When a medical error occurs, physicians have ethical and professional obligations to inform the patient.1–3 Many countries have established standards of practice or legislation that mandate disclosure.4,5 In Canada, the Canadian Patient Safety Institute has developed guidelines for disclosure.6 These guidelines advise that error disclosure comprise a statement of the facts, information about the ongoing care of the patients, an expression of sympathy or regret, an overview of the investigative process that will occur, an offer of future meetings, time for questions, offers of support, and a follow-up plan. In the United States, the National Quality Forum7 includes error disclosure amongst its Safe Practices. Studies of physicians indicate that the majority agree that patients should be told about errors in their care.8–12 However, despite their belief in disclosure, many physicians do not disclose errors to patients, and if they do, often they do not fully disclose all the details or offer an explicit apology.10–16
The reasons why physicians may not be forthcoming when an error occurs are multifactorial.17 One important barrier is that physicians may feel ill equipped in the disclosure process. Three studies, one of surgeons,15 one of emergency room physicians,18 and one of pediatricians,12 found that only 10%, 12%, and 29%, respectively, had received formal training in how to disclose errors to patients. A nationwide survey of physicians found that 93% felt they needed more training in handling medical errors.19
Experience with medical error begins early in one's career. Many residents, across different specialties, report having made a serious error during their training.12,20–25 Like practicing physicians, most residents feel responsible for errors,20–23 experience a strong emotional reaction to errors,20–22,24,26,27 and believe that patients should be told of errors in their care.12,24,25,28 However, in practice, many residents do not disclose errors to patients or even to their attending physicians.21 A recent study in 2005 found that only 71% of residents discussed a serious error with their attending physician, and only 28% discussed it with the patient or family.22 Although trainees report increasing comfort in disclosing adverse events to patients as their training advances, overall only 56% of trainees in one study were comfortable with disclosing adverse events.28 Over 90% of residents believe that disclosing a serious error would be difficult.12,25 Residents may have never received training in talking with patients about errors,25 and many teachers and preceptors have themselves not been trained in how to respond to trainees’ errors.29
Training residents to be prepared for these difficult discussions is warranted; such training will provide them with a skill that they will use throughout their careers. In this study, we used standardized patient (SP)-based assessment to determine internal medicine (IM) residents’ ability to disclose a medical error and to identify specific aspects of disclosure that are performed particularly well or poorly. We also surveyed the residents about their experiences disclosing medical errors.
We invited all 48 IM residents registered in the second postgraduate year (PGY) in the Department of Medicine at the University of Toronto to participate in the study during 2005. The residents received no incentive. After providing consent, each resident read a short vignette describing a medical error and then disclosed the error to an SP. After the encounter, each resident completed a questionnaire about error disclosure. The study received approval by the University of Toronto Ethics Review Unit.
The error scenario we developed for use in the study (Chart 1) involved the administration of an unintended overdose of insulin. We developed the scenario so that it would be realistic and so that the primary locus of responsibility would be on the resident. To ensure that the situation was realistic and that an error clearly existed, we piloted the error scenario on a group of residents and general internists.
To evaluate the residents’ abilities to disclose medical error, the SPs and an independent physician observer used a rating scale previously developed by Chan et al15 (List 1). The scale is based on studies that have examined what patients would want to know about medical errors in their care,8 and it has previously been used in a study of error disclosure by surgeons.15 The rating scale comprises five components: (1) explanation of the medical facts regarding the error, (2) honesty and truthfulness, (3) empathy, (4) explanation of steps taken to prevent future errors, and (5) general communication skills. Each of the components assesses the presence or absence of specific skills and is scored on a three-point scale (1 = not performed at all; 2 = attempted but either not complete or not effective; 3 = well performed). For each of the five, we calculated an aggregated mean score of 1 to 3. The SP and physician raters also gave the residents a separate, gestalt-based, overall assessment score of disclosure on a five-point Likert-like scale (1 being low and 5 being high).
Five SPs from the University of Toronto Standardized Patient Program participated in the study. All had significant prior experience in evaluating physicians’ communication skills. The SPs received two specific training sessions. During the first session, the SPs familiarized themselves with the scenario and the rating scale, and they practiced the role. During the second session, they met with a study investigator (L.S.) to practice the scenario in pilot interactions with a volunteer resident, and they used the rating scale to score the performances. This session was used to clarify specific items on the rating scale and to ensure that there was agreement between the ratings of the SPs and the study investigator.
To allow for all residents to complete the interaction, each resident had 1 minute to read the scenario and then a maximum of 10 minutes to interact with the SP (none needed more time). We videotaped each interaction with an unobtrusive video camera. At the end of each encounter, the SP provided the resident with verbal feedback and completed the rating scale. The residents completed a questionnaire (Appendix 1) immediately after they received the SP feedback. A physician rater, one of us (L.S.), viewed the videotapes at a later date and scored each encounter on a second rating form.
For scheduling purposes, we split the residents into two groups who completed the study during two different sessions—the second one a week after the first.
We performed all statistical analyses using SPSS 15.0 (Chicago, Illinois). To determine the group mean scores on each of the five components, we calculated the mean of the skills within that component. We also determined the mean of the overall scores. To detect any difference in the performance ratings by each of the two raters attributable to session, gender of the resident, prior error disclosure experience, prior error disclosure instruction, and specific SP, we performed one-way ANOVAs.
We used Cronbach alpha to calculate the overall internal consistency of the skills. It was also used to calculate the internal consistency of the skills within each component. We calculated interrater reliability using a two-way random, absolute-agreement, single-measure intraclass correlation coefficient.30
We performed one-way ANOVAs to detect any difference in the residents’ self-ratings by session, gender, prior error disclosure experience, and prior error disclosure instruction. We used Spearman rho correlation coefficients to compare the residents’ self-ratings with the observed (SP and physician) ratings.
Of the 48 eligible residents (29 male, 19 female), 39 graduated from Canadian medical schools (25 from the University of Toronto, 14 from other universities), and 9 were trained outside of Canada (7 in the Middle East, 2 in the United Kingdom). Five residents (3 Canadian, 2 foreign trained) declined to participate, and 1 resident (Canadian trained) was unable to participate because of a last-minute emergency. Therefore, we used data from 42 residents (25 male, 17 female; 35 Canadian trained, 7 foreign trained) in the analysis.
Error disclosure rating scale
Interrater reliability coefficients between the SP and the independent physician observer were good to excellent for the five components (0.51-0.80) and the overall score (0.74). The internal consistency coefficient for all 19 skills on the rating scale was 0.91. For skills within the components, internal consistency coefficients ranged from 0.49 to 0.88. The ratings given by the observer and the SP were very similar (Table 1).
One-way ANOVAs did not detect any differences in the residents’ performances based on session, sex of the resident, prior error disclosure experience, prior error disclosure training, or interaction with a specific SP.
All 42 residents completed the questionnaire; however, one resident failed to answer the questions on the second page. Therefore, most of the questionnaire results are for 42 residents, but for items that appeared on the second page, there are only 41 responses.
Face validity and perceived responsibility.
The majority of the residents (33/42; 78.5%) found the error disclosure scenario to be either somewhat or very realistic. Of 41 residents, 33 (80.5%) believed that responsibility for the error was shared between individuals, whereas 8 (19.5%) attributed sole responsibility for the error to a specific individual. Of 41 residents, 35 (85.4%) named the physician (themselves) as either fully (4/35; 11.4%) or partly (31/35; 88.6%) responsible for the error. Six residents (14.6%) did not assign any responsibility to themselves. Four of 41 residents (9.8%) named the patient as partly responsible; the remaining residents did not place any blame on the patient. Of the 41 who answered this question, 19 (46.3%) named the hospital as either fully or partly responsible for the error, whereas 28 (68.3%) named other health professionals. Ten of the residents (24.4%) named the general category of “other.” The most frequent combinations of shared responsibility were physician (the resident him- or herself), hospital, and other health professional (10/41; 24.4%), followed by just the physician (the resident) and other health professional (7/41; 17.1%).
Experience and feedback.
Over half of the residents (27/42; 64.3%) had previous experience in disclosing a medical error to a patient during their training. Of these residents, only 7 of 26 (26.9%) reported receiving any feedback from an attending or more senior resident about their performance. In contrast, the majority of residents, both those with and those without prior experience in error disclosure, reported receiving general communication feedback during their training: 21/26 (80.8%) and 13/15 (86.7%), respectively.
Training and needs.
Although, as stated above, 27 of the 42 residents (64.3%) had previously disclosed a medical error to a patient, only 21 of 41 (51.2%) had ever received any training in disclosing medical error. A greater number of residents who received their undergraduate medical degrees in Canada reported previous error disclosure training in medical school compared with those trained outside Canada (20/34, 58.8% versus 1/7, 14.3%). Of the 21 residents who had received training in disclosing medical error, most received training either during medical school (15; 71.4%), and a few received training in a combination of both medical school and residency (6; 28.6%). No residents had received training at only the resident level. The majority of residents (32/41; 78.0%) were either unsure about or thought that help in disclosure did not exist at their hospitals. Almost all the residents (38/41; 92.7%) thought that training in error disclosure would be useful and relevant to their careers.
Residents with previous instruction in error disclosure rated themselves significantly higher (on a scale from 1 to 5) than did residents who had no training with respect to error disclosure skills (3.76 versus 3.30, F[1, 39] = 5.32, P = .03) and comfort disclosing errors (3.76 versus 3.20, F[1, 39] = 5.48, P = .02) in the study scenario with SPs. They also rated their disclosure skills (3.59 versus 3.11, F[1, 24] = 4.59, P = .04), but not their comfort (3.47 versus 3.22, F[1, 24] = 0.74, P = .40), higher in previous clinical encounters than those who had received no training. In contrast, if the residents had prior experience with error disclosure, there were no differences compared with residents without prior experience on self-ratings of disclosure skill (3.59 versus 3.47, F[1, 40] = 0.34, P = .57) or comfort (3.56 versus 3.40, F[1, 40] = 0.36, P = .54) in the study scenario with SPs.
The residents’ self-ratings of their performance were significantly correlated both with the independent observer's ratings of communication and overall scores (0.571 and 0.474, respectively, P < .05 for each) and with the SPs’ ratings (0.465 and 0.459, respectively, P < .05 for each).
Discussion and Conclusions
Residents demonstrate skills in disclosure to SPs, but our study suggests that nonetheless, residents should improve these skills. Overall, residents score well on some aspects of error disclosure, such as explaining medical facts, but they do not completely or effectively perform well in other areas. Conveying future error prevention is a particularly deficient disclosure skill amongst residents.
There are several reasons why residents may fall short in error disclosure. Inconsistent messages about disclosure exist; many risk managers still advise against apology in fear of litigation.31 National guidelines for error disclosure were published only in 2008, and trainees may not yet be aware of them.6 Lack of training may also have impacted the residents’ performance; 49.8% of residents had not received any training. Residents may also be unaware of resources; 78.0% of the trainees in our study were either unsure about or thought that help in disclosure did not exist at their hospitals. In particular, the low scores on error prevention may reflect either that residents do not consider prevention at all, or that if they do, they do not think this is important to disclose to patients. Physicians often fail to recognize that patients want to know that physicians and hospitals have learned from an event so that the error is not repeated.8 Practicing surgeons also scored lower on prevention than on other components of disclosure.15
Encounters with medical errors begin early in training; almost two thirds of PGY2 IM residents at our institution already had experience with disclosing medical errors. We did not ask residents whether these were serious or minor errors. Our number is much higher than those in studies in which residents were asked about disclosure of their most serious errors,12,21,22 but it is comparable with the rate of resident disclosure of minor errors in a recent study.25 Although residents need to be equipped with knowledge and skills, the majority of those with error-disclosing experience had never received any feedback on their performance from a more senior resident or attending. This may reflect a lack of preceptors’ knowledge and comfort in dealing with errors made by trainees.29 Attending physicians also likely need training to improve their own skills and comfort in error management and disclosure.
We were pleased that half the residents had received some form of training in error disclosure during their medical education—the majority at the medical student level. A recent study of pediatric residents also found that about half had received training in error disclosure.12 Although there are still many residents not receiving training, many more are receiving it than are practicing surgeons, emergency physicians, and pediatricians.15,18,12 Similar to the finding that 93% of practicing physicians feel that they need further training in the overall handling of medical errors,19 the vast majority of the residents in our study (93%) said that they would find additional training on error disclosure to be useful. They are a receptive audience for learning these important skills.
Residents’ self-ratings were significantly higher if they had received prior instruction in error disclosure. Although it is possible that residents with prior instruction overestimate their ability, resident self-ratings on the scenario correlated with our independent ratings, suggesting that residents may be able to accurately critique their own performances. No significant differences in resident or observer ratings between trainees with and without prior experience with medical errors existed. This implies that mere experience does not increase skill and comfort with disclosure, and specific training is necessary. Other work has demonstrated that experience alone does not improve resident communication skills, and dedicated training programs are likely required.32 Efforts to teach medical students error disclosure using SPs have resulted in subjective improvement in students’ self-awareness, knowledge, and confidence in disclosure.33,34 Recently, a small study used SPs to pilot an objective structured clinical exam scenario and a rating tool for the disclosure skills of surgical residents.35
Numerous toolkits in error disclosure have been developed.36–38 The next steps for educators are to incorporate such resources into core residency education. Our study demonstrates that discussing future error prevention is an area for particular emphasis. Furthermore, educators should evaluate the effectiveness of any disclosure training, including longitudinal follow-up to determine whether the learned behaviors endure and whether education changes long-term practice and attitudes toward disclosure. It will also be important to study patient experiences with error disclosure as greater numbers of doctors receive training.
There were a number of limitations to our study. First, the number of participants was small, potentially limiting the power of our analyses. Despite the difference in the resident self-ratings by prior instruction and the correlations between resident self-ratings and our independent ratings, our independent ratings on the error disclosure rating scale failed to find an effect by prior instruction. Although it is possible that there truly was no observable difference in the disclosure behaviors between the two groups, it is also possible that our study lacked the power to detect a difference. Second, all the residents were from the same year and training program at one institution, limiting the generalizability of the results to residents in other specialties or programs. Third, we used the same scenario for all residents. Because we split the residents into two sessions that were spaced one week apart, a breach in the security of the scenario content was possible. There is debate in the literature regarding the extent of the effect that reusing scenarios may have on performance.39–42 We did not find any significant differences in performance between the two groups of residents.
Exposure to medical errors begins early in one's medical career, and disclosure of medical error is now the standard of practice. Physicians, including those in training, must be prepared for these discussions. Experience alone is not sufficient to accomplish this training. Establishing skills and comfort in error disclosure early will provide a valuable asset throughout a physician's career. Areas exist for improvement in residents’ error disclosure abilities, in particular with respect to expressing future error prevention. Disclosure curricula need to be implemented and their effectiveness studied. Almost all residents would welcome such training.
1 Wu AW, Cavanaugh TA, McPhee SJ, Lo B, Micco GP. To tell the truth: Ethical and practical issues in disclosing medical mistakes to patients. J Gen Intern Med. 1997;12:770–775.
2 Rosner F, Berger JT, Kark P, Potash J, Bennett AJ. Disclosure and prevention of medical error. Arch Intern Med. 2000;160:2089–2092.
3 Hebert PC, Levin AV, Robertson G. Bioethics for clinicians: Disclosure of medical error. CMAJ. 2001;164:509–513.
4 Kalra J, Massey KL, Mulla A. Disclosure of medical error: Policies and practice. J R Soc Med. 2005;98:307–309.
5 Gallagher TH, Studdert D, Levinson W. Disclosing harmful medical errors to patients. N Engl J Med. 2007;356:2713–2719.
8 Gallagher TH, Waterman AD, Ebers AG, Fraser VJ, Levinson W. Patients’ and physicians’ attitudes regarding the disclosure of medical errors. JAMA. 2003;289:1001–1007.
9 Gallagher TH, Waterman AD, Garbutt JM, et al. U.S. and Canadian physicians’ attitudes and experiences regarding disclosing errors to patients. Arch Intern Med. 2006;166:1605–1611.
10 Vincent JL. Information in the ICU: Are we being honest with our patients? The results of a European questionnaire. Intensive Care Med. 1998;24:1251–1256.
11 Gallagher TH, Garbutt JM, Waterman AD, et al. Choosing your words carefully: How physicians would disclose harmful medical errors to patients. Arch Intern Med. 2006;166:1585–1593.
12 Garbutt J, Brownstein DR, Klein EJ, et al. Reporting and disclosing medical errors: Pediatrician's attitudes and behaviours. Arch Pediatr Adolesc Med. 2007;161:179–185.
13 Blendon RJ, DesRoches CM, Brodie M, et al. Views of practicing physicians and the public on medical errors. N Engl J Med. 2002;347:1933–1940.
14 Espin S, Levinson W, Regehr G, Baker GR, Lingard L. Error or “act of God”? A study of patients’ and operating room team members’ perceptions of error definition, reporting, and disclosure. Surgery. 2006;139:6–14.
15 Chan DK, Gallagher TH, Reznick R, Levinson W. How surgeons disclose medical errors to patients: A study using standardized patients. Surgery. 2005;138:851–858.
16 Kaldjian LC, Jones EW, Wu BJ, Forman-Hoffman VL, Levi BH, Rosenthal GE. Disclosing medical errors to patients: Attitudes and practices of physicians and trainees. J Gen Intern Med. 2007;22: 988–996.
17 Kaldjian LC, Jones EW, Rosenthal GE, Tripp-Reimer T, Hillis SL. An empirically derived taxonomy of factors affecting physicians’ willingness to disclose medical errors. J Gen Intern Med. 2006;21:942–948.
18 Hobgood C, Xie J, Weiner B, Hooker J. Error identification, disclosure, and reporting: Practice patterns of three emergency medicine provider types. Acad Emerg Med. 2004;11:196–199.
19 Robinson AR, Hohmann KB, Rifkin JI, et al. Physician and public opinions of quality of health care and the problem of medical errors. Arch Intern Med. 2002;162:2186–2190.
20 Mizrahi T. Managing medical mistakes: Ideology, insularity, and accountability among internists-in-training. Soc Sci Med. 1984;19:135–146.
21 Wu AW, Folkman S, McPhee SJ, Lo B. Do house officers learn from their mistakes? JAMA. 1991;265:2089–2094.
22 Hobgood C, Hevia A, Tamayo-Sarver JH, Weiner B, Riviello R. The influences of the causes and contexts of medical errors on emergency medicine residents’ responses to their errors: An exploration. Acad Med. 2005;80:758–764.
23 Jagsi R, Barrett TK, Weinstein DF, Campbell EG, Hutter M, Weissman JS. Residents report on adverse events and their causes. Arch Intern Med. 2005;165:2607–2613.
24 Muller D, Ornstein K. Perceptions of and attitudes towards medical errors among medical trainees. Med Educ. 2007;41:645–652.
25 White AA, Gallagher TH, Krauss MJ, et al. The attitudes and experiences of trainees regarding disclosing medical errors to patients. Acad Med. 2008;83:250–256.
26 Engel KG, Rosenthal M, Sutcliffe KM. Residents’ responses to medical error: Coping, learning, and change. Acad Med. 2006;81:86–93.
27 West CP, Huschka MM, Novotny PJ, et al. Association of perceived medical errors with resident distress and empathy: A prospective longitudinal study. JAMA. 2006;296:1071–1078.
28 Sorokin R, Riggio JM, Hwang C. Attitudes about patient safety: A survey of physicians-in-training. Am J Med Qual. 2005;20:70–77.
29 Mazor KM, Fischer MA, Haley HL, Hatem D, Quirk ME. Teaching and medical errors: Primary care preceptors’ views. Med Educ. 2005;39:982–990.
30 Shrout PE, Fleiss JL. Intraclass correlations: Uses in assessing rater reliability. Psychol Bull. 1979;86:420–427.
32 Yudkowsky R, Downing SM, Ommert D. Prior experiences associated with residents’ scores on a communication and interpersonal skill OSCE. Patient Educ Couns. 2006;62:368–373.
33 Halbach JL, Sullivan LL. Teaching medical students about medical errors and patient safety: Evaluation of a required curriculum. Acad Med. 2005;80:600–606.
34 Gunderson A, Mayer D, Tekian A. Breaking the cycle of error: Patient safety training. Med Educ. 2007;41:518–519.
35 Chipman JG, Beilman GJ, Schmitz CC, Seatter SC. Development and pilot testing of an OSCE for difficult conversations in surgical intensive care. J Surg Educ. 2007;64:79–87.
39 Swanson DB, Stillman PL. Use of standardized patients for teaching and assessing clinical skills. Eval Health Prof. 1990;13:79–103.
40 Niehaus AH, DaRosa DA, Markwell SJ, Folse R. Is test security a concern when OSCE stations are repeated across clerkship rotations? Acad Med. 1996;71:2879.
41 Rutala PJ, Witzke DB, Leko EO, Fulginiti JV, Taylor PJ. Sharing of information by students in an objective structured clinical examination. Arch Intern Med. 1991;151:541–544.
42 Wilkinson TJ, Fontaine S, Egan T. Was a breach of examination security unfair in an objective structured clinical examination?A critical incident. Med Teach. 2003;25:42–46.