Special Focus |
|
|
|---|
Key Words: physicians' ethical obligations patient confidentiality public health violence prevention individual rights
On April 16, 2007 Virginia Polytechnic Institute and State University (Virginia Tech) student Seung-Hui Cho killed 32 people and wounded more than 20 others before committing suicide in the deadliest campus shooting in US history. In response to this tragedy, university administrators and violence prevention experts across the country are examining the details to prevent future incidents. At the center of discussions is Cho's history of mental health problems. According to court papers, Cho was found to be "mentally ill and in need of hospitalization" in December 2005. Involuntary outpatient mental health treatment was subsequently mandated, but it remains unclear whether Cho completed treatment or received any specific diagnosis.1 Questions have emerged regarding the extent of the university administration's knowledge of Cho's mental health status and violent tendencies and whether this information, if available, should have prompted any action to protect students and faculty. Controversy also surrounds Cho's purchase of 2 semiautomatic pistols despite federal laws designed to prevent those judged mentally ill by the courts from purchasing handguns.
Several days after the shootings, the Independent Virginia Tech Incident Review Panel was convened by Virginia governor Timothy M. Kaine to investigate Cho's background, including his medical history; the events and circumstances surrounding the shootings; and the university's response with the purpose of understanding causes and preventing future incidents.2 The panel's focus is on warning signs as well as administrative barriers, policies, or laws that may have kept school administrators from knowing critical information about Cho that could have prevented the shootings. Kaine said he would press lawmakers for reforms "if the panel found that state or federal privacy laws covering medical records need to be changed to give school administrators more information about troubled students."3 Similarly, in a statement to the panel, Virginia Tech president Charles W. Steger said, "We need to know if privacy laws can or should change so that school administrators, court officials, or the mental health profession itself has the information it needs to treat and handle those with mental illnesses on college campuses."4
Presently, individual medical records, including mental health records, are legally protected and cannot be released to third parties without patient permission except in limited circumstances, as provided in federal and state privacy laws. These privacy protections apply even to deceased patients. Therefore, the panel's access to Cho's medical records is limited, although Cho's family may choose to release any information to which it has access. Kaine's and Steger's comments draw attention to the inherent tension between an individual's right to have his or her medical information remain confidential and public duties to promote the health and safety of third parties. Elected officials and university presidents have responsibilities to communities and must balance the protection of individual rights with potential risks and benefits to the aggregate. Physicians' primary ethical obligations are to the health and well-being of individual patients. Strict confidentiality is a traditional professional ethical norm of medicine; however, physicians and other health care providers have obligations to the public health, in part because they are in a unique position to identify potential threats (eg, individuals with communicable diseases or intentions to commit violent acts).5 Arguably, breach of patient confidentiality may be ethically justifiable, or even obligatory, if that breach can be reasonably expected to prevent harm to others.
This article describes the public health approach to violence prevention, including the role of individual health care providers. Central to this discussion is the examination of ethical arguments that favor or oppose the disclosure of confidential medical information for the purposes of public protection. In addition, this analysis also discusses laws that permit or require health care providers to breach confidentiality to protect third parties. Following these analyses, the authors summarize additional considerations relevant to public policies regarding the disclosure of personal medical information for public health purposes.
|
|
|---|
From an ethical standpoint, the public health approach is founded upon utilitarian grounds. Under this rubric actions are preferable insofar as they promote the greatest good for the greatest number. In this case, infringements upon the rights of individuals can be justified if they can produce more substantial benefits (eg, prevention of potential harms) to the larger population. Public health surveillance and communicable disease reporting systems, long-accepted public health interventions, are justified on these grounds.
Like communicable disease reporting, a public health approach to violence prevention requires substantial cooperation from physicians and other health care providers working in clinical settings. As discussed below, health care providers in most states are encouraged or required to breach confidentiality when a patient makes a violent threat; however, unlike cases of communicable disease that are reported to a central public health agency, threats of violence are to be reported to the potential victim and/or local law enforcement. Accordingly, health care providers, particularly mental health professionals, are uniquely positioned to assist in identification and prevention efforts because some substantial root causes of violence derive from medical or psychological antecedents.9 These precursors may include addiction, substance abuse, low intelligence, or mental disorders such as depression, hyperactivity, and attention-deficit/hyperactivity disorder.10 Physicians possess the specialized knowledge that is necessary to identify these risk factors at early stages. Moreover, the patient–provider relationship creates the opportunity to speak candidly with individuals at risk and assess any predispositions to violence; this candor is arguably necessary to facilitate effective treatment.5
In general, the benefit of preventing additional harms to the public, which derives from the reporting of known incidences of communicable disease, is considered to outweigh the harms associated with the disclosure of patients' personal health information. This measurable benefit makes such disclosure appear more acceptable to physicians who are called upon to deviate from their standard obligations to protect patients' privacy.
There is some debate, however, whether the same rationale for disclosure is applicable to public health efforts to prevent violence. Violence is inherently different from other threats that public health authorities have sought to address. Foremost, violence does not spread like communicable diseases, nor are its causes as easy to isolate. Moreover, the potential harms of having one's mental health status disclosed or of being labeled as violent are somewhat different than identifying someone as having SARS or measles (sexually transmitted diseases are more sensitive, especially HIV). Physicians' cooperation with the legal and law enforcement systems in preventing violence therefore raises unique ethical issues that must be examined in greater detail.
|
|
|---|
"The information disclosed to a physician by a patient should be held in confidence. The patient should feel free to make a full disclosure of information to the physician in order that the physician may most effectively provide needed services. The patient should be able to make this disclosure with the knowledge that the physician will respect the confidential nature of the communication ..."11
The American Psychiatric Association has adopted the AMA Code of Medical Ethics and added "With Annotations Especially Applicable to Psychiatry"12; the Ethical Principles of Psychologists and Code of Conduct of the American Psychological Association is similar to the AMA code with regard to privacy and disclosure.13
Patient confidentiality is rooted in the more general ethical norm that physicians respect patients' autonomy, that is, patients' control over the influences that determine their actions.5 Confidentiality rules are also based on the professional ethical norm of nonmaleficence. Physicians have an obligation to promote the well-being of their patients. If physicians lose their patients' trust, then patients may not be completely forthcoming in providing personal information during the clinical encounter. In turn, this loss of trust can diminish physicians' ability to diagnose and treat their patients.
This consequentialist argument outlines the importance of confidentiality from the perspective of both patient and physicians. Such reasoning still allows for exceptions to absolute confidentiality when necessary to benefit the individual patient, but does not effectively establish the propriety of disclosing confidential information to benefit third parties, or even the community as a whole. More dubiously, this logic does not establish criteria to guide physicians in distinguishing between circumstances in which a breach in confidentially may be permissible and those circumstances in which the disclosure of patient information may be required.
Despite increasing emphasis in the United States on the privacy of personal medical records, contemporary medical ethics does allow for exceptions to absolute patient confidentiality, specifically when potential harm to third parties can be prevented. (There are ethical arguments against "qualified" medical confidentiality. In support of absolute confidentiality, Kenneth Kipnis argues that exceptions to confidentiality place the moral comfort of physicians above the well-being of patients and victims, and this conflicts with professional responsibility. In Kipnis's opinion, absolute confidentiality is the only way to maintain honor and respect for the profession of medicine, even if it means that physicians are aware of but do not intervene to prevent likely harm to third parties. Kipnis also claims that absolute confidentiality "is effective at getting more patients into therapeutic alliances more quickly, it is more effective in bringing about better outcomes for more of them and—counter-intuitively—it is most likely to prevent serious harm to the largest number of at-risk third parties."14) In addition to patient-centered ethical duties, physicians arguably concurrently possess a duty to prevent violence that is logically derived from physicians' ethical obligations to promote the well-being of the public. To this end, the AMA Code of Medical Ethics states that confidential patient information may be disclosed "(w)hen a patient threatens to inflict serious physical harm to another person or to him or herself and there is a reasonable probability that the patient may carry out the threat."11 In cases in which disclosure of confidential medical information is required by state or federal law or court order, the Code additionally states that physicians should notify the patient and disclose minimal information required by law11; however, professional ethics involves clinical judgment and the careful balance of competing norms. Just because breach of medical confidentiality may be ethically justifiable in a specific situation, it does not necessarily follow that laws requiring breach of confidentiality in specific situations will translate directly into preventing violence.
|
|
|---|
The California Supreme Court determined that the therapist's actions were insufficient because he informed only the police and not the victim herself or her family directly. A concise statement by the court for this new disclosure requirement provides that "the protective privilege ends where the public peril begins," reflecting the view that at times the safety of an individual (or individuals) outside the doctor–patient relationship may be more important than patient confidentiality.15 In this vein the court's analysis parallels the ethical basis for communicable disease surveillance and reporting.
The court cited the American Medical Association's Principles of Medical Ethics (1972): "a physician may not reveal the confidence entrusted to him in the course of medical assistance unless he is required to do so by law or unless it becomes necessary in order to protect the welfare of the individual or of the community."16 The court viewed this principle as proof that professional ethical standards already permitted such disclosures; however, if this principle is read in conjunction with the Opinion of the Judicial Council regarding "Patience, Delicacy and Secrecy," which requires secrecy except in instances when a communicable disease is diagnosed—in which case the physician may disclose the information—or when the law requires disclosure, the court's interpretation is not accurate.16 The AMA's policy did not consider breaching confidentiality for threats of violence, unless the state's laws already required it (which California's apparently did not). Interestingly, the decision of the court made such disclosures ethically permissible by making disclosure the law.
The Tarasoff court's decision leaves us with ambiguity when faced with large-scale events. (Cho was deemed a danger to himself, not others. The obligation to warn delineated in Tarasoff is not analogous to suicide, and there is less agreement between jurisdictions whether a health professional can be held liable for a patient's suicide. Typically, if the patient is in the custodial care of a physician or facility, where the patient's activities can be controlled and monitored, liability may be appropriate; however, if the patient is receiving treatment on an outpatient basis, it is less likely that treating mental health professional will be held responsible for the patient's actions.17) The rigid disclosure requirements apply only to identified individuals, so what happens if a patient makes valid but generalized threats to an identifiable yet amorphous group of people; an organization, business, or university; or, more generally, the public? States that have, through legislation, required disclosure generally apply the requirement to identifiable victims.18 Other states, however, have taken a less rigid approach and applied it to a broader class of victims. For example, Florida's law permits confidentiality to be waived "when there is a clear and immediate probability of physical harm to the patient or the client, to other individuals, or to society."19 Creating a legal basis for permitting disclosure in a wider range of circumstances may in the end be more beneficial than requiring disclosure in the limited circumstance delineated in Tarasoff.
Although the law may mandate or permit disclosure of sensitive medical information to prevent violence, physicians' professional ethical obligations of beneficence and nonmaleficence require them to consider patients' best interests ahead of public interests. To this extent, physicians must recognize that the disclosure of a patient's predisposition to violence does not mandate the disclosure of their entire medical record (eg, specific diagnoses, treatments, or facts revealed in therapy sessions), and the law does not generally specify the information to be disclosed. Health care providers should disclose only the minimum information to meet the requirements of the law to prevent an anticipated act of violence.11
|
|
|---|
Dr Anderson, Mr Black, and Mr Bostick are with the American Medical Association.
Correspondence and reprint requests to Dr Emily Anderson, American Medical Association, 515 N State St, Chicago, IL 60610 (e-mail: emily.anderson{at}ama-assn.org).
Received for publication June 11, 2007; accepted June 28, 2007.
|
|
|---|
148.975 (2006).
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||