What Is Physician Noncompliance?

I begin with a case to illustrate the problem I examine in this article:

This is a case of what I call physician noncompliance. As I define it in this article, it refers to a spectrum of oppositional, rule-breaking stances by physicians that are nonviolent and morally motivated.Footnote ¹ More specifically, they are motivated by moral concern or outrage at what they take to be injustice or unfairness that negatively affects the care their patients receive. The injustice or unfairness may reside in particular rules, in the administration of morally unobjectionable rules, or in global structural features of the healthcare system. In addition, physician noncompliers are motivated by the conviction that in breaking the rules they are adhering to a fundamental principle of professional medical ethics and acting as patient advocates.

Although there are various types of principled noncompliance, including civil disobedience and conscientious refusal, I am chiefly concerned here with what Rawls calls “conscientious evasion,”Footnote ² an example of which is the preceding case. Conscientious evasion refers to covert acts of noncompliance done for reasons of conscience or principle. Physicians who perform acts of conscientious evasion are not engaged in public protest, and they may entertain no expectation or hope of changing laws or policies. In fact, they do not assume their actions are known to the authorities, and they are generally actively engaged in trying to avoid detection. Conscientious evasion may be the most common type of noncompliance among physicians, and arguably it is the most morally problematic one as well.

I begin with an argument in defense of physician noncompliance that I call the argument from professional role definition. Although a common line of reasoning for the argument from professional role definition has some initial plausibility, on closer examination it is unconvincing. The following section presses the case more forcefully against physicians by offering two arguments: one consequentialist and the other an argument based in justice, commonly known as the free-rider objection. These arguments set out some of the moral hurdles that have to be overcome before physician noncompliance can be justified. They are taken up again in the penultimate section, in the context of a discussion of general conditions for the justification of physician noncompliance. In the final section, I return to the opening case.

An Argument in Defense of Physician Noncompliance

The scenario at the opening of this article describes a type of rule-breaking conduct undertaken in response to rules that constrain the ability of physicians to protect and advance the well-being of their patients. What considerations speak in favor of acts like these, and is there a conception of physician advocacy that permits or countenances them?

Consider the following argument. Noncompliance is sometimes morally permissible, although perhaps not obligatory, because there are rules of the distributive system and rules and norms of the physician-patient relationship, and the two sets of prescriptions do not always coincide. This is the argument from professional role definition. Because the physician’s primary loyalty is to her patients, physicians are permitted to do things to aid their patients that might be unacceptable from the standpoint of the rules and procedures of the distributive scheme. The argument does not merely purport to justify the physician basing her actions on considerations that might be irrelevant from the standpoint of the system. Rather, it purports to justify her acting in ways that would not be approvable from that standpoint. Note also that the argument as stated does not require that the scheme of resource allocation be unjust. It might not be, but physician noncompliance could nonetheless sometimes be justified because of the different levels at which the two sets of rules and norms operate.

To bolster this argument, one might appeal to another kind of role differentiation, one that characterizes legal advocacy in an adversarial justice system. The point of the comparison to legal advocacy is not to suggest that physicians are or should be adversaries acting on behalf of their individual patients. Indeed, there are good reasons why they should not be.Footnote ³ Rather, the suggestion is that if we model physician advocacy on the sort of aggressive, zealous advocacy that lawyers are expected to exercise on behalf of their clients, we might see more clearly how to defend a conception of physician advocacy sufficiently robust to encompass physician noncompliance under some circumstances.

Consider Richard Wasserstrom’s description of one common view of the advocacy role of the lawyer: “In the course of defending an accused, an attorney may have, as part of his or her duty of representation, the obligation to invoke procedures and practices which are themselves morally objectionable and of which the lawyer in other contexts might thoroughly disapprove.”Footnote ⁴ For example, it may be appropriate, even obligatory, for an attorney to engage in manipulative tactics within the courtroom, as part of his or her advocacy role. Indeed, a conception of the advocacy role of the lawyer that includes such role-specific permissions and obligations is embraced by many lawyers on principled grounds. Perhaps we should not expect the physician to adhere to the normal rules of fair play any more than we should expect the legal advocate to adhere to norms of ordinary morality, and perhaps this is for similar reasons in the two cases.

To assess whether the analogy can be useful in defending the sort of physician noncompliance at issue here, we need to understand how zealous legal advocacy is typically defended in our legal system. According to the traditional view described previously, the lawyer is a pure legal advocate who may, and indeed is required to, do things on behalf of his client that would in ordinary circumstances be regarded as morally objectionable. An alternative and, in the view of many in the legal profession, preferable view holds that a good lawyer is one who is effective not only in legally advocating for his client’s cause but in morally advocating as well, so he is considerably less free in his advocacy to disregard ordinary moral considerations that nonprofessional citizens might take to be important.Footnote ⁵ On neither conception is the justice of the adversary system within which the lawyer operates an irrelevant consideration in the justification of the lawyer’s conduct. Rather, zealous advocacy on behalf of one’s client is commonly defended on rule-utilitarian grounds as an essential part of a legal system in which truth is most likely to be discovered and justice most likely to be done if lawyers advocate zealously for their clients. According to the standard justification, that is, zealous advocacy is necessary for the proper and effective operation of a system of justice, a system that seeks to discover the facts, to protect the legal rights of the accused, and to afford the maximum possible protection for those who are innocent, which is particularly important in criminal trials. Moreover, the purposes for which zealous advocacy is thought to be necessary and that provide its justification set limits to what lawyers may permissibly do in representing the interests of their clients. The only permissible sorts of ways in which lawyers may zealously represent their clients’ interests are those that sustain and are allowed by the existing system, assuming it is just, because zealous advocacy derives its justification from the role it plays within a system that professes to be devoted to the pursuit of justice. Legal advocates do not have unrestricted discretion to do whatever advances the interests of their clients.

To return to physician noncompliance, we can see that legal advocacy provides no reason to think that among the sorts of actions that physicians may engage in on behalf of their patients are acts of noncompliance with system rules that regulate physician conduct, whether or not they are evasive. It is no part of the conception of legal advocacy defended in the aforementioned way that lawyers are permitted to break the rules to promote justice for their clients. On the contrary, they have a duty to uphold the legal system as well as to zealously represent the interests of their individual clients, because it is only as permitted by the rules of the system that justice is to be pursued for them. Although superficially the analogy with legal advocacy seems promising, further reflection shows that it breaks down. A comparison with the role-based advocacy of lawyers fails to support the argument from professional role definition for physician noncompliance, because it is one thing to justify role definition on the grounds that it maintains the integrity of a system of rules, and something quite different to appeal to role definition to justify breaking the rules.

If reasons similar to the ones supporting zealous legal advocacy in an adversarial system do not vindicate a permission, let alone obligation, for the physician to engage in acts of noncompliance on behalf of his or her patients, they do suggest something important about physician advocacy and the meaning of “zealousness” in this context. The suggestion is that there can be ample space for physicians to aggressively promote the interests of their individual patients, while still respecting the rules that determine how benefits are to be distributed to them as well as to others (including other patients of theirs) who are subject to the rules. The scope of zealous physician advocacy, like that of zealous legal advocacy, may be limited to means that are compatible with the purposes of the system within which they operate, that is, with the just distribution of the relevant goods (medical or legal). However, the constraints leave physicians with a range of options for how to meet their patient’s needs, including acts that, although compliant with the rules, test their limits.

One way that physicians can advocate within the constraints of rules is by taking advantage of the indeterminacy of the language in which the rules are expressed. For example, rules for insurance coverage use terms like “medically necessary” and “medically appropriate,” terms that are inherently vague and whose implications are to some extent indeterminate. The authoritative language in which the rules are expressed guides only in an uncertain way, so a choice has to be made between different interpretive possibilities at the point of actual application. Physicians who exercise discretion in interpreting the general language in which a rule is expressed are not necessarily breaking the rule, and so not necessarily failing to comply with it. Rather, they may only be selecting, among the variety of ways of reasonably construing the rule, the interpretation that makes the best possible case for their patients.

Negative Consequences and Free Riding

An objection that is likely to be raised to some kinds of rule breaking applies a universalizing test: however well intentioned your actions might be, what if everyone did what you are doing? What if every physician were to break the rules because they believed their patients were being unfairly treated? The questions are intended to leave no doubt that something quite terrible would happen.

E. Haavi Morreim, in her oft-cited article on gaming the system,Footnote ⁶ takes such questions seriously. She criticizes gaming in part by asking us to take note of what the consequences would be if large numbers of physicians did it. They would be dire, she says, for “no resource system can long survive widespread abuse and dishonesty”; “gaming, if widespread enough, can destroy any system of resource rules.”Footnote ⁷ Also hypothesizing its extensive use, Catherine Regis criticizes gaming on the grounds that “if such a practice becomes prevalent, patients, whose physicians refuse to do so, will be penalized.”Footnote ⁸ The conditionals (if everyone, then . . .) may well be correct, but the obvious response to criticisms based on them is that these doomsday scenarios are so unlikely to materialize as not to be reasonably entertained. After all, breaking the rules is not completely risk free. Even if physicians who conscientiously break health-related rules to help their patients believe that the risk of exposure is small, discovery cannot be ruled out entirely. And if they are discovered, their professional standing and their ability to help other patients may be in jeopardy. They may also lose their license to practice medicine or be reprimanded in some other official manner. Moreover, it may be that most physicians simply don’t feel comfortable breaking the rules and wouldn’t do it, not only because they fear the repercussions but because it doesn’t feel right to them, even if they can’t fully articulate why. Perhaps the combination of self-interested motives and powerful moral inhibitions would be sufficient to steer most physicians away from rule breaking, or if they are going to break the rules, to induce them to do so only infrequently and very selectively. As long as instances of noncompliance are isolated and it remains a marginal activity, the actions of a few outliers shouldn’t make a difference to the survival of the system of resource rules within which the noncompliance takes place.Footnote ⁹

The aforementioned assumes that the “What if everyone did it” (i.e., conscientiously broke the rules) argument is intended to persuade us of the wrongness of rule breaking by describing the consequences of its widespread practice. Understood in this way, the argument loses its moral force in the face of strong evidence that everyone will not in fact do it.Footnote ¹⁰ But this may not be how the argument is intended to work. Perhaps the point of asking the question is not to imply the prediction that everyone will or is likely to break the rules, but rather to suggest that even if others complied with the rules, breaking them would still constitute a particular sort of wrong. That is, the point of appealing to a universalizing test may be that, in order for rule breaking by some physicians to succeed, at least most physicians must continue to adhere to, and be known to adhere to, the rules. Successful rule breaking by the few, therefore, is parasitic on rule conformity by the many. This is the basis of the free-rider objection, and it does not presume or allege that others will actually do what the rule breaker does.

The free-rider objection, to say more, is this. The physician rule breaker depends on the willingness of other physicians (and others in the healthcare system whose support of the rules is critical to their maintenance) to subject themselves to the requirements of the rules, without being willing to do so herself. The rule breaker is only able to help her patients by breaking the rules because others limit themselves to helping their patients by adhering to them: in helping her patients by rule breaking, the noncompliant physician arrogates to herself a privilege, while depending on the renunciation of that privilege by others. What makes this unethical is that it is unfair, that is to say, it amounts to giving her patients objectionably preferential treatment. Even if the conscientious rule breaker has a legitimate moral complaint against the rules or how they are implemented in a particular case, not every moral complaint is serious enough to outweigh the unfairness of arrogating to herself the privilege of rule breaking.

We should note, however, that not all instances of physician noncompliance raise a free-rider objection. Free riding is unethical because it violates what Rawls calls the principle of fairness, and not all instances of physician noncompliance take place against a background to which this principle applies. The principle is formulated as follows:

The free rider unfairly helps herself to the benefits of a cooperative scheme, so in situations in which the rules that the noncompliant physician violates are not part of such a scheme, free riding, as a conceptual matter, is not possible. This is the case, for example, with the physician who refuses to comply with institutional rules forbidding physicians from telling their patients when they have made medical mistakes in their care. Her noncompliance with the rules seems commendable. But whether or not it is, this is not the sort of situation to which the free rider objection might apply; the physician doesn’t take advantage of others’ conformity with the rules in order to be able to break them herself.

Still, the charge of free riding is something that has to be addressed in many cases of physician noncompliance. Sometimes the charge can be rebutted, because giving preferential treatment to one’s patients is not always objectionable, or objectionable all things considered. This depends on the sort of moral objection that the physician can raise to the distributive scheme with which he is noncompliant. If the cooperative scheme itself is seriously unjust, or if the cost to her patients and to other patients in the system of compliance with the rules is excessive relative to the benefit from compliance with them, then the physician’s refusal to comply with the rules, and her decision to break them to help her patients, may be morally permissible, perhaps even morally obligatory.

There is one objection to this that we can set aside rather quickly, namely, breaking the rules, especially by means of conscientious evasion, is not a solution. What is it not a solution to? If the problem is how to replace the current scheme with a morally superior one, then this may or may not be a solution. However, if the problem to which the physician seeks a solution is how to help her patients who are being unjustly treated now, questions about how the system or its constituent rules can be changed may be beside the point. Desirable though changes in the distributive scheme may be—and physicians who break the rules to help their patients can fight for these as vigorously as those who do not—these changes might only be realizable over the long term. In the meantime, her patients are being treated unfairly, and, as I have assumed, it is their current plight that chiefly motivates the physician’s rule breaking. The thought motivating the rule breaking, presumably, is that it will be much too late for these patients, these patients for whom I am currently responsible, if they have to wait for the rules to change. The questions that need to be answered, therefore, are whether she is justified in thinking this and, if she is, whether this justifies her particular mode of noncompliance. If she is wrong about the timeline for change, which in fact is not a distant prospect, the case for rule breaking or her particular form of it will to that extent be weakened.

The Justification of Physician Noncompliance

There is one justificatory burden that any physician contemplating breaking health-related rules for her patients must confront, at least if the rules are not so thoroughly unjust that there is no obligation to obey them in the first place. The burden, in very general terms, is to show that in the particular case in which noncompliance is contemplated, the ethical norm or obligation of physician advocacy trumps the obligation of fidelity to the rules. How difficult it is to meet this burden, or whether it can be met at all, depends on several considerations.

For one thing, this burden is more or less easily met depending on the conception of physician advocacy one adopts. For example, the free-rider objection to physician noncompliance is easier to rebut under what Norman Daniels calls the unrestricted ideal advocate conception of physician advocacy than the ideal advocate conception.Footnote ¹² For if, as the former holds, there are no external constraints on what treatment physicians may provide their patients, then there are no reasons grounded in justice to have moral concerns about whether in helping one’s patients one is free riding on the rule-compliant behavior of others. This, however, is a sufficient reason to reject the unrestricted ideal advocate conception. Any plausible conception of physician advocacy cannot be morally indifferent to whether the physician is giving her patients unfair advantages, that is, objectionably preferential treatment. Even under a plausible conception, the justificatory burden on physicians contemplating noncompliance is not equally weighty in all situations. Rule breaking that amounts to a minor departure from the rules, what some might prefer to call rule bending rather than rule breaking, might be more easily justifiable than serious rule breaking. At the same time, noncompliance to resist a serious injustice might be easier to justify than noncompliance that resists a slight one.

Let me say more about how this burden can be met. It is a minimally necessary condition of justified noncompliance that the method of noncompliance be suitable to the end and the end to the method, and that the end be realizable by the means chosen. Different methods of noncompliance have characteristically different ends, and employing a method to achieve an end that is not characteristic of it is problematic. If the method cannot plausibly be supposed to advance the end, then either the end or the method of noncompliance should be rethought. So, for example, if the reason for noncompliance is to bring about change in rules or policies, not only is conscientious evasion ill advised and unlikely to be successful, but to the extent that it is ineffective it is likely to be morally unjustified as well. It would not be morally unjustified were conscientious evasion a totally innocuous act that did not rise to the level of a morally assessable act. Were this the case, conscientious evasion might be a waste of time, but it could not be unethical. Conscientious evasion is not like this, however. The rules that it breaks may have some legitimate regulatory purpose, and in seeking to benefit one, it may impose costs on others.Footnote ¹³

There is, of course, more involved in the justification of physician noncompliance than the congruence of method and end and the achievability of the end. The end itself, whatever the method of noncompliance, must be a worthwhile one that protects or advances morally legitimate interests, specifically, interests in just or fair treatment.

Moreover, in general noncompliance is not justified if there are ways of securing these legitimate interests by taking advantage of opportunities for revision and improvement provided by the rules themselves. This is the meaning of the “only as a last resort” condition usually mentioned in connection with civil disobedience,Footnote ¹⁴ but it applies as well to the type of noncompliance I am discussing here. Securing medical services for one’s patients by breaking the rules, assuming this is a morally principled act, amounts to a statement by the physician that political or institutional procedures cannot be relied on to protect her patients from the consequences of injustice or unfairness or to get them what they are entitled to as a matter of justice or fairness. In a system in which these procedures are normally efficacious, rule breakers bear a heavy burden of justification.

Further, in order to justify noncompliance, the free-rider objection will have to be disarmed. One way to do this is to show that the free-rider objection gets no traction, because there is no cooperative distributive scheme within which the rule-breaking takes place. However, many, perhaps most, instances of physician noncompliance in the current healthcare regime are protests against various limits placed on access to care by managed care organizations or government programs, and these cases involve a cooperative distributive scheme of some sort. Here the free-rider objection cannot be dismissed by pointing out that the conditions for its application have not been met. To justify noncompliance in these cases it is necessary that the injustice from which the patient suffers be substantial and clear. In addition, the patient’s situation must be pressing. That is, it must allow no time for the physician to help her patient by making a case for him by some other means, whether by complying with the rules or by being more openly noncompliant. Although being a free rider is pro tanto wrong, its wrongness can be outweighed by other moral reasons in favor of the act, including, as necessary conditions, the severity and transparency of the wrong and the urgency of addressing it.

Consequentialist considerations are also part of the moral calculus, whatever kind of noncompliance is entertained. Some of the negative consequences were noted earlier in discussing the question, What if everyone did that? The method of noncompliance might match the end sought, and the end might be achievable by the means chosen, but only at the cost of considerable collateral damage. Acts of noncompliance might have significant negative consequences for the provision of fair healthcare to all who depend on the system whose rules the physician breaks. There are other consequences as well, consequences for the patients on whose behalf the physician is breaking the rules and for the physician herself. Noncompliance, if it becomes known, can create distrust and resentment among those who are adversely affected by it, and this can set back the interests of her patients, even those whom the physician did not attempt to aid by noncompliance. And evasive noncompliance can have corrosive effects on the character of the physician herself. In short, the consequences are evaluated along different dimensions and involve multiple participants in the healthcare system.

Finally, I want to comment briefly on the moral significance of noncompliance for the physician, because this is one of the considerations that should be factored into the complex moral calculus that determines the justifiability of noncompliance. There is a self-referential component to some physician noncompliance that consists in a moral concern to avoid complicity in injustice. Margaret Little characterizes complicity as follows: “One is complicitous when one endorses, promotes, or unduly benefits from norms and practices that are morally suspect.”Footnote ¹⁵ Concerns about complicity arise most acutely when one’s moral and/or professional integrity is believed to be imperiled by it, which is not to imply that one’s aim in avoiding complicity is to safeguard one’s integrity.Footnote ¹⁶

One is complicit in injustice when one acquiesces to it although means of resistance are possible, likely to be effective, and at a cost that is not prohibitive. Acquiescence to injustice under these circumstances is a tacit endorsement of it. However, there is no easy answer in many cases to the question whether nonresistance to an unjust status quo constitutes objectionable complicity in injustice. Nor, as I hope to have shown, is it a simple matter to decide how one should go about challenging it if one wants to avoid complicity. There may be considerable uncertainty about the efficacy and consequences of one’s actions and about the sorts of compromises with injustice that are compatible with the preservation of one’s moral and professional integrity, an integrity that, for physicians, is rooted in the commitment to patient welfare.

Avoiding complicity in injustice is not limited to engaging in acts of noncompliance on behalf of one’s patients. The injustices from which one’s patients suffer may be rooted in morally suspect norms to which the profession of medicine, or some influential part of it, has lent its support or that it has not opposed and from which it and its practitioners have benefited. There may also be injustices that the profession has condemned but that remain. In general, avoiding complicity in wrongdoing involves, as a baseline, understanding that the norms and practices responsible for it have contributed to making noncompliance an option that at least deserves serious moral consideration, if not endorsement. A physician may then decide to engage in some form of rule breaking in order to act on this understanding and express her refusal to be complicit. But even if she decides not to do so, there is a range of actions, apart from rule breaking, from which physicians can choose in order to express, as Little puts it, a “stance of fighting the norms.”Footnote ¹⁷ If a physician wants to avoid the moral taint of complicity in injustice, then plainly she ought to do something to fight the norms, and her circumstances, the resources at her disposal, and the seriousness of the injustice will determine what she can and should do.

The physician’s desire to avoid complicity in injustice that harms her patients is an important moral motive that helps explain why physicians sometimes push their advocacy role farther than they would ordinarily take it, and why they are sometimes prepared to take non-negligible risks on behalf of their patients who are unjustly or unfairly treated. The importance for the physician of avoiding complicity in injustice that harms her patients should also be acknowledged as one factor that bears on the justifiability of physician noncompliance. Complicity threatens the moral and professional integrity of the physician, and noncompliance may be warranted in part because it is the only way that a physician can meet the threat.

The Medicaid Case Revisited

To conclude, I want to return to the case at the beginning of this article and ask whether the physician’s action on behalf of her patient is morally justified. Note first that I characterized her action as a falsification. Were it instead an instance of a physician doing the best she can for her patient by reasonably interpreting an indeterminate rule in his favor, this would not be a case of noncompliance at all. Rather, it would be just what we expect physicians who are advocates for their patients to do. The Medicaid rules under which the physician is operating, however, are quite clear in excluding from coverage patients like the one in this case.

Insofar as this act of noncompliance seems especially morally troubling, it is, I think, because of the concealment, as well as the deception the physician engages in to accomplish it. In general, the sort of conscientious evasion physicians engage in is problematic for a number of interrelated reasons. First, because the noncompliance is secretive and evasive, the only check on the physician’s actions is likely to be her own conscience, and this may not be entirely reliable. Second, because the act is done in secret, the physician may be prone to self-deception about her motivations for engaging in noncompliance. And third, the physician’s grasp on the difference between right and wrong may weaken if she permits herself to furtively break the rules when she judges this to be in her patients’ best interests.

Of course, it is still possible for an individual act of conscientious evasion to be justified in certain circumstances. But the bar for justification is set fairly high, in part because of the generally problematic character of such actions. First, whatever negative consequences noncompliance may have for other patients, not only other patients of the rule-breaking physician but the total group of patients in the system, must be tolerable in light of what is gained for the patient she is seeking to help. Second, even if the overall consequences are favorable to rule breaking, the physician may be breaking the rules of a mutually advantageous cooperative venture, in which case additional elements are required. In particular, the harm to which her patient is exposed by her adherence to these rules should be substantial and undeniable; the patient’s need for the treatment he is denied under existing rules and procedures should be urgent; and alternative means of securing this treatment that do not involve noncompliance must be unavailable or too slow to achieve results in a timely fashion. Given only the facts of the Medicaid case as presented, we can’t tell whether these conditions are satisfied. But at least we know what else we would need to know about the case in order to justify this act of noncompliance.