Improving Feedback From Residents to Preceptors: A Quality Improvement Project

Preceptors providing clinical supervision for residents in outpatient settings need feedback to improve their teaching. Feedback on one’s performance is considered vital for improvement and development of expertise.¹ However, when feedback is delivered as a summative evaluation, it may affect preceptors’ self-efficacy and self-concept, and cause anxiety, embarrassment, and other negative reactions.^2-4

Many tools for evaluation of clinical teaching^5-9 are expressly summative, in which judgments are rendered rather than descriptions of how particular behaviors of preceptors affect learners. Feedback, instead, should identify specific preceptor behaviors that hinder or support resident learning.¹⁰ Preceptors, then, can understand how their specific teaching behaviors are perceived in a psychologically safe environment that facilitates their objective examination of their attitudes, emotions, feelings, skills, and views about teaching.^3,11,12

Realizing that our standard method of providing feedback from residents to preceptors was evaluative, we used a quality improvement approach to implement a new format for generating behavior-specific feedback that communicates residents’ observations as requests for change, making feedback more formative than summative. We used the Standards for QUality Improvement Reporting Excellence in Education (SQUIRE-EDU) to formulate this report.¹³

From 2018 to 2025, with a hiatus during the pandemic, we instituted twice monthly feedback sessions in a family medicine residency training program, which included 24 residents (eight per class). Residents were precepted by core teaching faculty (N = 6) and family physicians (N = 32) in clinical practice. Most preceptors have at least 3 years of precepting experience.

We adopted a novel approach to giving and receiving feedback using a model of communication that had not previously been used to deliver feedback in this setting. Nonviolent Communication (NVC), also called compassionate communication, is a philosophy and style of communication that seeks to bring out feelings of affection, respect, and empathy in interpersonal relationships.¹⁴ It follows a four-step format: (1) When you (specific behavior); (2) I feel (a specific feeling); (3) I need (a need not met); (4) Please (request for specific behavior). Examples of the transformation of feedback are in Table 1.

Monthly feedback collecting sessions were incorporated into the protected didactic time of the second- and third-year residents’ classes. Interns were excluded because they had only limited interaction with preceptors. Sessions were facilitated by one or two consistent nonprecepting faculty (J.F. and A.F.S.). These faculty had basic training in using the NVC approach; one (A.F.S.) learned the philosophy and practice by reading the book about it,¹⁴ and the other, with background in psychology, learned the process as part of her training. At each feedback collecting session, the class of residents identified one or two preceptors to whom they wanted to give feedback. For each preceptor, residents then described behaviors that they valued and would like the preceptor to keep doing and behaviors that they would like the preceptor to change. The process typically identified four to 10 instances or behaviors for each preceptor. Through moderated group discussion, residents converted this raw feedback into four-part statements using the NVC format. The completed NVC feedback statements (without the raw feedback) were sent to the preceptor accompanied by an invitation (Appendix) for the recipient to work with one of the NVC facilitators to further develop their teaching and communication skills if they desired.

Residents, at the beginning of their second year, received a 1.5 hour introduction that taught the NVC format and allowed them to practice differentiating judgments (eg, “the preceptor talks too much”) from behaviors (eg, “You frequently interrupt our presentations”). Residents were taught how to elucidate their feelings and how to name their needs using the NVC feelings and needs inventory.¹⁵ To create a safe and honest space for residents to share, the feedback was collective and anonymous and was not shared with the program director or department chair, or used for punitive purposes. Residents were reminded that they had other venues (including Accreditation Council for Graduate Medical Education surveys and labor union meetings) to evaluate faculty members.

To continuously improve the process, at the end of each session residents shared effective parts of the session and areas that could be improved. We also devoted a significant part of a single session with experienced (ie, third year) residents to gain their insights on the NVC feedback process.

To determine preceptors’ reactions to NVC feedback, we conducted semistructured interviews with a purposeful sample of preceptors to hear their reactions and reflections. We analyzed notes from these interviews to identify themes in their reactions to the NVC process.

In addition, we used several ways to evaluate how the NVC structure changed the quality of the feedback. For each initial (raw) feedback statement converted to the NVC format, two investigators (A.F.S., R.S.) independently determined how the process changed the descriptions of the precepting behavior. Similarly, two investigators also independently evaluated the initial feedback statements for the presence of feelings and needs conveyed by the residents.

This project was considered a quality improvement project and thus did not constitute human subject research. It was exempted by the Chair of the Cambridge Health Alliance Institutional Review Board.

We conducted 63 feedback sessions with 72 residents. Many preceptors received feedback several times over the time span of our study, though the measures to ensure anonymity prevented us from tracking exact rates of feedback.

Through iterative assessments, we conducted seven plan-do-study-act cycles of quality improvement,¹⁶ which are outlined in Table 2. The original process collected feedback in person; with the requirement to move many activities online during the pandemic, we now conduct the process online. We use a shared online document (available from the authors) that each resident can add to or edit in real time, which allows rapid collection of the initial feedback statements. As residents become accustomed to the process and require less facilitation, they can be split into online “rooms” to generate feedback for more than one preceptor.

The NVC process changes original, raw feedback statements in several ways. Initial feedback statements are often general judgments of the behavior rather than the action itself. For example, a typical raw feedback statement might contain a judgment, such as “It’s fun to work with her!” The NVC process changed this statement to a behavior-based comment: “When you ask how things went at the end of a session . . .,” which allows the preceptor to understand the behavior being commented upon. Descriptions of behavior in the raw feedback were changed in 67.3% of instances. If a specific behavior was mentioned in the raw feedback, the NVC process resulted in minor changes to the wording in 18.5%, major changes to content in 19.0%, and from a judgment to a specific behavior in 29.9% of instances.

None of the initial feedback statements expressed a need, and residents infrequently spontaneously described how they were affected (ie, their feelings) by a precepting behavior (23.3%). The NVC process encouraged residents to express a wider array of feelings; raw feedback statements that described a feeling used only one of five common feeling words; over the course of this analysis, we identified 147 distinct feeling words used because of the process.

Residents described how they valued having some agency to provide feedback on their precepting experience. They appreciated the protected time and felt that the group process improved the feedback quality, preventing an isolated experience of one resident being conveyed to a preceptor. Experienced residents noted that the communication format is transferable to other interpersonal contexts, including patient care and personal relationships. Several residents mentioned noticing a change in a preceptor’s behavior after receiving feedback. Residents reported that some preceptors had discussed the NVC feedback with them in a positive way or shared with residents their efforts to change behavior or continue doing what is valued.

Thirteen preceptors were interviewed, comprising a mix of early and more experienced preceptors, and those who received feedback recently or one or more years before the interview. All interviewed preceptors were appreciative they received feedback, even if they didn’t agree with all of it. Preceptors remarked that the feedback was “helpful to know what’s working”; “increased my confidence”; and was “a positive experience.” Although receiving feedback in this format was new to most of the preceptors, they respected the process. Some preceptors recalled specific behaviors mentioned in the feedback given to them several years ago and described how they used the feedback.

We identified several themes across the interviews. Preceptors often mentioned they felt “invisible” and that the NVC feedback acknowledged the effort they put into precepting. Some preceptors commented on the complexity of providing the appropriate teaching and supervision of residents while also ensuring patient safety. Other preceptors were surprised that their nonteaching behaviors, such as bringing snacks, sharing personal information, or simply expressing concern to the residents regarding their well-being, were valued by residents.

Over the past 7 years, we have implemented and continually refined a system of collecting feedback from residents and providing it to their preceptors via a novel approach. Using two faculty facilitators, we incorporated NVC skills training into residency didactics to provide preceptors with nonjudgmental, actionable behavior requests that were valued by both residents and preceptors.

NVC is designed to improve direct interpersonal communication. While not used previously to shape feedback, NVC has been used to improve physician-patient relationships,¹⁷ interprofessional collaboration,^18,19 education,^20,21 and mentoring and coaching,²² and is effective in helping health care professionals to express their observations in a nonjudgmental way.¹⁹ Our goal for using the NVC approach is to create feedback to preceptors that is behavior-specific, actionable, timely, and personalized, yet not personal (ie, that could be interpreted as a personal criticism). Unlike many other forms of feedback, this approach explicitly conveys to the recipient the needs and feelings of the feedback givers. This approach also reflects a feed-forward stance, seeking to influence future actions rather than evaluating only past behavior.²³

This process is different from traditional feedback approaches that typically involve evaluative statements. Rather, the NVC format conveys to preceptors how their teaching is experienced by residents and the degree to which it meets the residents’ perceived needs, honoring the concept that every interpersonal interaction has a content and relationship aspect and should thus be approached with intention and respect.²⁴ Indeed, facilitators often must help residents find the right nuance of emotional valence¹⁹ when describing feelings (eg, “stressed” vs “overwhelmed”). Additionally, residents often are unaccustomed to identifying and expressing their needs and must be prompted to use the provided inventory. Eliciting the needs and feelings of the residents keeps the emotional component alive, an important component to relationship-based dialogue. This approach may be more palatable to preceptors, allowing the message to be processed and acted on rather than generating a defensive response.²⁵

The use of this approach is limited by the size and structure of our residency program, which is medium-sized, with regularly scheduled protected didactic time. Programs that are larger or organized differently might find it more challenging to implement. Implementation requires continual involvement of dedicated faculty to ensure consistency and integrity of the process and to create psychological safety. Additionally, preceptors who received feedback via this process were identified by residents, rather than selected in a systematic manner, limiting those who received feedback. Programs should ensure that other feedback mechanisms are in place to promote a comprehensive approach to faculty development.

While we have focused on the process of incorporating NVC into a residency curriculum and the resulting responses from residents and preceptors who were part of this process, future research should explore outcomes in more depth—whether the NVC approach results in preceptor growth and subsequent improvements in teaching behaviors. Future research also should explore how teaching the NVC format to residents and faculty results in more generalized improvements in communication skills.

We were able to implement a group-based method for collecting feedback from groups of residents and refine it using the NVC approach. The process was valued by residents, and the resulting feedback was valued by preceptors.

Presented at the North American Primary Care Research Group (NAPCRG) 52nd Annual Meeting, November 20–24, 2024, Quebec City, Quebec, Canada.