Characterizing Resident Preferences for Faculty Involvement and Support in Disclosing Medical Errors to Patients Open Access

Faculty role modeling and formal training are critical for residents to develop error disclosure skills and positive attitudes regarding disclosure.^1–7  In academic medical centers, residents often share the responsibility to disclose errors alongside their faculty⁶  and want to be directly involved in communicating with patients and families.¹  Therefore, faculty supervisors and training programs would benefit from knowing specific ways to support residents when the need for error disclosure arises in the clinical setting.

To our knowledge, little is known about resident preferences as they relate to faculty involvement and supervision during error disclosure. Characterizing resident needs and whether they vary depending on training program, learner stage, or error characteristic (error type, severity, and proximity) is essential if faculty are to ensure appropriate error disclosure to patients and involve residents in the process. We surveyed residents across 6 programs at the University of Toronto about their preferences for faculty supervisor involvement and support during error disclosure.

The University of Toronto has a large graduate medical education department with 75 accredited residency programs. There are 5 fully affiliated adult teaching hospitals and 1 major pediatric hospital. We surveyed residents in 3 medical (internal medicine, pediatrics, and emergency medicine) and 3 surgical (general surgery, orthopedic surgery, and obstetrics and gynecology) programs.

We created a survey (provided as online supplemental material) to determine the degree of faculty supervisor involvement perceived to be most helpful in specific error contexts, and what elements of faculty support are most important. We focused on content and response processes to enhance the validity of our tool.⁸  We derived content from literature searches and rich qualitative data previously collected by the authors on this subject.¹  Draft versions of the survey were reviewed by 2 patient safety content experts. To determine response processes, 12 fourth-year internal medicine residents pilot tested the survey for readability, clarity, and length. We incorporated their feedback into the final version of the survey and removed items consistently reported to be redundant.

We defined a medical error in our survey as “the failure of a planned action to be completed as intended (ie, error of execution) or the use of a wrong plan to achieve an aim (ie, error of planning).”⁹  We asked residents about their prior experiences with error disclosure and their preferences for faculty supervisor involvement in specific error contexts. These contexts included error type (diagnostic, management, prescribing, procedural, and communication); error severity¹⁰  (death or severe permanent disability, mild/moderate permanent disability, severe temporary pain or suffering, or mild/moderate temporary pain or suffering); and error proximity¹¹  (either the resident, another resident or medical student, or nursing or other health professionals felt most responsible). Residents rated their preference for faculty involvement for each scenario among 5 levels of resident autonomy in the disclosure process that we collapsed into 3 categories: (1) resident led without direct faculty supervision, (2) resident led with faculty involvement, and (3) faculty led with or without resident involvement.

Residents used a 5-point Likert scale (1, unimportant, to 5, very important) to rate the importance of faculty guidance to prepare for specific elements of error disclosure communication based on a well-established framework¹²  that includes what to disclose, explaining facts, explaining consequences, managing consequences, and offering an apology. They also rated the importance of different types of faculty support around the disclosure period, including providing feedback, examples from personal experience, external resources, institutional resources, and personal support.

Between November 2014 and March 2015, we distributed electronic surveys via SurveyMonkey (SurveyMonkey Inc, San Mateo, CA) to all residents in the 6 medical and surgical programs. We sent 2 e-mail reminders to encourage participation and offered residents the opportunity to enter a drawing for 1 of 20 $50 gift certificates to a bookstore as an incentive.

The University of Toronto Research Ethics Board approved this study.

We analyzed survey data descriptively, reporting continuous data using means and standard deviations and categorical data using counts and percentages. We grouped participants into medical (internal medicine, pediatrics, and emergency medicine) or surgical (general surgery, obstetrics and gynecology, and orthopedic surgery) training programs and junior (postgraduate year 1 [PGY-1] and PGY-2) or senior (PGY-3 through PGY-5) residents for subgroup analyses. We performed a chi-square test for comparisons across error type, severity, and proximity to determine differences in resident preferences for faculty involvement and support, and followed up with a Mantel-Haenszel chi-square analysis to identify differences in the distribution of responses between medical and surgical residents and junior and senior residents. We carried out data analysis using SPSS version 23.0 software (SPSS Inc, Chicago, IL). We considered P < .05 statistically significant and applied a Bonferroni correction by dividing P = .05 by the number of subgroups for all subgroup analyses (eg, for error type, we divided P = .05 by 15 subgroups and considered P = .003 to be statistically significant).

Of 538 eligible residents, 192 responded to the survey (response rate 36%). Table 1 summarizes study participant characteristics, which were representative of the total population when comparing resident distribution across training program and level. The majority of residents reported having been involved in a harmful medical error in the past (78%, 150 of 192), having prior experience disclosing a medical error (72%, 106 of 148), and having had formal training in error disclosure (83%, 160 of 192). A significantly greater proportion of medicine residents reported having prior experience disclosing medical errors compared with surgical residents (78% [76 of 98] versus 60% [30 of 50], P = .025).

For all errors, irrespective of error type or severity, more than 90% of the residents wanted to have direct involvement in the error disclosure process. Their preference, however, for the level of faculty involvement varied by type of error (table 2). Specifically, for diagnostic and management errors, residents were more likely to seek faculty involvement in disclosing the error, whereas for prescribing and communication errors, residents were more likely to feel comfortable disclosing the error independently without direct faculty supervision. Subgroup analysis revealed that surgical residents were more likely to prefer a faculty-led approach to error disclosure than medical residents for management, prescribing, and procedural type errors. Senior residents were more comfortable independently disclosing all error types except for communication errors.

Residents' preference for faculty involvement in an error disclosure also varied by error severity (table 2). For severe life-threatening errors, residents were significantly more likely to have faculty lead the disclosure (61%, 106 of 173), although 38% (65 of 173) of respondents indicated that they still wanted to participate in the discussion with faculty support, and senior residents had a stronger preference for a resident-led approach with faculty involvement than junior residents. For minor errors, residents preferred a resident-led approach, with 47% (81 of 173) specifying that they would be comfortable leading the conversation independently without direct supervision. Finally, for errors where residents felt most responsible, 95% of residents (164 of 173) preferred a resident-independent or resident-led approach, which was significantly higher than for errors where another resident (64%, 111 of 173) or another health professional (65%, 112 of 173) was primarily responsible (P < .001).

We explored the types of faculty support residents preferred in the period leading up to and following an error disclosure encounter. Before disclosing an error, residents were significantly more likely to seek faculty advice on how to manage consequences, and less likely to seek faculty advice on how to explain facts and make an apology (figure 1a). After an error disclosure has taken place, residents reported that the most important way faculty could support residents was to provide feedback (figure 1b).

Our study confirms results of prior studies demonstrating that the majority of residents have personally disclosed medical errors to patients and families and have received formal training in how to carry out disclosure conversations effectively. Importantly, most residents wanted to participate directly in error disclosure conversations. However, residents' desired levels of faculty involvement in error disclosure conversations varied by error type and severity, with greater faculty involvement sought for diagnostic and management errors as well as for life-threatening errors.

Medical residents were more likely to disclose errors without direct faculty involvement than surgical residents. This may relate to the fact that a larger proportion of medical residents had prior experience disclosing errors to patients (78% versus 60%). However, there may be other differences in training culture, faculty expectation, or organizational processes in medical versus surgical programs that influence how residents view their role relative to faculty in the error disclosure process. A survey of surgical and nonsurgical residents suggested that punitive responses to error by senior members of the health care team might impede transparent disclosure of errors, with surgical residents disproportionally affected.³ 

These findings help to inform faculty development efforts to support error disclosure training. Whereas residents strongly value many aspects of faculty guidance and support before and after an error disclosure conversation, certain aspects (eg, describing how to manage consequences of the error, explaining how to prevent future errors) were seen by residents as particularly relevant and warrant specific attention. Studies exploring the patient's perspective on error disclosure have also highlighted these elements as being particularly important.¹²  Evaluations of resident error disclosure skills have consistently identified the need to develop skills to prevent future errors.^1,13 

Residents responded that they highly valued receiving feedback from faculty about the disclosure process. This lends further credence to our prior research that identified informal learning in the clinical setting as critical to the development of error disclosure skills in residency.¹  This finding is also consistent with the increased focus on formative feedback in graduate medical education and the implementation of competency-based medical education.¹⁴  While residents may seek autonomy and, at times, disclose errors independently without direct supervision, they must also recognize the importance of “getting it right.” Therefore, it is imperative for faculty to directly observe residents when they disclose errors and provide them with coaching and feedback. Given the high stakes surrounding medical errors, it is also critical to develop guiding principles regarding direct observation and assessment to inform entrustment decisions related to error disclosure.

This study is limited by a moderate response rate (36%), and our findings may not be fully representative. We also did not include certain disciplines in our study (eg, family medicine, anesthesiology, or neurological surgery). The self-reported nature of our data means that our findings center primarily on perceived needs. Although we found no differences in residents seeking personal support from their faculty, the survey may not have adequately addressed this topic, as faculty may play an important role in helping residents cope with the negative feelings of being involved in medical error.¹⁵  While our findings provide an important lens to inform curriculum design and faculty development efforts, programs should combine them with needs assessment data that reflect the local training environment. Future research should emphasize how best to deliver feedback to residents and focus on establishing criteria to guide entrustment decisions with respect to error disclosure.

Residents are willing participants in the error disclosure process but have specific preferences for faculty involvement and support. These findings can inform faculty development to ensure appropriate support and supervision for residents when disclosing errors to patients.