Article Text

  1. M. B. Duke1,
  2. J. F. Wilson1
  1. 1University of Kentucky, Lexington, KY


Background When a medical error occurs, studies indicate that patients prefer to be informed, be given an explanation of what happened and why, be informed of the measures that will be taken to prevent recurrences, and offered an apology from the responsible parties. Most patients endorse a preference for full disclosure, even when no harm is done. Data suggest that as few as one-third of harmful errors are disclosed to patients and that although physicians are supportive in principle, they are hesitant to actually disclose error. Resident physicians, by virtue of their inexperience, may be involved in medical errors at a rate higher than more experienced physicians, yet their responses to error, including their attitudes toward disclosure, have not been well studied. We sought to examine resident attitudes toward disclosure by asking them to predict their actions after being involved in a medical error in the hospital setting.

Methods We anonymously surveyed 44 internal medicine residents attending a resident retreat. Two clinical cases were presented. In the first case, a diabetic patient is given insulin and kept NPO with resultant hypoglycemia. In the second case, a misread chest radiograph delays the diagnosis of thoracic aortic aneurysm. For each case, the outcomes ranged from no harm to serious harm. The resident was given four possible responses: (1) no explanation, (2) an explanation of the medical facts of the case, (3) an explanation and an admission of error, and (4) an explanation, admission of error, and an apology. The residents were directed to choose the response that most closely resembled what they would say and do.

Results Surveys were completed by 34 (77%) residents, including 14 (34%) women, with all levels of training equally represented. In every instance, the most likely response was to explain (54%) rather than to admit error (22%) or to apologize (22%). In the case of the diabetic patient with hypoglycemia, responses did not differ regardless of whether the error caused harm. In the case of the misread chest radiograph, however, if the patient was harmed, residents were significantly more likely to include an admission of error and an apology than if no harm was done. Neither gender nor level of training affected results.

Conclusions Our findings suggest that, with respect to disclosure of medical error, the actions of residents are at odds with the stated preferences of patients. When presented with hypothetical cases of medical error, residents were significantly more likely to limit their response to an explanation of medical facts rather than an admission of error or an apology. Results also suggest that the type of error may influence the response. Although both cases involved error, the first case may be viewed as a “system” error in which responsibility is shared by other members of the health care team. In the case of the thoracic aortic aneurysm, the error more clearly rests with the physician, and this sense of personal responsibility for the error may be influencing the response toward full disclosure when the patient suffered greater harm. Factors influencing the response to error by residents should be considered as educational interventions are developed to bridge the gap between patient preferences for disclosure and physician practice.

Statistics from

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.