by the continuing relationship of the doctor and patient. Autonomy and shared decision making are thus constrained and liberated by the nature of the interaction where power differentials, flow of communication and diversity of thought play important parts.

A morally appropriate outcome should recognise the influence of these contextual issues as well as the unpredictable and emergent nature of such complex interactions. We suggest that traditional moral theory has not in fact recognised, other than rather superficially, the dilemma central to our case or other aspects of primary care practice.

5 Moral theory and primary care

Primary care is not particularly unusual in its access to traditional moral theory: including inter alia utilitarianism, deontological theory and virtue theory. A consequentialist view of our patient’s case would suggest a straightforward involvement of secondary care in the aim of a better, life extending, clinical outcome. A deontological view would likely do the opposite—making the patients’ decision a question of respect for persons, and the autonomy which flows from it.

Utilitarianism is a consequentialist theory which involves practicing in a way which the clinician believes will result in the greatest good for the greatest number of people. This usually involves rationing resources based on a judgement of what seems to the clinician to be fair or reasonable; and this judgement may conflict with that of their patients. Such a judgement rests on a ‘dual role’ for primary care practitioners, conflating the duty to the individual patient with the responsibly for the community served. In some jurisdictions, this population role is formalised into that of a commissioner or gate keeper. In our case, the drive for early cancer diagnosis could be described as a population health outcome, which improves health overall. Utilitarianism assumes that outcomes resulting from actions or policy are predictable. Complexity theory by contrast asserts that the future is essentially unpredictable.

The ethical theories of consequentialism and deontology consider that the stance adopted by a practitioner is chosen purely on its own merits rather than being influenced by factors we have discussed above—guidelines, performative technologies which incentivise particular outcomes, external quality regulators and the complex interactions which occur between members of the group and which are responsible for the emergence of the moral rules adopted by the group. They also assume that there are predetermined moral positions which are not influenced by the interaction between clinician and patient.

The links between virtue theory and primary care are perhaps harder to see. Virtue theory is not concerned as much with the approach taken by the GP to ensure a satisfactory outcome for her patient; but more with the definable attributes that underlie her practice.16 17 Some authors, such as Mercer, have argued that the empathetic skills of the clinician are a particularly valuable element of the therapeutic relationship, motivated by a kind of moral altruism.18 We might argue that the doctor in our case should be honest with her patient and his relatives, that she should show courage in dealing with the clinical situation as she finds it, and she should not let self-interest influence her decision making and exercise wisdom in applying her knowledge and skills to the dilemma. In doing this she will have displayed three of the five focal virtues put forward by Beauchamp and Childress.19 That there are virtues which primary care practitioners should aspire to has been proposed20 in recent years; and to some extent the key characteristics of primary care practice as described above could be held to be relevant. For example, being a first contact clinician demands the focal virtue of courage more than in encounters where patients have been ‘processed’ through the layers of initial exploration and investigation later down pathways of care. In considering virtue theory in the context of complexity, it is worth questioning whether the virtues described above are immutable within individuals or whether they develop with continuing care over time, involving evolving conversations with patients. We suggest that choice and decision-making (autonomy) are in fact framed within these interactions between the two protagonists rather than being wholly dependent on any inherent characteristic of the protagonists themselves. This is reminiscent of the theory of narrative practice, where values are shared and confirmed between clinicians and patients.21 Conversely, it is possible that if virtues are intrinsic, then they are subject to erosion by external and internal influences within a system.