AMA Journal of Ethics^®

Ethicists have long endorsed the full disclosure of medical errors to patients.^1,2,3 In addition, recent hospital accreditation standards and some state laws require that patients be informed about "unanticipated outcomes" in their care.⁴ Nevertheless, the available evidence suggests that full disclosure of medical errors may be uncommon.^5,6,7

This gap between disclosure recommendations and actual clinical practice has many causes. Barriers to disclosing errors range from physicians' concern that disclosure could precipitate a malpractice suit to discomfort in conducting the conversation itself.^8,9,10 The disclosure gap also reveals important unanswered questions about how errors should be disclosed. One such question is exactly what information should be disclosed to a patient following a harmful medical error. While patients and physicians largely agree that harmful errors should be disclosed, they differ regarding the optimal content of disclosure.¹¹ Although physicians appear willing to disclose that an adverse event occurred, they often hesitate to state explicitly that the adverse event was due to an error or to discuss what caused the error and how recurrences will be prevented.

Developing a consensus on the basic content of disclosure could help physicians better inform patients about medical errors. However, promulgating a minimum standard for error disclosure can seem artificial. Patients clearly vary in their preferences for receiving health information,^12,13 so physicians may wish to customize their error disclosure conversations. Yet allowing physicians complete discretion regarding the content of error disclosure is potentially problematic. Physicians are generally unable to predict patient preferences for information in other clinical situations, such as end-of-life decision making.¹⁴ There is no reason to presume their ability to predict a patient's preference for content of error disclosure conversation would be any better. Some physicians report that they typically disclose only very basic information about medical errors and assume that interested patients will ask clarifying questions.¹¹ Relying on the patient to extract desired information about an error from the physician may actually be counter-productive, creating the impression that the physician is trying to hide important details about what happened.

The effort to define a minimum content of error disclosure has a historical precedent in the evolution of informed consent.¹⁵ Informed consent is a formalized approach to sharing information between doctor and patient. Through scholarly debate and court cases, a basic standard for informed consent has emerged, and guidelines for informed consent in specific clinical situations have been delineated. For example, there is a general consensus that patients should be told about the serious risks of a procedure even when the probability of those risks is extremely low. Some purported exceptions to the need to obtain informed consent, such as the therapeutic privilege, have been analyzed and rejected or severely restricted.

Absent an existing consensus on the content of disclosure, there are various approaches for establishing it. First, one could assert that the minimally acceptable content for error disclosure is the information most patients say they would want. In prior research, patients, many who had experienced a medical error in their care or the care of a family member, reported wanting the physician to state explicitly that an error had occurred and to provide basic information about what the error was, why it had happened, and how recurrences would be prevented, and they wanted an apology.^11,16,17 This research has important limitations, however. The studies involved primarily individuals who are not acutely ill and whose preferences for information about errors may differ from patients who have recently experienced a medical error.

Another approach to defining the content of error disclosure would be to consider different ethical rationales for disclosure. However, it is unclear if this approach would gain consensus. To some ethicists, the justification for disclosure is that it promotes patient autonomy and enhances decision making. From the perspective of enhancing decision making, clinicians are required to disclose only error information that patients need to make subsequent health care decisions. The patient autonomy standard of disclosure has limitations also. Consider a fatal medical error. Disclosing this error could certainly not enhance that patient's decision making, and some might argue that this information would not be helpful to the patient's family either.

As a third alternative, one could ground error disclosure in physicians' ethical obligation to tell the truth. Error disclosure as a form of truth telling implies that all errors should be disclosed to patients, including those errors that cause no harm. Theories of justice would direct disclosure of information patients need in order to know whether they are being fairly compensated for a harmful error. Finally, utilitarians might suggest that the content of disclosure be oriented toward producing the greatest net benefit. Suppose disclosing a harmful error was certain to precipitate a lawsuit. Are the harmful consequences of disclosure for health care workers relevant in deciding what to disclose about an error, and, if so, how should this harm be weighed against the benefit of disclosure to the patient?

One aspect of error disclosure is especially controversial, namely whether physicians should apologize. Patients uniformly desire an apology following a harmful error; an apology demonstrates that the involved health care workers and institutions recognize the magnitude of what has happened to the patient. While physicians want to apologize, they fear that an apology could be construed as an admission of legal liability.

Many states have adopted or are considering "apology laws" that exempt expressions of regret from being considered as admissions of liability.¹⁸ But it is unclear how much legal protection such apology laws actually provide. While protecting apologies, many laws also note that statements concerning "culpable conduct" are still admissible. The California statute reads:

The portion of statements, writings, or benevolent gestures expressing sympathy or a general sense of benevolence relating to the pain, suffering, or death of a person involved in an accident... shall be inadmissible as evidence of liability in a civil action. A statement of fault, however, which is part of, or in addition to, any of the above, shall not be inadmissible pursuant to this section.¹⁹

Whether the physician's apology constitutes a "general sense of benevolence" or a "statement of fault" is then debated in court. From the patient's perspective, what likely matters most is whether a health care worker's apology seems sincere. A legally approved apology delivered without empathy could actually backfire and increase the chances that a patient would sue.

Ultimately, consensus regarding a minimum standard for error disclosure will require continued debate about the relevant ethical norms, additional empirical research about patient's preferences for disclosure, and clarification of the legal status of disclosure statements. In the meantime, the following guidelines can help clinicians approach these difficult situations:

Let patients' preferences guide error disclosure.

Patients desire an explicit statement that a harmful error occurred, information about what the error was, why the error happened, how recurrences will be prevented, and an apology. Physicians should provide this basic information about harmful errors to patients regardless of whether the patient asks.

Get help.

Disclosing errors is difficult and requires careful planning and consultation with risk managers or other knowledgeable colleagues. Oftentimes, whether an event actually was an error becomes clear only after thorough analysis.

Increase institutional support for error disclosure.

Error disclosure, like any other communication skill, requires training and practice. Institutions should provide such training and should insure that their policies support clinicians in disclosing errors to patients. Institutions should also increase their emotional support for health care workers affected by errors, thereby allowing clinicians to better meet the needs of the affected patient.