Fidelity and Deceit at the Bedside

To what extent should physicians act as their patients' advocates in the struggle to gain access to health care resources? Neither the Hippocratic ethical tradition nor contemporary bioethics nor US law speak clearly to this question.¹

The physician who recites the Hippocratic Oath pledges, "I will prescribe regimen for the good of my patients according to my ability and my judgment. . . ."² But the oath says nothing about what, if anything, its adherents should do to secure their patients' access to the "regimen" so prescribed. The dominant stream in contemporary bioethics subsumes professional obligation under 4 principles—respect for autonomy, nonmaleficence, beneficence, and justice.³ Fidelity to patients and advocacy on their behalf are derived from respect for autonomy and, thus, treated as taking their content, case by case, from patients' legitimate expectations. This, of course, begs the question of what expectations are "legitimate" when insured patients want all possibly beneficial care, regardless of cost. The law charges physicians with some duties of loyalty, including the keeping of confidences and avoidance of some conflicts of interest.⁴ But the law has not developed a duty of patient advocacy,¹ akin to the lawyer's duty of zealous advocacy on his or her client's behalf.⁵

The question of the scope and limits of physician advocacy vis-à-vis health care payers is a new issue for US medicine. Until about 1920, medical care was affordable to most working Americans without recourse to health insurers. The advent of hospital-based, technology-intensive care put prices out of reach of most Americans by 1930. Over the next 20 years, insurance for the cost of serious illness became commonplace.

Until the 1980s, insurers typically paid for care on demand when treating physicians certified need, and the question of professional advocacy vis-à-vis health care payers rarely arose. Only in the last 20 years have health plans intervened aggressively to restrain spending. Their principal approach has been to claim the authority to determine what is "medically necessary" and to withhold or preauthorize payment accordingly. As Wynia et al⁶ imply in their article in this issue of THE JOURNAL, health plans' detailed reimbursement rules are typically not part of their contracts with either patients or physicians. Almost without exception, these contracts commit plans to cover "medically necessary" services without defining this term. Indeed, plans often take the position that their coverage rules are trade secrets,⁷ not to be revealed to physicians or patients, to forestall lawyer-like advocacy within the rules' interstices.

Were medical necessity always a clear-cut matter, physicians and health plans would make "correct" clinical and coverage decisions except in cases of negligence or fraud. But medical need is often open to differing understandings, as our increasing knowledge about clinical practice variations demonstrates.^{8 - 9} The subjectivity of clinical assessment, scientific uncertainty about the efficacy of many treatments, and the impossibility of anticipating and writing rules for all clinical circumstances ensure conflict over medical need.¹⁰ Society's irresolution about whether costs should be weighed against clinical benefits—and, if so, in what manner—magnifies this conflict.

These sources of ambiguity leave much room for health plans to interpret subscribers' needs frugally and for physicians to represent their patients' interests vigorously. For at least the past 10 years, the lines of conflict over clinical resources have been clearly drawn, and the 2 sides have "gamed" the system in reciprocal ways. Health plans have promised "medically necessary" care without regard for expense, but they have used extracontractual coverage rules that economized, sometimes aggressively,¹¹ by trading off possible clinical benefits against costs. To varying degrees, physicians crafted their presentations to utilization managers with an eye toward gaining coverage. Particularly in cases in which clinical findings are subjective and ambiguous (or fall within the interstices of plan coverage rules), the possibilities for advocacy-driven interpretation and presentation of clinical data are great.

The study by Wynia et al⁶ in this issue presents the most reliable information to date on the extent of physician willingness to exploit these possibilities. Going beyond earlier studies, which used hypothetical scenarios to investigate professional norms influencing representations to health care payers,^{12 - 14} Wynia et al show that many physicians report regularly engaging in patient advocacy to the point of consciously deceiving payers. On the other hand, this study documents the persistence of deep divisions within the medical community concerning the relative weight of norms of fidelity to patients, truthfulness, and compliance with perceived contractual and social obligations.

The authors' findings concerning conscious deception are stunning. Of the 720 physicians who responded to their survey (64% response rate), 39% reported that they had "sometimes," "often," or "very often" "exaggerate[d] the severity" of a patient's condition, "change[d] a patient's official (billing) diagnosis," or "report[ed] signs or symptoms that a patient did not actually have." Of these physicians, 54% reported "using deception of third-party payers to obtain needed benefits" more often "now" (in 1998) than 5 years before, while only 12% said they did so less often. More than a quarter (28.5%) of all respondents agreed that "it is necessary to game the system to provide high-quality care."

However, the study by Wynia et al has important limitations and ambiguities. First, the term exaggerate suggests a clear line between accurate and misleading clinical reporting, a notion belied by the subjectivity and uncertainty of clinical perception and judgment. Within clinical gray zones of subjectivity and uncertainty, one physician's exaggeration is another's truth-telling. A confession of exaggeration to third-party payers requires consciousness of a difference between what one tells a payer about a clinical situation and one's understanding of the situation. To the extent that survey respondents who believe they tell the truth to payers are influenced in their clinical perceptions (and reporting to health plans) by preauthorization and reimbursement pressures, the responses to the survey question about exaggeration understate the distorting effects of these pressures on clinical reporting.

Moreover, the pejorative connotation of the term exaggerate gives short shrift to the possibility of legitimate, advocacy-oriented presentation of ambiguous clinical information. Different physician beliefs about the location of the line, if any, between advocacy-oriented presentation and exaggeration can lead to different responses to the survey question about exaggeration. Some respondents with a broad view of permissible advocacy-oriented presentation might be expected to deny engaging in exaggeration (on the ground that advocacy-oriented presentation is not exaggeration). Others with a similarly broad view of acceptable advocacy-oriented presentation might be comfortable characterizing such presentation as exaggeration—and thus could be inclined to admit making exaggerated statements to payers. Conversely, respondents with a negative view of advocacy-oriented presentation to payers might be expected to deny engaging in exaggeration. Responses to this survey question are thus confounded by diverse understandings of exaggeration—and of the acceptable scope of physician advocacy.

It is also not possible from the findings of Wynia et al to assess clearly the significance of a respondent's admission that he or she "change[d] patients' official (billing) diagnoses." To what extent did such changes reflect plausible reinterpretations of ambiguous clinical data and to what degree did these changes go beyond the bounds of reasonable bedside ambiguity in pursuit of payment or preauthorization? This question is central to assessment of whether any given change constitutes plausible patient advocacy or clear-cut deception.

By contrast, all physicians who admitted to "recording signs or symptoms that patients did not actually have" thereby confessed unambiguously to lying—and to breaking the law—to obtain payment or preauthorization. That 10% of respondents admitted doing this "sometimes" or more often is a remarkable finding. It is potent evidence that the professional norm of truthfulness is under pressure from the competing norm of fidelity to patients, although, as Wynia et al acknowledge, physician financial self-interest may be playing a role.

The fact that 54% of all respondents admitted to engaging in "deception" of payers more often in 1998 than 5 years earlier suggests that the norm of truthfulness is under mounting pressure from utilization management. Wynia et al sought to better understand this pressure by asking respondents whether "gam[ing] the system" is "necessary . . . to provide high-quality care" and "ethical" when done "for your patients' benefit." But the term game the system is an imprecise tool for exploring professional beliefs about the necessity and ethics of deception. For respondents, gaming might refer to anything from adroit clinical reporting to fit coverage rules (without distorting findings) to fabrication of signs, symptoms, or laboratory data. The former is consistent with the norm of truthfulness while the latter flagrantly violates it.

Physicians' varying understandings of what "gaming" means are sure to influence their answers to inquiries about its necessity and ethics. The fact that more than a quarter of respondents agreed that "it is necessary to game the system to provide high-quality care" (and that 15.3% agreed that "gam[ing] the system" for patients is "ethical") thus cannot be taken as proof that this many physicians countenance deception of payers. Indeed, to the extent that respondents interpreted "gaming" as a reference to patient advocacy without deception, these results could be taken as evidence that many physicians view such advocacy as neither necessary nor ethical. This uncertainty points to the need for more focused research into physicians' beliefs about the roles of particular advocacy practices (with and without deception) in their clinical work.

Advocacy without deception is consistent with a robust conception of the professional duty of fidelity to patients. But lying to health care payers to obtain preauthorization or payment crosses the line. The findings of Wynia et al that 10% of more than 700 surveyed physicians admit to fabricating signs or symptoms—and that 54% admit to "deception" of payers more often "now" (in 1998, when the survey was taken) than 5 years before—are deeply troublesome. The authors' findings concerning "exaggeration" of illness severity and physicians' attitudes toward "gam[ing] the system" also raise concern, although, for reasons just discussed, they are more difficult to interpret.

What accounts for this willingness to deceive? A clue lies in the medical profession's historical reliance on informal social norms—norms inculcated during clinical training^{15 - 16} and reinforced through fraternal interaction with professional peers.¹⁷ Norms imposed from outside the medical community, whether by law or by other sources of authority, until recently have had relatively little role in the governance of medicine. Obligations assumed by contract until recently have had similarly little role, although US law has for at least a century treated the formation of a patient-physician relationship as a matter of contract.¹⁸

In this regard, medical practice has been akin to other endeavors engaged in by close-knit groups.^{19 - 20} Such groups tend to prefer their own custom-tailored norms (and systems of enforcement) to the one-size-fits-all rigidities of the law, and legal authorities tend to defer to this self-governance unless the parochial norms of the group clash with wider social concerns.¹⁹ Group members, moreover, tend to see formal, arms-length contracting, and duties thereby assumed, as corrosive of fraternal feeling and trust.

Until the last several decades of the 20th century, medicine remained a largely self-governed enterprise.²¹ The ethic of fidelity to patients was a centerpiece of this self-governance²² and was widely regarded as desirable from a social welfare perspective.²³ But the transformation of health care payers from passive poolers of risk to aggressive managers of cost has injected outside actors into the governance of medicine to an unprecedented degree. Obligations and norms imposed by law or formally adopted by contract now vie with long-standing professional values.

The public's contradictory expectations of medicine magnify this conflict. As taxpayers and private-sector stakeholders, Americans want medical costs contained. But as patients and as friends and family of the sick, Americans demand access to the finest care and expect their physicians to be faithful, uncompromising agents. The remarkable finding of Wynia et al that 37% of physician respondents said their patients "sometimes, often, or very often" asked them to deceive health care payers reveals how ardent—even intemperate—this expectation can become.

Physicians today are caught up in the moral turbulence created by these conflicting norms, obligations, and expectations. The medical profession lacks what Ellickson has called "controller-selecting norms"¹⁹ —agreed-on rules for determining which sources of social control should govern. Most physicians remain deeply committed to the ethic of undivided loyalty to patients,²⁴ yet they are pulled by perceived contractual duties, legal requirements, and ideals of social justice.

People who are squeezed between pressures they cannot reconcile sometimes cope with their dilemma through deceit. The dishonest practices that Wynia et al have described merit moral condemnation. Yet these practices are understandable, at the epidemiological level, as markers of ethical stress that challenges the integrity of the medical profession. Professional leaders, legal and ethics scholars, and public officials should give high priority to the task of defining the scope and limits of clinical advocacy when conflicts arise between individual patients' claims and stewardship of collective resources. Perfect alignment of professional norms, legal requirements, and contract-based expectations is beyond reach. Tension among multiple sources of social control is a fact of professional life in a pluralistic society. But this tension can be reduced to a level that does not undermine professional integrity and public confidence.

I have argued for a robust conception of fidelity to patients, rooted in the moral urgency of standing by people when they are most vulnerable.²² This argument, akin to Fried's case for lawyers' commitment to their clients,⁵ supports a duty to champion patients' interests to the limit of what is possible without making false statements or breaching contractual duties that can reasonably be taken as consented to by patients. This duty encompasses advocacy-oriented presentation of clinical data, including strategic emphasis on facts most favorable to a patient's case. It excludes selective withholding of clinical information that a reasonable reviewer would find material. It does not obligate physicians to follow coverage rules that are not clearly part of health plans' contracts with subscribers. But it does require physicians to honor coverage exclusions and cost-benefit trade-off principles that are explicit in these contracts.

Professional advocacy along these lines would discourage health plans from "gaming" patients by promising all "medically necessary" care, then economizing sub rosa through coverage policies not made plain in subscriber contracts. It would encourage plans to be up-front with prospective subscribers about coverage practices and exclusions. Such candor might, in turn, spur Americans to come to terms with the contradiction between their expectations of reasonable cost and uncompromising intensity and quality of care.