Can “Giving Preference to My Patients” be Explained as a Role Related Duty in Public Health Care Systems?

Abstract

Most of us have two strong intuitions (or sets of intuitions) in relation to fairness in health care systems that are funded by public money, whether through taxation or compulsory insurance. The first intuition is that such a system has to treat patients (and other users) fairly, equitably, impartially, justly and without discrimination. The second intuition is that doctors, nurses and other health care professionals are allowed to, and may even in some cases be obligated to give preference to the interests of their particular patients or clients over the interests of other patients or clients of the system. These two intuitions are in potential conflict. One of the most obvious ways in which to ensure impartiality in a health care system is to require impartiality of all actors in the system, i.e. to give health care professionals a duty to treat everyone impartially and to deny them the ‘right’ to give their patients preferential treatment. And one of the possible side-effects of allowing individual health care professionals to give preference to ‘their clients’ is to create inequality in health care. This paper explores the conflict and proposes that it can be right to give preference to ‘your’ patients in certain circumstances.

Introduction

Most of us have two strong intuitions (or sets of intuitions) in relation to fairness in health care systems that are funded by public money, whether through taxation or compulsory insurance.¹

The first intuition is that such a system has to treat patients (and other users) fairly, equitably, impartially, justly and without discrimination. The system essentially has to live up to a requirement to provide equal service for equal need.²

The second intuition is that doctors, nurses and other health care professionals are allowed to, and may even in some cases be obligated to give preference to the interests of their particular patients or clients over the interests of other patients or clients of the system.³

These two intuitions are in potential conflict. One of the most obvious ways in which to ensure impartiality in a health care system is to require impartiality of all actors in the system, i.e. to give health care professionals a duty to treat everyone impartially and to deny them the ‘right’ to give their patients preferential treatment. And one of the possible side-effects of allowing individual health care professionals to give preference to ‘their clients’ is to create inequality in health care.

This potential conflict can be further sharpened by the possible claim that health care professionals have an independent moral obligation to treat everyone impartially and not to discriminate in favour of anyone. If we parse ‘discriminate’ as ‘giving unjust preference’ then it follows straightforwardly that health care professionals should not discriminate. But it follows as an analytic truth in the same way that, for instance murder is always wrong, simply because murder is unjustified killing. And it is, as an analytic truth not particularly interesting. It still leaves us with the task of locating the unjustifiable injustice.

If we parse ‘discriminate’ more neutrally as ‘giving preference to’ then it is an open question whether it is true that health care professionals should never discriminate and an open question what the class of people is among whom they should not discriminate.

In the Danish Physicians Oath from 1815 we can for instance find the promise that “… I will always as conscientiously care for the poor as for the rich without concern for status…” (my translation),⁴ but this was not originally taken to imply that a physician who happened to have a sufficient number of rich patients did anything wrong by caring for them and not for the many poor people who were not his patients. It was only if he treated two of his patients differently because one was rich and one was poor that he was seen as doing wrong.

Finally, we could parse ‘discriminate’ as ‘showing a preference against’. On this reading it is again reasonably clear that a health care professional should not show a preference against any patient in the health care system. But this does not settle the matter against giving preference to some patients, because it is not obvious that a health care professional who gives preference to her own patients thereby necessarily shows a preference against those patients that are not hers. These other patients may be spatially and temporally distant. Helping the patient that is in front of me now without thinking about a, to me unknown patient with the same condition somewhere else is not showing a preference against the, to me unknown patient in any straightforward way. By reflecting on the totality of patients I may of course come to realise that I or the health care system have obligations towards them all and may be brought to reflect on the content and strength of these obligations, but that is a different matter.

It is important to note that whereas the core meaning of the verb ‘to prefer’ is comparative, as in ‘I prefer apples to pears as an afternoon snack’, it is often used elliptically as in ‘I prefer apples’ where both the comparator and the comparative context needs further explication. It is only by further explication of the elliptical statement that we can ascertain whether the person who prefers apples prefer them to say pineapples and in what context and with what consequences this preference is likely to be acted upon.

In this paper I will analyse a particular way of dissolving the tension between the two intuitions mentioned at the beginning. The strategy will be to argue that whereas the system has to be impartial, just, fair etc. this is best achieved if health care professionals are given specific role related permissions to be partial in relation to their patients.

The reason for pursuing the particular argumentative strategy is that it is to a considerable extent theory free. If it is successful we can justify provider partiality without committing to a theoretical account that allows for or justifies partiality at a more fundamental level, such as an ethics of care, a particular account of agent relativity or some version of a proximity ethics. This is an advantage because it allows agreement on the substantive issue of partiality in health care encounter even among persons with differing theoretical commitments.

Impartiality, Arbitrariness and Discrimination

Before moving on to the constructive part of the argument it is necessary to clear some conceptual and theoretical ground. It might be argued that the project pursued here is bound to failure because we have (strong) reasons for requiring impartiality at the interpersonal level that are unrelated to the goals of the health care system. Impartiality is simply a foundational principle of ethics (or very close to being one). This is, for instance exemplified in the consequentialist dictum that “everybody to count as one, nobody as more than one”.⁵ But even strict consequentialists have to find ways to accommodate the intuition that some kinds of partiality are not only acceptable, but perhaps even preferable. The standard example being the partiality of parents towards their own children.⁶ Consequentialism becomes a much less attractive theory if such partiality cannot be justified. The consequentialist solution to this problem is to assert (without much supporting evidence) that allowing this particular type of partiality optimises good consequences. As soon as this move is made the gate is opened for any other kind of partiality that can also plausibly be claimed to produce net good consequences, including, I submit the partiality of health care professionals towards their patients.

Kantian universalisation is not of much help to us either in this context since both the maxim ‘no one should show partiality under any circumstances’ and, for instance the strong maxim ‘parents should show partiality towards their children’ as well as the slightly weaker ‘parents are allowed to show partiality towards their children’ seem to be validly universalisable. And the other famous formulation of the categorical imperative in terms of treating others as ends in themselves, not merely as means also seems open to certain forms of partiality.

And finally it is generally recognised that partiality is acceptable in cases where two parties commit themselves to partiality for instance through an explicit or implicit contract. In US jurisprudence on the doctor-patient relationship it is, for instance assumed that when a doctor-patient relationship is established the physician has to act as a fiduciary for the patient. This is in the context of a non-public health care system, but is never the less illuminating for the discussion here. According to Furrow the content of the physician’s obligations can be explicated in the following way:

“A fiduciary obligation in medicine means that the physician focuses exclusively on the patient’s health; the patient assumes the doctor’s single-minded devotion to him; and the doctor-patient relationship is expected to be free of conflict” [3].⁷

There are two things it is important to note in relation to this traditional account of the doctor-patient relationship. The first is that any workable morality must allow for some kind of commitment to partiality (e.g. in many types of promises) and the second that the possibility of the doctor to credibly commit to a fiduciary relationship may be a precondition for the patient to trust that the doctor is single-minded in his devotion to the patient’s interest. We worry a great deal about conflicts of interest in medicine and medical research especially in relation to whether the pharmaceutical industry are seducing health care professionals, but from the point of view of the patient it may not really matter why I can’t trust that my doctor is devoted to my interests or how his conflict of interest has been created. I may not care whether he is a servant of the state or a servant of the capitalists. What matters to me is that I can no longer rely on him but have to protect my interests myself.

Role Related Permissions: Some Examples

There are many situations where we allow actors in social systems to transgress prima facie ethical norms in order to achieve the overall goods of the system. We allow boxers to assault each other in the ring in ways which would be ethically problematic and in many jurisdictions illegal if done outside of the ring, even with mutual consent. This is permitted because the activity of boxing is thought to make possible the achievement of some higher goal, e.g. a certain aesthetic pleasure or a certain display and development of manly virtues.⁸

Perhaps more relevant for our analysis of the current issue is the fact that we have in many countries instituted an adversarial criminal justice system where the prosecutor and the defender, both trained lawyers have been allocated roles that allow them not to be impartial. The defender is allowed only to emphasise evidence that benefits the defendant, to use highly questionable techniques for examining witnesses and not to mention evidence that may jeopardise the case for the defence. Defenders are not allowed to lie in court, i.e. to state something as fact that they absolutely know not to be the fact, but they are allowed and in some sense expected to try to mislead the court by stating the facts selectively; and vice versa for the prosecutor.⁹

The reason that we have institutionalised such a seemingly strange system is partly historical, partly because we believe that institutionalising two adversarial roles is in the final analysis conducive to the discovery of the truth and thereby to the doing of justice.

Role Responsibility in Health Care

Can similar arguments be mounted for provider partiality in health care? Let us first note that we do in general expect health care professionals to pursue the best interest of their patients. In situations where there are no resource constraints a health care professional would be negligent if s/he did not provide the best possible treatment for the patient’s condition. But situations without resource constraint are special because they only occur when my treatment of one patient does not affect any other patient in the health care system.¹⁰

But there are also a range of other situation where we expect (or at least allow) health care professionals to be partial. We expect health care professionals to engage in debates about priority setting as representatives of their group of patients; and in cases where a patient’s case has to be presented to others in relation to a treatment decision with priority implications we expect the patient’s health care professional to present the case in the best possible light and act as an advocate for the patient. In the latter context we might not allow the degree of partiality we allow in the legal system, but we do as a matter of fact allow considerable partiality.

The more specific question we therefore have to answer is what degree of partiality, if any we can justify as promoting the goals of the health care system.

The question of what the proper goals of the health care system are is itself contentious. I have elsewhere argued that the goals of the health care system cannot be reduced to one single goal and will not repeat that argument here.¹¹ Let us therefore, for the sake of argument assume that the goals of the health care system are complex but include:

• Delivering high quality care to the whole population

• Delivering such care effectively

• Delivering such care equitably

• Engendering trust in the system, both with regard to overall performance and with regard to individual patient-provider interactions

Nothing much in the analysis hinges on whether this stipulation of goals is accurate or complete, except for the inclusion of other goals than equitable delivery of care. If the only goal of the health care system is the equitable delivery of care the following arguments may not be sound.

What is the best way of achieving these goals? What kind of health care system and what allocation of roles within this system are best suited to achieve the goals? Stated in this way it is probably an unanswerable question because it is unclear (1) whether there is a single “best way” and (2) whether we have the epistemic resources to find the answer to the question even if there is one. But it may be possible to give an answer in broad outline, especially if we assume that at least some of the features of the current health care system can be held constant and that we are not beginning with a blank organisational canvas.

The features I will assume as constant are:

• 1. That the primary responsibility for the care and treatment of a patient is allocated to one particular health care professional or one particular, relatively stable team of health care professionals

• 2. That health care decision making is not and cannot be micro-managed all the way down, i.e. that guidelines can never be made sufficiently specific to cover every or even most clinical decisions in detail

Let us move on to discuss whether two kinds of provider partiality can be acceptable in this context, that is: 1. partiality in relation to advocacy for patient groups and 2. partiality in relation to treatment of those a health care professional sees as his or her patients.

Effective advocacy for a patient group arguably requires partiality and as mentioned above we do as a matter of fact expect or at least allow health care professionals to act as advocates for ‘their’ patient group. Can this be justified in relation to the goals of the health care system? Let us first note that patient groups can advocate for their own interests and that we would need a very restrictive account of what might be called citizen impartiality to justify a claim that such advocacy is wrong. It is generally assumed that a citizen in a democratic society has a right to advocate for her own interests and also a right to organise with others to advocate more successfully. This has the inevitable side-effect that patient groups that can organise most effectively are also likely to be able to advocate most effectively for their interests and likely to get their interests better met, unless we believe against all available evidence that decision makers in democracy can successfully balance the ‘real’ interests and completely resist the advocacy.

Success organising as a patient group is dependent on a number of factors, but there are clearly patient groups that will have difficulty organising or difficulty organising in sufficient numbers. The sufferers from appendicitis will not organise because their problem will be solved shortly after it is diagnosed, patients with severe dementia will not organise because they are unable to and patients with very rare diseases may only be able to organise trans-nationally if at all simply because of low numbers.

If that is true it seems compatible with the goals of health care that health care professionals act at advocates for patient groups that are badly organised, or whose conditions are not ‘popular’. Such advocacy can in many contexts be directly equity promoting and will just be justifiable even if equity in health care is the only or main goal of the health care system.

It also seems compatible with the goals of health care to advocate for a patient group if the quality of care falls below a threshold of acceptable quality. What is acceptable quality cannot be fixed without reference to the rest of the health care system, but if a reasonable case can be made that a patient group receives treatment or care that is of a significantly lower quality than that available to other groups in the system it would be perfectly justified for a health care professional to advocate for better care for that particular group and thus to evince partiality in the advocacy.

Let us move to the more contentious case of partiality in relation to the care and treatment of individuals. As I have already alluded to previously some degree of partiality may be necessary to engender a successful and maximally effective patient-professional relationship. It is important for choosing the best and most effective treatment that patients provide full information to health care professionals about their symptoms and circumstances, and that they answer questions both truthfully and (reasonably) exhaustively. But a patient may only feel able to engage with and disclose fully to a health care professional if he feels that this will not endanger his interests. If the health care professional is simply ‘an agent of the health care system’ or a Lipskyan ‘street level bureaucrat’¹² and if patients accurately perceive this they may engage with the health care professional in exactly the same way as they engage with other bureaucrats, i.e. they may act strategically or tactically in order to get their interests promoted.

The professional cannot at the same time be the agent of the patient and the agent of the system. Any double agency creates potential conflicts of interest, and it is unclear that the best way to resolve this particular instance of double agency is by removing the professional’s responsibility to the patient, or by giving the professional the task of balancing these interests.

This is compatible with the professional working within a set of guidelines, for instance guidelines specifying the criteria for access to particular treatments. A professional who was conscientious in making sure that all of her patients who fulfilled the criteria got the treatment and did this without any thoughts concerning the overall resource use this entailed, or whether her colleagues were equally conscientious, would show partiality, but arguably in an acceptable way.

In a public health care system it is clear that some body has to take responsibility for resource allocation but this seems to be a task much better placed at a higher level of the system than the level of the individual health care professional.

The account of reasonable partiality that has been developed here is denied by Bærøe in a recent paper in which she claims that all clinicians are always responsible for:

“a. horizontal equity regarding equal treatment of cases considered equal within the clinician’s own patient population

b. vertical equity regarding discrimination between the needs of the clinician’s own patients

c. contributing to vertical equity by unequal treatment of unequal cases within the whole patient population of a healthcare system

d. contributing to horizontal equity by equal treatment of equal cases within the whole patient population of a healthcare system”.¹³ (p. 491)

She develops these four obligations from the clinician’s two-fold responsibility (i.e. to the patients and to the system) and the existence of both horizontal and vertical equity demands. As it should be obvious from the above I agree with Bærøe that health care professionals have all of these four obligations, but I disagree concerning (1) whether the overall aim of equitable health care is best met by pursuing them all, equally in all health care encounters and (2) the extent to which always pursuing them all is the right thing to do in a health care system with multiple goals, only one of which is equity.

Finally, it is perhaps also worth noting that we have little reason to believe that imposing formal equality all the way down by requiring health care professionals to act completely impartially will actually lead to actual equality. It might do if we could micro-manage each and every health care decision. But in the real world we have reason to expect that well educated, canny and socially resourceful patients are still likely to get a better deal. How big an improvement, if any in actual equality we would get by requiring impartiality ‘all the way down’ is an empirical question; but it is an important question because if we are not able to answer it we do not know whether our sacrifice of a better and probably more effective patient-professional relationship is repaid by significant improvements in equality.

Conclusion

In this paper I have argued that we may have reasons to permit health care professionals a carefully circumscribed area of partiality in relation to the patients they come into contact with and properly consider as their patients or clients.

The argument falls in two parts (1) a negative argument in the form of the rejection of arguments attempting to show that we have strong moral reasons to require strict impartiality of health care professionals and (2) a positive argument showing that it is plausible that the goals of the health care system are promoted if we allow some degree of partiality. However, the argument does not rely on any particular positive theory of acceptable partiality. The limits to acceptable partiality are not generated directly by moral theory but by considerations of effective delivery of high quality and equitable health care to all patients.