Words That Heal

Document Type


Publication Date



An apology is a complicated thing. Coming clean is hard; if done poorly, it can make things worse. Just ask Bill Clinton or Pete Rose. But when done right, talking can be powerfully productive, as suggested in the article by Mazor and colleagues in this issue (1). As former litigators and now mediators and clinical law teachers, we have spent our careers learning how conflicts arise and what is really at stake when people sue. We have also learned how words can be used to change perceptions and, ultimately, to find constructive and less costly ways to resolve conflicts. Our experience has taught us how nuanced the subject of apology can be. Apologies have a potential for healing that is matched only by the difficulty most people have in offering them. Physicians as a group face special challenges in this realm. Like the rest of us, physicians are subject to shame, guilt, and grief when their actions harm others, even if fault is unclear. These feelings, coupled with the fear of litigation that permeates medicine, may get in the way of effective communication with patients and their families after a medical error. In addition, since statements of sympathy, regret, or responsibility can be used to prove legal liability in medical malpractice cases in many states, lawyers and risk managers routinely advise physicians against being open about what happened and against apologizing, especially at the time when such communications might be most valued by a patient. These personal and institutional impediments to openness are at odds with what research is beginning to show us: that ineffective communication is the single largest factor in producing patient litigation (2) and that good communication, including effective apologies, can avert or help end conflict, especially litigation. Moreover, laws in some states mandate doctors to come clean after serious adverse events (3-5). In light of this tension and the emerging legal requirement of disclosure, this study by Mazor and colleagues is both welcome and timely (1). In the study, questionnaire respondents (all managed care patients) were presented with 2 hypothetical cases of undisputed medical error (failure to check for penicillin allergy and failure to monitor antiepileptic drugs) with varying degrees of clinical outcome (life-threatening and less serious) to the patient. Physician fault was clear. Respondents were given hypothetical doctor–patient dialogues that were randomized for these factors and 2 levels of physician disclosure: “full disclosure”—acceptance of responsibility, a clear description of the error, and an apology with a promise of reform to prevent recurrent error—and a “low” or partial level of disclosure—an expression of regret without acceptance of responsibility or an apology or accurate information about what went wrong. Participants were then asked to provide their reactions. Overall, Mazor and colleagues found that respondents are more likely to respond favorably—more trusting, more satisfied, and less likely to change physicians—when physicians fully disclosed medical errors. The strongest of these responses were in cases of more severe patient harm. Physician openness, however, did not seem to diminish the inclination to seek legal advice about the error. On one level, these results confirm what we have generally learned through our mediation work with litigants, what scholars in law are beginning to learn, and, indeed, what one might simply suspect on the basis of human experience. It is hardly surprising that patients want and respond well to full disclosure and an apology in which the physician takes responsibility for what happened in cases of clear error and clear fault. “Partial apology”—limited to an expression of sympathy without acknowledging fault—lacks the moral dimension of its fuller counterpart and, even worse, may seem insincere or evasive. Indeed, research in the context of civil litigation suggests that partial apologies may, where fault is clear, be worse than none at all (6). The positive response to full disclosure may be explained by another of the study's findings: Patients are somewhat torn and of two minds—on the one hand, they feel that physicians are only human and thus will make errors, but on the other hand, they think they have a right to be free from physician mistakes. Full disclosure may tip the balance to give the physician the benefit of the doubt. One of the study's findings may seem a bit anomalous at first glance: Full disclosure and apology did not have the same positive effect on patients with regard to seeking legal advice about the medical error, at least in the case of serious harm to the patient. The researchers themselves seemed surprised at this finding, but to us, it was not at all unexpected. Patients seeking advice from a lawyer about the possibility of obtaining compensation after a harmful medical error should be no more surprising than patients seeking the services of a physician if they were hurt in a road or workplace accident. Both professionals can help remedy the injury. Beyond this, it is significant that none of the study vignettes included an offer of a financial nature (for example, money or relief from past or future bills for treatment) in the discussions. While apologies by definition deal with the intangible aspects of injury, they do not eliminate the need to address the tangible aspects as well. Moreover, not every legal consultation produces litigation. Indeed, given the cost of prosecuting a lawsuit on the basis of medical error, the likelihood of an actual claim being brought in a minor harm scenario is remote. Finally, even where a patient–victim sues after receiving full disclosure and an apology, the apology may nonetheless have a positive and significant (albeit delayed) effect. As shown in the case of Veterans Affairs hospitals (7) and other industries (as recently exhibited by such celebrated apologizers as the Ford Motor Company [8]), those who file lawsuits are inclined to accept reasonable compensation more quickly if they perceive the wrongdoer as having a heart and taking responsibility. The study by Mazor and colleagues points to the need for more work of this kind. This study dealt with doctor–patient communication in the relatively “easy” area of undisputable professional negligence causing clear harm. But many, if not most, claims of patient harm take place under cloudy circumstances. What if a nonpreventable adverse event occurs? What about cases where a physician makes a mistake, but her or his conduct is not negligent (that is, within the acceptable standard of care in the community)? In such cases, physicians will understandably be reluctant to take responsibility. But what kind of communication before and after such events will be most productive? And what about the effects of openness versus guardedness on the physicians in these different scenarios? They too must deal with the emotional consequences of being involved in an adverse event, both while disclosing the event to the patient and in the longer term. This study also points out significant policy choices on several fronts. Physicians and related professionals must learn the kind of communication skills that are often in short supply in medical training and practice and be rewarded when they use them skillfully. The study may also shed light on 2 important law reform questions: whether we should seek to encourage more open communication by doing what some states have done—making apologies or statements of regret inadmissible in court (9-12)—and whether more states should follow the lead of Pennsylvania, Florida, and Nevada by mandating disclosure of serious adverse events. On the level of institutional and risk management, the study should trigger more inquiry about what sorts of conversations between physicians and patients (regardless of whether the conversations receive evidentiary protection) should occur. Good discussion after an error might produce more positive patient attributions (13) and thus fewer law suits. More important, if such conversations promote physician reflection and institutional learning (especially if they provide patients and families with an opportunity to add their insights), they may lead to increased patient safety. Finally, this study indicates the potential value to be gained if lawyers, physicians, and patients talk together about these difficult issues. While discussion may not solve the “malpractice crisis,” it may shed some light on the enormous potential power of a few well-chosen, well-delivered, and well-timed words.


Adverse events, disclosure, emotions, patients, penicillin, perception, questionnaires, risk management, safety, serious adverse events

Publication Title

Annals of Internal Medicine