INTRODUCTION
In this analysis we challenge the conceptual honesty of ‘shared decision making’, arguing that, although it is held up as an ideal decision-making standard, it is used too often to describe (and justify) decision-making practices that are not shared at all. This is problematic because, if the legitimacy of a decision relies on it being ‘shared’, but it is not in fact shared, the decision loses legitimacy, or is falsely legitimised by the appearance of being shared. We argue that the realities of clinical practice mean that genuinely shared decision making is not completely impossible but difficult to achieve in a sincere and just manner. We articulate an intentionally controversial stance, with the aim of generating thought and debate.
DECISION MAKING AND THE CLINICIAN/PATIENT RELATIONSHIP
Shared decision making has been offered as a way of better respecting patient autonomy, leading to management plans or actions that are in tune with patient values and therefore improving concordance. It provides a platform for respecting autonomy while avoiding abandoning patients to their autonomy, by giving them information without assistance in interpretation; this allows them to arrive at informed decisions that are in accord with their beliefs, values, and preferences. This ideal accords with the deliberative model of the clinician/patient relationship put forward by Emmanuel and Emmanuel being morally preferable to a paternalistic or informative/commercial relationship.1
In arguing that shared decision making is the pinnacle of patient-centred care, Barry and Edgman-Levitan describe it in the following terms:
‘… the clinician offers options and describes their risks and benefits, and the patient expresses his or her preferences and values. Each participant is thus armed with a better understanding of the relevant factors and shares responsibility in the decision about how to proceed.’2
Commentators lament, however, that:
‘Unfortunately, this ideal is inconsistently realized today. The care patients receive doesn’t always align with their preferences.’ 3
FAILURE TO ATTAIN FULLY SHARED DECISION MAKING
It is important not to miss the important point that the care actually received is not a good barometer for the extent to which the decision was shared. Numerous external factors can impact on the translation of that decision into treatment.
It is probably true, however, that fully shared decision making is often not achieved, and in that case we ought to be motivated to change our practice to ensure that we meet the standard more often. In making this resolution we accept in principle that a failure to engage in ethically optimal shared decision making is an ethical failure, because we are failing to do something that we ought to do.
We believe however there are certain contingent and essential features of the clinical encounter that frequently make attainment of fully shared decision making impossible. We offer the following three arguments in support of this:
Ideal forms of shared decision making seem to rely on the assumption that the two parties engage in a deliberative and rational information exchange, through which they come to shared understanding and harmonious consensus on the most appropriate course of action. The majority of clinical encounters simply do not, and cannot, allow sufficient time to engage in the kind of dialogue required to arrive at genuinely shared decisions. Instead, what we tend to find is that options are presented by the clinician, which are simply agreed to by the patient.
The Anglo-American medico-legal tradition has shaped consent into a negotiation where (with the exception of forced treatment and in the absence of external constraints of resource) the clinician and the patient each get one veto. Patients can refuse treatment that they do not consent to, and clinicians can refuse to provide treatment they believe to be non-beneficial or will cause clinically unacceptable harm. The fact that this right of veto exists — and that GPs can choose, for example, to: not mention homeopathic treatment as an option based on their scientific beliefs; refuse patient requests for antibiotics on grounds of futility; not suggest certain options on resource grounds — limits the extent to which we can see very many decisions as genuinely shared. Any healthcare worker may have a distinct set of goals, beliefs, and values, and these often go unobserved unless they give rise to conflict — a phenomenon remarked upon by Petrova and colleagues in their work on values.4
The assumption that information can be objectively conveyed without selection or interpretation (what is referred to as a ‘conduit’ model) is problematic, and affected by (inter alia): knowledge and power differentials, memory retention, fear, embarrassment, cognitive bias, cultural difference, and so on. As Manson and O’Neil note, while critiquing contemporary notions of informed consent, the practice of giving information is complex and often opaque:
‘Neither full explicitness nor complete specificity is possible … For human agents, with varying beliefs, varying inferential commitments and varying vocabularies, the best we can hope for is a mutually agreed level of specificity in the disclosure of a particular transaction.’5
If what we have said is plausible, then there are both contingent (limited time and right of veto) and essential (the imperfectness of communication) factors that will very often make fully shared decision making untenable.
ETHICAL STANDARDS AND ETHICAL PRACTICE
This leads us onto the horns of a dilemma. We must either accept that the majority of decisions made in clinical practice fall short of shared decision making and consequently represent some form of ethical failure, or we must acknowledge that fully shared decision making, in most cases, is unachievable in current practice, and that this ought not be viewed as an ethical failure on the part of the clinician.
We suggest a move towards the second horn. There is something troubling about doctors being held up to an ethical standard that is, for the most part, untenable. This highlights a critique of ethics as a theoretical enterprise, often driven by political commitments, which makes proclamations about ideal forms of practice that fail to take account of context. Consider, instead, contemporary accounts of empirical ethics, which view practice as a starting point, and theory as a guide, in which the aim is to balance competing interests, practical features, and theoretical demands, to arrive at a workable, justifiable, solution. Ives characterises this as a process of ‘noble failure’, in which we must accept that the ethical judgements we make are often necessarily constrained by the choices available to us, and the solution often requires justifiable compromise between what we might want in terms of ideal theory and what is possible.6
Essentially, this is a matter of recognising that we are ‘succeeding in doing the best we can’ rather than lamenting that we are ‘failing to do the best imaginable’.
CONCLUSION
We suggest that, far from fully shared decision making being required on every occasion, simple (informed) agreement, in the context of a fiduciary relationship, is often good enough, and the level of collaboration required for genuinely shared decision making is only ethically required for substantively important decisions. Of course, what counts as a substantively important decision is itself an important question — with a general rule of thumb that this is determined by patients and their representatives. We suggest that greater honesty about when shared decision making is genuine, and genuinely required, and when it is not, is needed. We call for an acknowledgment that a ‘failure’ to implement shared decision making in every clinical encounter is not a failure at all — because one ought not to feel morally obligated to do something that is not possible. This is not to say we shouldn’t attempt to share information and responsibility, but the question of how far we ought to go to make fully shared decision making possible remains moot.
Notes
Provenance
Freely submitted; externally peer reviewed.
Competing interests
All authors are members of the RCGP Ethics Committee. The views expressed in the paper belong to the authors as independent persons, and do not represent the views of the RCGP or the RCGP Ethics Committee.
- © British Journal of General Practice 2018