Progressive autonomy, culminating in entrustment for unsupervised practice, is a hallmark of medical training.1–6 Trust relates to both the person doing the trusting (supervising physician) and the person who is trusted (trainee) and can entail a belief, decision, and action. Trust is incremental and variable, based on individuals and circumstances.7 In medicine, workplace-based assessment of trustworthiness is complex, occurring in a dynamic environment with rotational learning and duty hours restrictions that can limit contact between supervisors and trainees.8–12 Recently, medical educators have explored how faculty physicians (attendings) make decisions to trust trainees5,13–20 and have identified five key factors influencing these decisions: supervisor, trainee, their relationship, task, and context.4 How supervisors tailor their supervisory practices based on trust continues to be a high-priority area for research and faculty development.1–4,13–15,18
Although attending physicians play a critical role in medical training, significant supervision of junior residents also comes from senior residents. The Accreditation Council for Graduate Medical Education asserts that residents can supervise more junior trainees once the former have “demonstrated competency in medical expertise and supervisory capability,” as guided by milestones.21 However, resident-as-teacher literature suggests that although residents spend a quarter of their time supervising more junior colleagues, they receive little education on how to supervise effectively.22–25 As new supervisors, residents may approach trust as a guide to their supervisory behaviors differently than attendings.
Drawing from concepts of situated learning and workplace learning, we hypothesized that interactions, activities, and work practices influence how residents develop trust to guide their supervisory decisions.26–28 Situated learning theory posits that social norms and the interactions between residents and their team members will shape residents’ approaches to trusting the interns they supervise.26 Workplace learning suggests that trainees learn through authentic and central participation.27 As residents oversee interns, they develop their own way to assess trust, guided by opportunities for modeling or coaching from their own former supervisors.27,28
In this study, we use a multi-institutional mixed-methods design, informed by the frameworks of situated learning and workplace learning, to explore how residents develop trust in the interns they supervise.
This is a mixed-methods study with an exploratory sequential design. We aimed to understand resident trust in greater depth than is possible through applying either qualitative or quantitative approaches alone.29–31 First, we interviewed postgraduate year (PGY) 2 and 3 internal medicine (IM) residents at a single institution to explore their trust in interns (PGY1) for providing clinical care. We used these insights to generate a model for resident trust, which we then shared in focus groups to confirm and, as necessary, revise the model. Using this model, we developed a quantitative survey to determine which factors most influenced residents’ trust. The institutional review board at the following institutions approved this study: University of California, San Francisco (UCSF); University of Colorado (UC); University of Florida (UF); University of Minnesota (UMN); and University of Pennsylvania (UPenn).
We considered reflexivity in this work32 by reflecting on the composition of the research team that included an IM resident, clinician–educators with experience supervising resident teams, a residency program director (who did not participate in data collection to avoid influencing respondents), and a medical education researcher. During e-mail and in-person discussions throughout the project, the team considered how their stance influenced their interpretations of findings.
Phase 1: Interviews and focus groups
We adapted interview questions from a guide used to interview attendings about trust.20 We wanted to elicit from residents information about their experiences with forming trust and supervising, and how they learned to trust and supervise (Supplemental Digital Appendix 1, https://links.lww.com/ACADMED/A336). Interviewees provided demographic information including their age, gender, year in training, and the number of months they had worked in the wards (wards months) as a supervisor.
Participants and setting.
We invited PGY2 and PGY3 IM residents at UCSF for interviews. We chose to focus on IM because IM residents have supervisory responsibilities within a complex context, and we felt our work could build on the existing literature on trust and supervision in IM settings.5,7,13,14,33 Of 118 total residents, we e-mailed invitations to 30 using random purposive sampling based on training year. Eligible residents had completed at least one month of inpatient supervisory experience.
We piloted the interview guide with 7 IM residents in March 2014 and revised it for clarity and flow of questions. We included these pilot interviews in the dataset as the changes did not alter the interview content. One of us (L.S.) and a research assistant conducted 22 additional interviews in person or by telephone between May and September 2014. All interviews were professionally transcribed and deidentified. Participants received a $20 gift card.
Outcome measures and analysis.
We analyzed interviews using grounded theory and sensitizing concepts (e.g., the five factors of trust) from the literature.4,34,35 Two of us (L.S., K.E.H.) and the research assistant each independently reviewed three transcripts and then met to discuss, develop, and reach consensus on coding categories. Then, using these codes, one of us (L.S.) and the research assistant independently coded all transcripts and met to discuss any discrepancies and reach consensus.29 We continued interviews as we coded and noted the same concepts emerging after approximately 20 interviews. However, to ensure that no new concepts arose, we interviewed 29 residents before concluding the interviews and beginning to develop our model.
Triangulation and confirmation.
The coding categories informed themes, which then formed the model for resident trust. To triangulate our findings to confirm and add credibility to findings from a single institution, we conducted two focus groups with four or five participants each at two different study institutions (UC, UF) in September and December 2014. Three recent graduates and six residents of other programs where residents had similar responsibilities (pediatrics, general surgery) participated (without incentive), to avoid including IM residents who would subsequently be invited to complete the survey. Three of us (E.M.A., D.T., H.E.H.) moderated the focus groups. Questions addressed the meaning of trust, challenges of balancing trust and supervision, and how trust changes over time (Supplemental Digital Appendix 2, https://links.lww.com/ACADMED/A336). During the focus groups, each of which lasted about an hour, participants also reviewed the resident trust model we had developed and discussed whether it resonated with their experiences, and whether anything was missing. Notes from each focus group affirmed and helped refine our final resident trust model.
Phase 2: Quantitative surveys
On the basis of our final model, we developed quantitative questions mapped to each of the model themes. We piloted the survey for clarity and timing with 7 attendings and 15 recent IM graduates not included in the interviews, focus groups, or survey. After incorporating feedback from the pilot participants, all authors finalized and approved the survey.
The final survey (Supplemental Digital Appendix 3, https://links.lww.com/ACADMED/A336) contained 5 demographic questions and 38 trust questions in two sections: (1) factors contributing to trust (31 questions on a continuous scale of 0–100 where 0 = not at all, 100 = very much); and (2) influence of resident experiences on trusting (7 questions on a continuous scale of 0–100 where 0 = strongly disagree, 100 = strongly agree).
Participants and setting.
We invited all PGY2-and-above IM residents at UCSF and four other institutions (UC, UF, UMN, and UPenn) to participate in a survey. The five schools provided a convenience sample of IM residents (n = 478) from dispersed geographic areas. IM residents at all sites worked on inpatient medicine teams as supervisors for one or two interns and one to three students, supervised by an attending physician. The 478 residents included IM–pediatrics and IM–dermatology residents through the PGY5 level because they have the same IM supervisory responsibilities.
Site investigators e-mailed, via resident listservs, invitations that included a link to a Web-based, anonymous survey (2015 Qualtrics, LLC), between January and March 2015. Site investigators sent four weekly reminder e-mails. Residents who completed the survey received a $10 electronic gift card.
Outcome measures and analysis.
We analyzed survey results using explor atory factor analysis; we assigned items to factors based on factor loading values greater than or equal to 0.3.36 We analyzed the two survey sections (contributors to trust, influence of resident experiences on trusting) separately because the question stems were different and preliminary analyses confirmed that these items did not load within factors across question types. Three of us (L.S., P.S.O., K.E.H.) reviewed our findings for interpretability and fit within our model’s predetermined constructs. Then, for each factor, we calculated the Cronbach alpha coefficient to determine internal consistency, determined means to describe relative importance, and computed Pearson coefficients to assess for associations between months of supervisory experience and influence of factors. We completed all statistical analyses using SPSS version 23 (IBM, Armonk, New York).
Phase 1: Interviews and focus groups
We interviewed 29 residents (16 PGY2 and 13 PGY3 residents) who had a mean of 3.4 wards months’ experience as supervisors (standard deviation [SD] = 2.2). Fourteen (48%) were women, and the average age was 29.8 years (SD = 1.9). The interviews lasted, on average, 30 minutes (range 20–43 minutes).
We identified 14 themes in our resident trust model (Figure 1), which mapped to all five factors in the previously described model of trust.4 Our focus group data reinforced these themes, particularly the importance of intern confidence, context, and the relationship between supervisor and supervisee. Theme descriptions with representative interview quotes are shown in Table 1.
Phase 2: Quantitative surveys
Overall, 376 of 478 (79%) residents responded to the survey. Response rates ranged from 61% to 85% by institution. On average, participants had completed 5.3 months (SD = 2.9) as supervisors. Additional demographic information is shown in Table 2.
Exploratory factor analysis yielded 11 factors (8 factors representing 67.1% of variance in items related to contributors to trust; 3 factors explaining 64.2% of variance in items dependent on the influence of resident experience on trusting). The factors mapped to themes in the model from our qualitative findings under context, resident, and intern factors (Figure 2). Reliability scores (Cronbach alpha) and descriptive statistics are shown in Table 3.
Themes within the intern factor were most important to residents in considering trust (Figure 2). Respondents rated interns’ reliability, clinical competence, and propensity to make errors highest when indicating importance to trust (respective means are 86.3 [SD = 9.7], 76.4 [12.9], and 75.8 [20.0]). Contextual factors were also important for trust determinations; these included access to an electronic medical record (EMR), duty hours, and patient characteristics (respective means are 79.8 [15.3], 73.1 [14.4], and 71.9 [20.0]). Resident’s own experience and learning to supervise were important, though less so (means are, respectively, 63.2 [20.3] and 61.0 [17.6]).
In exploring associations between months of supervisory experience and influence of factors, the Pearson coefficient for correlation ranged from −0.07 to 0.17, which was statistically significant but not considered meaningful given that it explains less than 2% of the variance (Table 3).
Discussion and Conclusions
Through this multi-institutional mixed-methods study, we aimed to understand how residents develop trust in the interns they supervise. We found that residents form trust around three of five previously identified factors that contribute to trust formation: intern (trainee), context, and resident as supervisor.4 Residents appear to develop trust in interns in a way that prioritizes interns’ execution of essential patient care tasks safely within the complexities and constraints of the hospital environment. The other two factors, task and relationship, seemed to be important based on our interviews, but were not emphasized in our quantitative results (see below).
The qualitative portion of our mixed-methods design captured residents’ perspectives on how their own evolution of comfort and skill as supervisors influenced their trust decisions. Residents described starting out unsure of how to be effective supervisors and not yet trusting the interns they supervised (Table 1). Over time, they felt better able to determine how much trust was warranted and when they could provide less supervision. This growth in their own ability to develop trust in interns may follow a natural course, similar to that described in the counseling and psychology literature: New supervisors begin with role ambiguity and uncertainty but, through experience, establish greater awareness of their supervisory role, which enables them to trust and relinquish responsibilities to their supervisees.37 Curricular design could facilitate this increasing ability to judge trustworthiness by making the process more conscious, structured, and deliberate.38,39
We learned that when residents see the interns they supervise complete patient care tasks reliably, they become more comfortable with providing less direct supervision, thus allowing interns to have more autonomy. The context, particularly access to an EMR, affords residents the ability to quickly and unobtrusively determine whether intern tasks are completed reliably and accurately, a style of supervision that Kennedy and colleagues13 describe as backstage oversight. Because EMRs allow residents remote access to patient records, interns may be unaware of residents scrutinizing patient orders and records, and thus may feel a greater degree of autonomy. Other influential contextual factors for resident trust included patient characteristics and duty hours. From a patient safety perspective, a large census may obligate more trust. Sick or complex patients necessitate closer and more direct supervision because the stakes of erroneous judgment are higher.4 Interestingly, while duty hours restrictions typically limit residents’ and interns’ time together because of different days off, the resulting patient care transitions enabled residents in our study to get a closer look at the interns’ attention to detail and ability to communicate with consultants and patients. This scrutiny informed the trust assessments of residents and guided their subsequent supervision.
Unexpectedly, our quantitative analysis did not yield separate task- or relationship-specific factors in factor analysis. The reasons these two factors did not manifest as strongly in our quantitative results may reflect the multifaceted nature of how residents form trust in their interns, which is difficult to disentangle in individual survey items. Residents’ trust in the interns they supervise is driven by the interns’ task-oriented role, and as such, residents seem to think about task and intern factors (i.e., ability to complete the task reliably) as intertwined. Our results suggest that response to one item may incorporate other influences on trust simultaneously. For example, residents’ use of EMRs to provide background supervision and determine intern reliability is not only intern and task based but also reliant on the context. Many complex intern tasks (e.g., discussing goals of care) necessitate interpersonal communication, and our survey results revealed that residents considered many specific task items together with interns’ ability to communicate effectively with consultants, nurses, patients and their families, and others. Additionally, simple tasks are those that interns can confidently perform with little or no supervision, and we found that some task items (e.g., ability to call consults) clustered with intern confidence, a finding supported by previous literature.40
As for relationship, trust inherently occurs within a relationship.4,7,20,41,42 Organizational trust literature suggests that the depth and strength of relationships influence trust, and in the absence of a strong relationship, trust becomes more dependent on external factors.7 In our study, relationship was not a strong factor in residents’ trust in interns. This may be because residents do not consciously realize how relationship affects trust, as shown in the literature on the unconscious preferences individuals have for others who are similar to them.38,43 Alternatively, opportunities for relationship and team building are brief in the context of current U.S. training environments.11,12 Residents may have learned to adopt role-based patterns of interaction that enable successful collaboration despite lack of opportunity to form deeper relationships.41 The current structure of medical training seems to foster residents’ ability to gauge trust quickly based on reliable completion of patient care tasks, but it may compromise the depth of their assessments of trust. Lastly, our survey items may have oversimplified the complexities of determinants of trust related to relationship. A survey with more questions about relationship, or perhaps questions on how relationship interacts with other model themes, may have yielded quantitative results more consistent with our qualitative interview findings about the importance of relationship to trust.
Compared with findings from previous studies of attendings’ trust in resident trainees, we found that resident trust in interns is based more on interns’ skill and reliability, key components of overall clinical competence. Although attendings likewise consider competence, they also focus heavily on more holistic and relationship-oriented aspects of trust such as leadership skills, professionalism, disposition, and interactions with the team, staff, and patients.8,20 Perhaps this distinction results from residents’ proximity to their own intern year. Alternatively, the differences in resident and attending trust may represent their differing roles on the team: The resident’s role is to ensure that interns complete patient care tasks reliably while continuing to hone their own clinical knowledge and clinical reasoning. In contrast, the attending’s role is to ensure that residents are providing good oversight, leadership, and role modeling for more junior team members while also overseeing patient care. Some attendings delegate a high degree of responsibility to residents, motivated either by high trust or a minimalist supervisory style44; most residents in our study did not report applying this more hands-off approach. Finally, another possibility is that differences in resident and attending trust determinations could derive from their amount of supervisory experience. The ways that supervisors approach trust change with experience, from residency into early and later practice. Residents are very early supervisors who are expected to continue to develop supervisory skills, and their trust may evolve from a task-based to a holistic and relationship-based model with experience. How experience informs supervisees’ learning is a rich area for further study.
This study has several limitations. Although we included residents from diverse institutions, the generalizability of our results may be limited to IM or other medical specialties in which residents play similar roles. Results may be affected by nonresponse bias, although the overall survey response rate was high. Our data derive from resident self-report; we did not corroborate findings with observations or reports from other team members. Different wording of some of our questions, or additional questions, could have improved the clarity of our factor analysis, although our survey pilots confirmed that trainees found the questions clear and relevant. The residents in our study had, on average, only a few months of supervisory experience, and results may not apply to more experienced supervisors.
In conclusion, we found that resident trust in interns is driven primarily by intern- and context-specific factors as residents seek confirmation that daily patient care occurs safely and reliably. This understanding of how residents trust the interns they supervise can inform residency program efforts to design curricula for residents as supervisors and can extend to supervision training in other health professions. Learning to supervise is an important skill unique from other aspects of teaching and learning in medical training.42 Providing supervision based on trust constitutes a crucial resident skill that not only enables postgraduate trainees to effectively lead their teams but also prepares them for potential roles as attending physicians. A firm understanding of how trust informs supervisory behaviors can serve as a platform for the use of entrustable professional activities to monitor trainees’ development, an increasingly popular method of trainee assessment.45 Our insights on resident trust can help align faculty and resident expectations regarding trust in interns and can inform areas in which faculty may help residents develop as supervisors. Taken together, these findings confirm the relevance of trust as the basis for supervisory decisions even among junior supervisors, and offer rich areas for further study.
Acknowledgments: The authors thank study assistant Joanne Batt for help with conducting interviews and coding transcripts.
1. Schumacher DJ, Bria C, Frohna JG. The quest toward unsupervised practice: Promoting autonomy, not independence. JAMA. 2013;310:26132614.
2. Babbott S. Commentary: Watching closely at a distance: Key tensions in supervising resident physicians. Acad Med. 2010;85:13991400.
3. Halpern SD, Detsky AS. Graded autonomy in medical education—Managing things that go bump in the night. N Engl J Med. 2014;370:10861089.
4. Hauer KE, Ten Cate O, Boscardin C, Irby DM, Iobst W, O’Sullivan PS. Understanding trust as an essential element of trainee supervision and learning in the workplace. Adv Health Sci Educ Theory Pract. 2014;19:435456.
5. Martin SK, Farnan JM, Mayo A, Vekhter B, Meltzer DO, Arora VM. How do attendings perceive housestaff autonomy? Attending experience, hospitalists, and trends over time. J Hosp Med. 2013;8:292297.
6. Dijksterhuis MG, Voorhuis M, Teunissen PW, et al. Assessment of competence and progressive independence in postgraduate clinical training. Med Educ. 2009;43:11561165.
7. Dietz G, Den Hartog DN. Measuring trust inside organisations. Pers Rev. 2006;35:557588.
8. Ginsburg S, McIlroy J, Oulanova O, Eva K, Regehr G. Toward authentic clinical evaluation: Pitfalls in the pursuit of competency. Acad Med. 2010;85:780786.
9. Govaerts MJ, Schuwirth LW, Van der Vleuten CP, Muijtjens AM. Workplace-based assessment: Effects of rater expertise. Adv Health Sci Educ Theory Pract. 2011;16:151165.
10. Weinstein DF, Arora V, Drolet B, Reynolds EE. Residency training—A decade of duty-hours regulations. N Engl J Med. 2013;369:e32.
11. Bernabeo EC, Holtman MC, Ginsburg S, Rosenbaum JR, Holmboe ES. Lost in transition: The experience and impact of frequent changes in the inpatient learning environment. Acad Med. 2011;86:591598.
12. Holmboe E, Ginsburg S, Bernabeo E. The rotational approach to medical education: Time to confront our assumptions? Med Educ. 2011;45:6980.
13. Kennedy TJ, Lingard L, Baker GR, Kitchen L, Regehr G. Clinical oversight: Conceptualizing the relationship between supervision and safety. J Gen Intern Med. 2007;22:10801085.
14. Kennedy TJ, Regehr G, Baker GR, Lingard L. Point-of-care assessment of medical trainee competence for independent clinical work. Acad Med. 2008;83(10 suppl):S89S92.
15. Kennedy TJ, Regehr G, Baker GR, Lingard LA. Progressive independence in clinical training: A tradition worth defending? Acad Med. 2005;80(10 suppl):S106S111.
16. Wijnen-Meijer M, van der Schaaf M, Nillesen K, Harendza S, Ten Cate O. Essential facets of competence that enable trust in graduates: A delphi study among physician educators in the Netherlands. J Grad Med Educ. 2013;5:4653.
17. Stenfors-Hayes T, Hult H, Dahlgren LO. What does it mean to be a good teacher and clinical supervisor in medical education? Adv Health Sci Educ Theory Pract. 2011;16:197210.
18. Sterkenburg A, Barach P, Kalkman C, Gielen M, ten Cate O. When do supervising physicians decide to entrust residents with unsupervised tasks? Acad Med. 2010;85:14081417.
19. Cianciolo AT, Kegg JA. Behavioral specification of the entrustment process. J Grad Med Educ. 2013;5:1012.
20. Hauer KE, Oza SK, Kogan JR, et al. How clinical supervisors develop trust in their trainees: A qualitative study. Med Educ. 2015;49:783795.
21. Accreditation Council for Graduate Medical Education. Frequently asked questions: ACGME common duty hour requirements. Effective July 1, 2011. Updated June 18, 2014. https://www.acgme.org/acgmeweb/Portals/0/PDFs/dh-faqs2011.pdf
. Accessed February 3, 2016.
22. Post RE, Quattlebaum RG, Benich JJ 3rd. Residents-as-teachers curricula: A critical review. Acad Med. 2009;84:374380.
23. Wamsley MA, Julian KA, Wipf JE. A literature review of “resident-as-teacher” curricula: Do teaching courses make a difference? J Gen Intern Med. 2004;19(5 pt 2):574581.
24. Hill AG, Yu TC, Barrow M, Hattie J. A systematic review of resident-as-teacher programmes. Med Educ. 2009;43:11291140.
25. Snell L. The resident-as-teacher: It’s more than just about student learning. J Grad Med Educ. 2011;3:440441.
26. Lave J, Wenger E. Situated Learning: Legitimate Peripheral Participation. 1991.New York, NY: Cambridge University Press.
27. Billett S. Learning in the Workplace: Strategies for Effective Practice. 2001.Crows Nest, Australia: Allen & Unwin.
28. Dornan T, Boshuizen H, King N, Scherpbier A. Experience-based learning: A model linking the processes and outcomes of medical students’ workplace learning. Med Educ. 2007;41:8491.
29. Creswell JW. Research Design: Qualitative, Quantitative, and Mixed Methods Approaches. 2014.4th ed. Thousand Oaks, Calif: SAGE.
30. Creswell JW. Qualitative Inquiry and Research Design: Choosing Among Five Approaches. 2012.3rd ed. Thousand Oaks, Calif: SAGE.
31. Caracelli VJ, Greene JC. Crafting mixed-method evaluation designs. New Dir Eval. 1997;1997:1932.
32. O’Brien BC, Harris IB, Beckman TJ, Reed DA, Cook DA. Standards for reporting qualitative research: A synthesis of recommendations. Acad Med. 2014;89:12451251.
33. Hinchey KT, Iwata I, Picchioni M, McArdle PJ. “I can do patient care on my own”: Autonomy and the manager role. Acad Med. 2009;84:15161521.
34. Bowen GA. Grounded theory and sensitizing concepts. Int J Qual Methods. September 2006;5. https://www.ualberta.ca/~iiqm/backissues/5_3/PDF/bowen.pdf
. Accessed February 3, 2016.
35. Corbin J, Strauss A. Basics of Qualitative Research: Techniques and Procedures for Developing Grounded Theory. 2008.3rd ed. Thousand Oaks, Calif: Sage Publications.
36. Stevens JP. Applied Multivariate Statistics for the Social Sciences. 2009.5th ed. New York, NY: Routledge.
37. Majcher J-A, Daniluk JC. The process of becoming a supervisor for students in a doctoral supervision training course. Train Educ Psychol. 2009;3:6371.
38. Hill CA, O’Hara O’Connor EA. A cognitive theory of trust. Minnesota legal studies research paper no. 05-51. December 2005. http://papers.ssrn.com/sol3/papers.cfm?abstract_id=869423
. Accessed February 3, 2016.
39. ten Cate O, Snell L, Carraccio C. Medical competence: The interplay between individual ability and the health care environment. Med Teach. 2010;32:669675.
40. Kennedy TJ, Regehr G, Baker GR, Lingard L. Preserving professional credibility: Grounded theory study of medical trainees’ requests for clinical support. BMJ. 2009;338:b128.
41. Lewicki RJ, Tomlinson EC, Gillespie N. Models of interpersonal trust development: Theoretical approaches, empirical evidence, and future directions. J Manag. 2006;32:9911022.
42. Kilminster S, Cottrell D, Grant J, Jolly B. AMEE guide no. 27: Effective educational and clinical supervision. Med Teach. 2007;29:219.
43. Gingerich A, Regehr G, Eva KW. Rater-based assessments as social judgments: Rethinking the etiology of rater errors. Acad Med. 2011;86(10 suppl):S1S7.
44. Goldszmidt M, Faden L, Dornan T, van Merriënboer J, Bordage G, Lingard L. Attending physician variability: A model of four supervisory styles. Acad Med. 2015;90:15411546.
45. Ten Cate O. AM last page: What entrustable professional activities add to a competency-based curriculum. Acad Med. 2014;89:691.