When Liability and Ethics Diverge

Civil liability in medical malpractice cases serves two distinct purposes: to provide compensation to injured patients or render them “whole” after negative clinical outcomes and to discourage undesirable conduct on the part of practitioners. Restitution might be achieved in a more cost-effective manner through the implementation of a universal, no-fault insurance system analogous to the workers’ compensation system used for on-the-job injuries. However, in the U.S. health care system, concern for civil liability plays a crucial role in shaping the behavior of physicians. At the extreme, providers whose past conduct renders them incapable of qualifying for affordable insurance coverage will find themselves unable to gain hospital privileges or unable to practice in most settings. As losing one’s medical license for ineptitude or general incompetence, as opposed to intentional misconduct, proves a relatively rare occurrence in many jurisdictions, an inability to procure liability coverage at a reasonable rate can serve to drive the least skilled providers from the profession. Like the adage goes about Hollywood, one stops working in health care when one becomes uninsurable (1).

Legal rules are generally designed to reflect and further ethical standards. William H. Shaw, a leading authority on business ethics, has observed that “[t]o a significant extent, law codifies a society's customs, ideals, norms, and moral values” and “changes in the law tend to reflect changes in what a society takes to be right and wrong” (2). In American medicine, civil liability (i.e., court-ordered payment of damages for harming another party) is among the mechanisms used to foster ethical conduct. Ideally, it is imposed—either through common law, which derives from the decisions of judges, or legislation—in circumstances where doing so reflects society’s underlying ethical values and where the aim is to further behavior by providers that conforms to those norms. In fact, the implicit assumption underlying this liability regime is that, if it reflects society’s ideals, it will inevitably further ethical ends. This principle is generally true. Even laypeople understand that tort law—the set of legal rules that addresses compensation for personal injuries—deters both physician misconduct (e.g., breaches of confidentiality) and second-rate care. Of course, much has been written about the fear of malpractice claims leading to defensive medicine and excess medical costs—at least, in the United States, which is, by far, the most litigious nation in the world when it comes to health care (3). It is not at all clear, however, that such risk-averse care truly runs against societal values.

Rarer and more concerning are situations in which a liability rule, or an approach to liability insurance that reflects one ethical norm, results in a policy outcome that achieves precisely the opposite effect. Three examples of this phenomenon are outlined in the following text: the imposition of liability for outpatient suicide, the limitation of malpractice coverage for certain forms of sexual misconduct, and the broadening of Tarasoff-like duties to include warnings to nonidentified potential third-party victims. Each of these examples reveals a distinct way in which imposing or adjusting liability to achieve an ethical ideal may backfire and risks producing results that are incompatible with that ideal.

Outpatient Suicide Liability

Suicide is the 10th leading cause of death in the United States and one of the most feared outcomes in psychiatric practice. Rates have risen precipitously in recent years, leading to increased awareness and various public health interventions (4). However, liability for outpatient suicides is a relatively new phenomenon in psychiatry. Before the 1960s, no outpatient suicide liability claims are known to have reached U.S. appellate courts (5). The California courts first considered the possibility of liability for outpatient suicides in Bellah v. Greenson (1978), followed by New York in Bell v. New York City Health and Hospitals Corporation (1982) (6). Once permitted, the number of such claims grew rapidly. By the start of the 21st century, suicide liability was the leading cause of financial judgments in psychiatric malpractice (7). One can assume that the physicians and insurers responsible for paying these claims were displeased with this result. However, the crucial question—seldom asked—is whether the rise of outpatient suicide liability achieved the policy goals for which it was designed.

The purpose of holding outpatient providers accountable for failure to take reasonable measures to prevent outpatient suicide is twofold. On the one hand, doing so is intended to increase the care that psychiatrists render in making risk assessments, presumably leading to more caution, increased hospitalization rates, and fewer preventable suicides. The second purpose is to compensate surviving relatives for their losses after such deaths. These goals reflect the widely held ethical norm that suicide is undesirable—at least when motivated by mental illness—and should be prevented whenever possible. But does suicide liability reduce suicide? Probably not.

A generation of experience suggests that the imposition of liability for outpatient suicide, however well intentioned, has achieved neither of its aims. Suicide is extremely difficult to predict with existing tools. To protect themselves from liability, psychiatrists have to engage in overinclusive approaches to hospitalization: denying freedom to 99 patients at risk who will not kill themselves, so to speak, to protect the one patient who will do so. In the short term, this may protect that particular individual. However, the negative consequences for the other 99 patients should not be underestimated. How many of them will avoid necessary mental health care in the future? How many will conceal genuine suicidal intent or plans—however inchoate—when they arise again, to prevent another unwanted hospitalization? Overinclusive involuntary hospitalizations may save some patients’ lives in the short run, and will likely shield providers from some malpractice claims, but the result may lead to patients avoiding the care they need in the long run—possibly even raising suicide rates.

A second consequence of outpatient suicide liability has been that—at least in large urban areas where demand usually exceeds supply—psychiatrists screen out potentially suicidal patients from their practices. In New York City, for example, patients who are most in want of psychiatric care because of high suicide risk are precisely those who are least likely to be able to find private providers to meet their needs (8). This systemic imbalance is a logical outgrowth of the outpatient suicide liability regime. If there is a sizable pool of potential patients all paying the same rate and a provider has a significantly higher risk of a malpractice claim for treating some individuals rather than others, rational self-interest argues for cherry picking the patients at lower risk. The result is an artificial shortage of care.

One might justify outpatient suicide liability on the grounds that it achieves its second purported goal: compensating survivors. However, this result appears to be grim consolation. What family member would not give up the prospect of future compensation—no matter how large or how certain—to increase the odds of their loved one receiving high-quality care while alive? Similarly, would not most family members accept the increased short-term risk of not hospitalizing certain potentially suicidal patients in order to prevent the overinclusive commitment of many others if doing so could reduce the long-term overall risk to their kin? Of course, once a suicide has occurred, families and estates want their due. However, viewed prospectively, nearly all would prefer better access to care and better long-term outcomes over compensation after a loss.

Sexual Misconduct

Sexual relationships between psychiatrists and patients have been proscribed by the American Psychiatric Association since 1973; the ban was expanded in 1992 to include former patients (9, 10). Such conduct was widespread in the 1970s, with estimates ranging from 5% to 10% of psychiatrists engaged in such endeavors and volumes such as psychiatrist Martin Shepard’s The Love Treatment: Sexual Intimacy Between Patients and Psychotherapists claiming that a “sexual involvement can indeed be a useful part of the psychotherapeutic process” (11, 12). Such conduct retreated from public view in the wake of the Phyllis Chesler’s groundbreaking New York Magazine exposé, “The Sensuous Psychiatrists,” which questioned the practice, and the $350,000 verdict in Roy v. Hartogs (1975) (13). Such sexual relationships are now categorically unacceptable, yet how they should be addressed by the tort system remains a question of considerable controversy and one that reveals tension between ethics and policy.

One often overlooked aspect of the decision in Roy v. Hartogs is that Renatus Hartogs, the physician who had seduced his patient, sued his insurance provider for indemnification (i.e., liability protection)—and lost. New York’s appellate court found that, “as a matter of public policy,” such indemnification would be “immoral” and “pay the expenses of prurience” (14). In a number of subsequent cases, insurers argued that such intentional sexual misconduct could not be malpractice because it was not psychiatric practice, per se, and therefore fell beyond the scope of insurance coverage (15). Although early outcomes were mixed, eventually some courts relied on theories of “mismanagement of transference” to justify payouts when psychiatrists—as opposed to other medical providers—were concerned (15). In rendering such judgments, courts prioritized compensation to victims over any potential moral hazard created by indemnifying egregious behavior.

The result was not widespread coverage. Instead, the American Psychiatric Association, which at the time provided insurance to a large portion of psychiatrists in private practice, overtly dropped such coverage from its policies in 1984. The “leadership of the APA looked upon sexual misconduct with patients as a most serious form of ethical violation” and thought it “inappropriate for an APA-sponsored program to defend and indemnify members who had engaged in such serious willful misconduct” (16). (Of note, concerns for high premiums and pressure from secondary insurers may also have played a role in this decision [17].) Although the APA-approved plan was willing to provide limited funding to defend such cases, the organization refused to cover verdicts against offenders—or, as a result, compensate plaintiffs. At the time, a committee composed of many of the most prominent women in the APA strongly dissented from the organization’s decision, arguing that the new policy offered little recourse to victims of abuse (15). At present, some liability carriers “explicitly exclude coverage for sexual misconduct,” often using a “Reservation of Rights” letter, whereas others cap such payouts with a limit such as $25,000 (18).

Unlike with the imposition of outpatient suicide liability, the changes to sexual misconduct liability did not occur through the legal system directly but through private insurers, yet the outcomes reveal the same underlying tension between ethics and policy. Taking the APA leaders at their word, the organization’s goal was to reflect a widely held ethical norm against sexual misconduct and, in doing so, to deter unethical behavior. Whether the exclusion they created deterred such behavior is hard to determine, especially in the light of other social changes that rendered psychiatric–sexual relations less frequent, but the answer probably is no. What provider believes that he (or she) will get caught engaging in sexual relations with a patient? Moreover, who would ever think, “If I get caught, I’ll lose my reputation and my license and my livelihood, but at least I’ll be indemnified by my liability carrier”? In short, it seems highly unlikely that the exclusion of insurance coverage for sexual misconduct led to a reduction in such misconduct.

However, an unintended consequence of these exclusions may be that victims go uncompensated or undercompensated. Many defendants simply do not possess the personal resources to cover large payouts; others have shielded their assets through complex legal machinations. Not only does this mean less compensation for victims but, arguably, less incentive for patients—and their attorneys—to bring forth their legal claims in the first place. What psychiatric patient, already the victim of sexual misconduct, is going to endure the challenges of litigation without the prospect of receiving meaningful compensation if he or she wins? Also, as a practical matter, what attorney is likely to handle such a matter? (It is assuredly not the case that this approach to liability is the only reason why victims of sexual misconduct do not come forward, as patients may experience shame, fear reprisal, or want to avoid attacks upon their credibility; however, in embracing this liability regime, the psychiatric profession and insurers have created yet another disincentive for taking action.)

In sticking to what they perceived to be the moral high ground, insurers undermined the widely held ethical norm that victims of mistreatment should be compensated. It is even possible that they deterred the very forms of litigation that would prove likely to expose, and thereby reduce, such misconduct.

Expanded Duty to Warn

Psychiatrists in many jurisdictions often have a duty either to warn or to protect identifiable third parties when they are threatened with physical violence by patients. This principle was originally established in a pair of California cases known as Tarasoff and is now imposed through a state-by-state patchwork of judicial rulings and statutes. A model statute proposed by the APA and roughly followed in a number of jurisdictions states that: “no cause of action shall lie against a [physician], nor shall legal liability be imposed, for breaching a duty to prevent harm to person or property caused by a patient unless a) the patient has communicated to the [physician] an explicit threat to kill or seriously injure a clearly identified or reasonably identifiable victim or victims, or to destroy property under circumstances likely to lead to serious personal injury or death, and the patient has the apparent intent and ability to carry out the threat; and b) the [physician] fails to take such reasonable precautions to prevent the threatened harm as would be taken by a reasonably prudent [physician] under the same circumstances” (19).

The key elements here are that the victim be readily identifiable and that the danger be overt. These carefully tailored limitations balance the need to protect third parties against other essential values, including confidentiality, patient welfare, and the dangers of excessive entanglement of medical practice with law enforcement.

The Washington State Supreme Court radically expanded the duty to warn in the matter of Volk v. DeMeerleer (2016) (20). In that tragic case, a patient with bipolar disorder murdered his ex-fiancée and her son, leading the ex-fiancée’s surviving son and her mother to sue the patient’s outpatient treating psychiatrist for damages. According to the facts presented at a motion for summary judgment, the patient had never voiced any particular threats toward his ultimate victims, nor had the patient more broadly threatened to harm himself or others. Nonetheless, an appellate court ruled against the physician, and the state’s highest court concurred, finding that a psychiatrist “incurs a duty to take reasonable precautions to protect anyone [italics in original] who might foreseeably be endangered by the patient’s condition” (21). No longer does the threat need to be overt, nor does it need to specify a target. The goals of such a ruling were presumably to encourage greater care in taking measures to protect the public while ensuring a larger swath of third-party victims are eligible to receive compensation—but will the new standard achieve either goal?

Being able to guarantee confidentiality to a psychiatric patient in all but a handful of extreme circumstances is essential for the development of a meaningful therapeutic alliance. Many providers explicitly warn their patients in advance that everything they share is confidential, unless they pose a direct threat to themselves or others. To most patients—and, likely, to most nonpatients—this caveat is reasonably interpreted to mean that the exception applies when the patient takes concrete steps (either through words or actions) to indicate the intent to harm a third party. That alone is a broad hole in the fabric of medical confidentiality but one that most patients have learned to tolerate. In contrast, the Washington ruling requires a much broader qualification, something like: If I think you might be a danger at some point to anybody, I may have to breach your confidentiality. Thus, this approach may well deter many patients from candid interactions with their providers, and some patients might forgo therapy entirely.

The consequences of operationalizing such a policy are also drastic. Countless third parties, such as relatives, neighbors, and employers, will be informed of hypothetical dangers. Many of these individuals might not otherwise ever have known the patient to suffer from mental illness. Because these third parties are not bound by the Health Insurance Portability and Accountability Act or any fiduciary duty to the allegedly dangerous party, psychiatric patients are liable to face social consequences in their communities. Rather than preserving the physician–patient relationship so that those truly in danger will be warned, the broader rule risks undermining that relationship and raises the possibility that fewer potential victims will be warned—because fewer potential perpetrators are in care or are being honest with their therapists.

What is crucial to understand is that the ethical norm (third-party welfare) that drove the underlying approach to policy also threatens to undermine its own goals. The difficulty with Washington’s approach is not that it values third-party welfare over confidentiality, which might, indeed, reflect society’s ethical norms. Rather, the problem is that in choosing a policy that prioritizes the former over the latter, the state may have inadvertently rendered protecting the public all the more difficult—which was certainly not its intent.

Conclusions

Liability regimes have short-term and long-term consequences. These consequences often differ. In all three scenarios described herein, the new approach to liability initially mirrored underlying (and often evolving) ethical norms. In rendering the new rules, the legislators, judges, or thought leaders in psychiatry sought just outcomes with specific parties in mind. However, as Justice Holmes reminds us, “hard cases” often “make bad law” (22). Achieving short-term justice can occasionally boomerang in ways that compromise long-term justice, and sometimes, such as in the use of truth and reconciliation processes, long-term justice can best be achieved by sacrificing short-term justice entirely. In all three of the circumstances I have outlined, there is likely an optimal “sweet spot” of policy, where the balance between two competing values leads to the closest approximation of the underlying ethical norm. Some limited degree of outpatient suicide liability may protect those patients in extreme danger and compensate victims of extremely poor care, whereas too much may lead to the overhospitalization of some patients and to limited access for others. Some compensation for victims of sexual misconduct may both help render them “whole” and motivate appropriate litigation, whereas too much may create a moral hazard. Encouraging the protection of unidentified third parties under some circumstances may lead to enhanced public welfare, whereas in other situations, it may undermine physician–patient confidence and pose a detriment to the welfare of those same third parties. To a significant extent, these are empirical questions—but, a priori, there are often little data to guide decision makers or to answer the underlying questions. That does not absolve psychiatrists or lawmakers from considering these tensions: By merely acknowledging that ethics and liability rules can often diverge—frequently in unexpected ways—and by keeping a watchful eye for such divergences, these thought leaders and rule makers can help to right the course of policy when it departs from its ethical underpinnings.