A Multisite Survey of Pediatric Hospitalists’ Practices in Soliciting Feedback From Trainees Free

In the winter of 2022, the authors designed a survey to evaluate attending physicians’ perceived value of feedback from medical students and residents, quantify the frequency with which attending physicians solicit feedback from trainees, and describe potential barriers to obtaining that feedback. Survey questions were developed utilizing previous literature,^2,6,8  trainee input, and faculty expertise with a framework based on the qualities of exemplary pediatric educators as described by Fromme et al³  and the Clinician Educator Milestone Project.⁴  The logic and rigor of questions were developed to meet the standards put forth by Artino et al.⁹  The validity and logic of questions were confirmed through cognitive think-aloud interviews with 5 hospitalists, with changes incorporated after discussions with the research team.¹⁰ 

The survey assessed each respondent’s opinion of the importance of trainee feedback, the frequency of asking for and receiving feedback from trainees in each domain of exemplary pediatric educators, satisfaction with existing feedback received from trainees, and potential barriers to upward feedback from trainees. The 4 domains were defined in our survey by using definitions from Fromme et al.³  The survey (Supplemental Fig 3) defines patient skills as “medical knowledge, knowledge acquisition, communication with patient, clinical care plans, system knowledge, etc,” personal qualities as “self-reflection, empathy, enthusiasm, motivation, respect for others, professional behavior,” teaching skills as “patient-centered teaching, setting expectations, stimulation of learner problem-solving, setting learning climate, etc,” and leadership skills as “encouraging autonomy, time management, conflict resolution, etc.” The survey had yes-no questions, multiitem Likert scale questions (1 = “Strongly Disagree,” 5 = “Strongly Agree” or 1 = “Always,” 5 = “Never”), and a predetermined list of answer choices to a question on barriers to upward feedback that included a free-text box if the respondent chose “other.”

Participating institutions were a convenience sample of 9 pediatric hospitalist programs in the Northeastern, Central, Southern, and Western regions of the United States. They were selected for diversity in program location, number of non-ICU inpatient pediatric hospital beds, size of medical school, and size of residency (Supplemental Table 4). Physicians who attend as pediatric hospitalists were sent invitations to an anonymous, web-based survey (REDCAP tool) via an e-mail from site leads. Sites included Advocate Children’s Hospital at Oak Lawn and at Park Ridge (Chicago, IL), Children’s Hospital of Richmond at Virginia Commonwealth University (Richmond, VA), Comer Children’s Hospital at University of Chicago (Chicago, IL), LA County-Harbor-UCLA Pediatrics (Los Angeles, CA), Lucile Packard Children’s Hospital at Stanford (Palo Alto, CA), NorthShore University Hospital (Evanston, IL), University of Florida Health Shands Children’s Hospital (Gainesville, FL), University of Rochester-Golisano Children’s Hospital (Rochester, NY), and University of South Alabama Health Children’s & Women’s Hospital (Mobile, AL). All respondents worked with residents and medical students. Respondents reported how many weeks per year they work with students. The answer choices are in intervals of 5 weeks because many hospitalists are on service for 1 or 2 weeks at a time; thus, 5 weeks captures 3 to 5 instances of being on service with students. Participants were sent 2 subsequent reminder emails over 4 weeks in March 2022. Participants received no compensation for completion. Site leads were offered anonymous composite data from their site after study completion.

Deidentified data were collected via REDCap and transferred to Microsoft Excel. The demographic characteristics of respondents were characterized by using frequency and percentage. χ-Square statistics were used to analyze yes-no questions. Wilcoxon rank test was used to compare Likert-scale questions. A P value <.05 was considered significant for an association between 2 questions. All analyses were conducted by using R (version 4.1.3) and Rstudio. The statistical methods of the study were determined with consultation from The University of Chicago’s Biostatistics Laboratory.

This study was determined to be exempt by The University of Chicago’s Institutional Review Board.

Results were gathered from 91 of the 189 surveyed individuals (response rate of 48.1%). Less than one-half of respondents (42, 46.2%) reported receiving formal training on feedback solicitation. More than one-half of respondents (60, 65.9%) reported working as hospitalist attending physicians for 5 or more years. The majority of respondents spent >5 weeks per year working with medical students on their clerkship inpatient pediatric rotation (77, 84.4%) (Table 1).

A majority of respondents strongly agreed or agreed (88, 96.7%) that feedback from medical students was valuable. Similarly, respondents strongly agreed or agreed (89, 97.8%) that resident feedback was valuable. Yet, feedback from residents was considered significantly more valuable (P <.001).

Attending physicians believed trainees could provide useful feedback in all 4 domains associated with a master clinical educator. When comparing feedback given by medical students across domains, feedback on teaching skills was reported to be potentially useful to the largest proportion of attending physicians (90, 98.9%), whereas feedback on patient skills could be useful to the least (71, 78.0%) (Table 2). For feedback given by residents, feedback on teaching skills could be useful to all attending physicians (91, 100%), whereas feedback on patient care skills could be useful to the least (86, 94.5%) (Table 2). In the domains of personal qualities and leadership skills, there was a significant difference between the perceived utility of feedback from medical students and the perceived utility of feedback from residents (P <.001) (Table 2).

Attending physicians were generally unsatisfied with the feedback they received from medical students and residents. Approximately one-quarter of attending physicians were satisfied with the quantity (23, 25.3%) and quality (25, 27.5%) of feedback from medical students (Table 2). Approximately 40% of attending physicians were satisfied by the quantity (38, 41.8%) or quality (37, 40.7%) of feedback from residents (Table 2). Satisfaction with the quantity and quality of feedback from residents was significantly greater than satisfaction with feedback from medical students (P <.05) (Table 2). Attending physicians believed the 3 largest barriers to upward feedback for both medical students and residents were (1) trainee lack of comfort with giving feedback (medical students: 81, 89.0%; residents: 69, 75.8%), (2) lack of awareness that providing feedback is part of the trainees’ role (medical students: 72, 79.1%; residents: 64, 70.3%), and (3) fear of retaliation (medical students: 63, 69.2%; residents: 55, 60.4%).

With medical students, attending physicians were most likely to frequently or always ask for feedback on their teaching skills (56, 62.2%) and least likely to frequently or always ask about leadership qualities (20, 24.1%; Table 3; Fig 1). With residents, they were most likely to frequently or always ask for feedback on their teaching skills (67, 74.5%) and least likely to frequently or always ask about patient care skills (37, 43.0%; Table 3; Fig 2). When comparing attending feedback solicitation practices from trainees, attending physicians were statistically more likely to report asking for feedback frequently or always from residents than medical students (P <.001; Table 3).

With medical students, attending physicians were most likely to frequently or always receive feedback on teaching skills (34, 39.5%) and least likely to frequently or always receive feedback on patient care skills (9, 14.5%; Table 3; Fig 1). With residents, they were most likely to frequently or always receive feedback from residents on teaching skills (42, 47.2%) and least likely to frequently or always receive feedback on patient care skills (20, 24.1%; Table 3; Fig 2). Similar to the data on asking for feedback, attending physicians were statistically more likely to report receiving feedback frequently or always from residents than from medical students (P <.001; Table 3).

This is the first study, to our knowledge, that explores pediatric hospitalists’ perceptions of upward feedback in clinical educator competencies and reveals the frequency with which attending physicians solicit such feedback from trainees. We identified that attending physicians would like feedback from trainees in all domains demonstrated by exemplary pediatric educators.³  We found variance in attending physicians’ feedback solicitation practices and found that they are dissatisfied with the quantity and quality of feedback from trainees. Given that there are barriers to providing upward feedback,^6,8,11  it is unsurprising that there is an unmet need for feedback from trainees.

We found that attending physicians value feedback from medical students and residents, which is consistent with previous literature and the fact that residents are expected to provide feedback to attending physicians (Supplemental Fig 4).^8,12–14  This study adds a layer of detail to our understanding of attending physicians’ perceptions of this feedback. Attending physicians value feedback more in certain domains, primarily teaching and personal qualities, than in other domains. This idea that attending physicians want feedback on particular components of their performance is consistent with publications revealing that attending physicians’ opinion of the usefulness of feedback is influenced by the type of feedback received.^6,8,11  Our findings suggest that, although all areas of feedback are valued by attending physicians and are worth addressing, additional learning opportunities for faculty and trainees that focus on increasing upward feedback in teaching skills and personal qualities might result in higher yield changes for attending physicians.

The surveyed pediatric hospitalists were dissatisfied with the quantity and quality of feedback from trainees, and they were significantly more dissatisfied with the quality and quantity from medical students than residents. This dissatisfaction may be, in part, explained by the variance in attending feedback solicitation practices. One conceptual framework of the components of effective upward feedback suggests that safety and trust in the deliverer-recipient relationship are critical to the attending physicians’ judgment of the validity and appropriateness of feedback.^11,15,16  Residents are more experienced members of the team and work more closely and often longitudinally with attending physicians, and our findings corroborate the idea that this connection may result in attending physicians’ more positive opinions of resident feedback. Moreover, the Association of American Medical Colleges explicitly encourages residents to provide feedback to attending physicians,¹⁴  which may help normalize residents giving feedback. For all trainees, there are multiple documented barriers to upward feedback in medical education that may be contributing to this dissatisfaction.^6,8  We identified that most attending physicians think a lack of trainee comfort, a lack of awareness that giving feedback is a trainee responsibility, and a fear of retaliation are barriers to trainees providing upward feedback. These barriers may contribute to the dissatisfaction that attending physicians experience with the feedback they obtain. The path to generating satisfying feedback is likely influenced by attending and institutional differences that are worth identifying at the local level. An ethnographic study on providing feedback at a children’s hospital provides evidence that efforts to change the institutional culture around feedback can result in improved experiences with communicating feedback.¹⁷  Attending physicians may be able to create safer environments for medical students and residents to voice concerns if they preface discussions with assurances that there will no repercussions on their evaluations and if they create tones of learning and opportunity for change before feedback. Focusing the conversation on learning and change for future learners may strengthen teacher-student alliances and create more comfortable environments for upward feedback.^15,17 

Asking for feedback is one strategy to generate specific and actionable feedback.^8,18,19  The attending physicians in our sample reported requesting feedback from residents more than students. Yet, attending physicians expressed potential value in medical student feedback and dissatisfaction with feedback from medical students. Currently, there is no guidance from professional societies (e.g., the Association of American Medical Colleges or American Academy of Pediatrics) regarding feedback solicitation by attending physicians from medical students. Formalizing recommendations regarding feedback solicitation from medical students may generate more frequent or satisfying feedback. Although we found 1 curriculum on how trainees can solicit feedback from attending physicians,²⁰  we could not find publicly available curricula to help attending physicians solicit feedback from trainees. Yet, nearly one-half of the respondents of this study reported formal training on feedback solicitation. In thinking about the development of upward feedback, providing both trainees and attending physicians with skills and advice on how to engage in bidirectional feedback could mitigate dissatisfaction with feedback.

Our study has several limitations. First, this survey is subject to response bias. Second, the generalizability of this study is limited by our study population and sample size. Next, the barriers to upward feedback that attending physicians could select were presented as a predetermined list on the basis of previous literature with an opportunity to add free text of an unlisted barrier.^6,8,11  Although minimal attending physicians selected other, we could be missing additional barriers to upward feedback. Next, de novo survey instruments have limitations in their ability to obtain empirical evidence, although the Fromme et al³  framework and survey piloting provided validity evidence for survey definitions. Finally, this was a self-report survey based on attending perceptions, therefore results do not reflect the perception of learners.

Although upward feedback from trainees to attending physicians was nearly unanimously viewed as valuable by the pediatric hospitalists in our sample, attending physicians were inconsistent in their solicitation practices, and they were dissatisfied with the feedback they received from trainees. Attending physicians were more likely to ask for and receive feedback from residents than medical students. The exploration of barriers to upward feedback and increased training of trainees in feedback delivery could improve attending physicians’ ability to generate helpful feedback. Future steps in researching upward feedback include investigating the trainee perspectives on upward feedback, including barriers to and strategies in providing it. Future work should also focus on developing curricula on upward feedback for both attending physicians and trainees and determining if curricula impact the quality and quantity of feedback given.