Exploring the Construct of Psychological Safety in Medical Education

Abstract

Purpose 

Psychological safety (PS) is recognized as key in health professions education. However, most studies exploring PS in medical education have focused on mistreatment, thus focusing on what PS is not. The authors set out to explicitly explore learners’ concept of PS in the context of medical education to better understand and define PS and its educational consequences for medical students.

Method 

This descriptive exploratory study was conducted in the context of a pilot peer-assisted learning (PAL) program. The program brought together residents and medical students for 16 semiformal learning sessions. Eight medical students from a PAL program were recruited for semistructured interviews to explore their experiences of PS. Transcripts were thematically analyzed using an inductive approach, and social ecological theory was integrated in the later stages of analysis.

Results 

Students described PS as not feeling judged. Having supportive relationships with peers and mentors improved PS. Students’ sense of PS appeared to free them to focus on learning in the present moment without considering the consequences for their image in the eyes of others. Feeling safe also seemed to facilitate relationship building with the mentors.

Conclusions 

A sense of PS appears to free learners from constantly being self-conscious about projecting an image of competence. This enables learners to be present in the moment and concentrate on engaging with the learning task at hand. The authors propose that the term “educational safety” be used to describe a relational construct that can capture the essence of what constitutes PS for learners.

Medical education has produced little formal literature defining psychological safety (PS) or exploring the methods that might be invoked to maximize PS in the educational context. Instead, many researchers have documented the social, verbal, physical, and sexual mistreatment facing medical students^1–7 and its consequences for student well-being.^8–17 This focus on the ways in which a safe learning environment is frequently absent in medical education seems to suggest the implicit assumption that removing these harmful behaviors will achieve PS. This inference is found in some broader literature also, which has focused on harmful workplace environments.^18–23 However, it may be an erroneous leap of logic to assume that the removal of such negative experiences would by itself create a psychologically safe environment. Similar to the psychological literature that actively explores happiness alongside unhappiness,^24,25 studying these clearly problematic behaviors in medical education might well be complemented by defining and exploring PS in positive terms—what it is—to understand not merely what actions to avoid but also what efforts can be exerted to directly promote PS.

PS as a construct has origins in management science where Edmondson’s²⁶ and Kahn’s²⁷ seminal works found that PS empowered interpersonal risk-taking. Since Edmondson’s work, modern workplace research has shifted to examine how to maximize PS and consequent benefits^28–33 rather than just minimizing harmful activities. The field of nursing has also grappled with PS, and they have defined it along similar veins of empowering risk-taking.^34,35 However, the large body of evidence on nursing mistreatment suggests that providing a definition may be only the first step toward operationalization of such concepts of trust and safety.^36–40

Because medicine has built concepts of safety by focusing on student mistreatment and its consequences, we have not fully captured the dimensions of a safe environment for medical learners. The narrow framing of PS limits the understanding of both the extent of the problem and approaches for improving safe learning environments. Developing a more nuanced understanding of PS’s relationship to learning in medical education may facilitate moving beyond merely mitigating negative experiences to encouraging the development of more supportive learning environments. Our aim was to inductively develop rich insights into the concept of PS for medical learners to galvanize the development of robust theoretical constructs on defining and applying the idea of PS for medical education initiatives.

Method

Setting

To study the elements of a safe learning environment, we situated this study within a Peer Mentorship in Medical Education (PMME) program, a peer-assisted learning (PAL) program founded in 2015 by a resident physician and a second-year medical student at the University of British Columbia (UBC). The literature suggests that PAL fosters students’ PS because it is known to foster a positive socioemotional environment.^41–45

The UBC PMME program was designed with 1 to 2 residents serving as mentors for a group of 6 to 8 first-year and second-year medical students. The mentees were selected via rigorous processes that included written application and interviews. Over the course of a year, each group completed 15 to 20 hours of workshops (each of which lasted about an hour) and 10 to 20 hours of consult scenarios (each of which lasted 3–5 hours and included significant hands-on practice). The workshops and consult scenarios each covered a clinical scenario from history taking and physical exam findings, through to clinical management. Each term concluded with a debriefing session and a dinner. Further details on the mentor recruitment and debriefing, mentee recruitment, and curriculum are provided in Appendix 1 and Table 1.

Table 1: Sample Curriculum

Participants

Following approval by UBC’s Behavioural Research Ethics Board (approval number H16-01105), student participants for the study were recruited from the active PMME cohort. Eight out of 8 PMME students approached agreed to participate in the study, consisting of 3 men and 5 women. All participants were recruited via informal conversation and emails for confidential, semistructured interviews to explore the constituents and effects of PS.

Data collection

Data collection and analysis were based on qualitative research principles. Data collection involved individual semistructured interviews. Each interview began with open-ended interview questions: “How did a sense of PS affect your learning?” or “Can you tell if there were times in PMME when you felt that the learning environment was emotionally or psychologically safe or unsafe?” Follow-up questions and probing techniques also helped elicit more information about PS in the broader medical curriculum (Appendix 2). The interviews were conducted over Skype or Google Hangouts, recorded, and transcribed verbatim with identifying information removed. In accordance with qualitative research principles, participants were recruited until a sufficient explanation of the phenomenon of interest was achieved.⁴⁶ We reached sufficient conceptual depth to nuance our understanding of PS by the eighth interview.

Data analysis

In this descriptive and exploratory study, data were thematically analyzed using an inductive approach where themes were strongly linked to the data.⁴⁷ S.H.-T.T. and C.H. engaged in thematic analysis by generating initial codes and then exploring patterns in the data⁴⁷ to examine any instances that addressed the sense of PS or indicated consequences of PS. Negative instances that were either unique or contradictory to other findings were sought to avoid premature judgments that could bias the final outcomes. After each iteration of analysis, S.H.-T.T. met with G.R. and L.N. Their feedback and questions deepened thematic analysis and generated insights that were developed into refined themes. As analysis progressed from description of themes to interpretation of themes,⁴⁷ S.H.-T.T. drew on social ecological theory (SET)⁴⁸ to better represent the meaning of the themes. SET conceptualizes an individual’s agency and action as a product of institutional, interpersonal, and intrapersonal factors. Based on previous theories of behavioral change and individual development, McLeroy and colleagues explicitly proposed SET to conceptualize health promotion within the following 5 spheres: (1) intrapersonal factors that pertain to characteristics of individuals; (2) interpersonal processes that describe social network dynamics; (3) institutional factors characterizing organizational characteristics, rules, and regulations; (4) community factors that delineate relationships between organizations, institutions, and networks; and (5) public policy which depicts laws and policies.⁴⁸ This flexible, multilevel theory served as an organizing framework for understanding the forces that drive PS. For the purposes of this study, the first 3 levels—intrapersonal, interpersonal, and institutional-level factors—merited the main consideration.

Results

The first section will depict how the students described the constituents of their sense of PS as embedded in the institutional, interpersonal, and intrapersonal aspects of their educational experience. The second section then elaborates on students’ rich descriptions of the affordances of a safe learning environment and illuminates students’ perceptions of the self-reinforcing networked nature of PS.

Constituents of PS

A dominant thread across themes was the students’ descriptions of the absence of a need to constantly self-monitor when they felt safe. Students could fully focus on learning in the moment without having to judge whether their performance was appropriate or adequate relative to perceived standards. This absence of a need to self-monitor appeared in the 3 major levels outlined in SET: institutional, interpersonal, and intrapersonal factors. For ease of readability, each aspect is described sequentially, beginning with the broader institutional-level factors and ending with the narrower intrapersonal levels.

Institutional factors.

Students suggested that both the lack of formal assessment and the unstructured nature of the learning agenda were critical to their sense of PS in PMME. The lack of formal assessments and specific learning objectives, which would otherwise force them to restrict their learning to what mattered to educational leaders and preceptors, enabled students the freedom to explore their curiosities. They described that learning was more self-directed and uninhibited without the constraint of expectations. As one participant stated,

There’s less formality in everything and that in itself makes it feel safer. Like we’re having conversations with each other… . It’s not like it’s set, we have to accomplish this task before we leave this room today … we tackle things as much as we’re able to and we do as much as [we] can on a case and the residents fill in gaps, and help direct us. And I guess yeah the word is safer. It just makes it more comfortable, the learning. Knowing that there’s no specifics, and we don’t have to hit some sort of criteria by the end of the day. (P#1)

Interpersonal factors.

Students appeared to also believe that their sense of PS arose in part because of the interpersonal relationships they had developed in the context of the PMME.

Their sense of PS through interpersonal connections arose because they felt understood and cared for as a person, not only by their peers but also by the preceptors. They emphasized that the preceptors appreciated and supported the mentees’ emotional state, abilities, and goals for themselves rather than focusing on “their own benefit or their own sort of aggrandizement” (P#6). As one student articulated,

[T]hese mentors, like whether they be second-year med students or residents, that they care about me as, like, a student, as a future clinician. But also as a person with my own goals and dreams and what I want to do. (P#5)

Students articulated the sense that because of these positive relationships, they did not feel constantly judged for their actions, comments, and questions. There was a strong sentiment that their participation would not incur negative opinions or repercussions. Thus, they could focus on the learning at hand without having to consider the consequences of demonstrating ignorance or a suboptimal performance:

I’d say I felt pretty safe to completely land on my face if I needed to. And knowing that you guys weren’t going to think I wasn’t trying hard or anything like that. So I was willing to do my best, but if I didn’t really know what I was doing knowing that you guys were there to help me through it and to teach me how to do it properly rather than judge me for not doing it right the first time. (P#7)

Participants also indicated that the friendly, humorous environment helped them feel at ease with each other, allowing them to collectively focus on learning instead of being distracted by individualistic competition. One student said,

The overall friendly environment. Like there’s so many jokes thrown around. It’s having a not very competitive environment where everyone has the best interest in particular to learn. (P#2)

This idea was elaborated by a student contrasting the experience in PMME with the overall competitiveness in medical school:

[I]n the hospital in like a group setting and I know that someone loves cardiology and they’re just asking all these intense questions about cardiology, I’m less likely to ask my stupid question about why the left heart is bigger than the right heart or something like that… . I kind of fade into the background, because there’s some gunner… . I think there’s a lot of comparing in med school, at least for me. (P#1)

Participants also suggested that as they became increasingly fluent with the group’s interpersonal dynamics, their sense of PS increased further because they could anticipate how others would interpret their actions or comments. One participant elaborated this aspect of interpersonal ease by stating,

You know how they’re going to behave. You can anticipate maybe their reactions or their willingness to answer your questions. Yeah, just how they behave in the group dynamic, that helps you anticipate, like, you know, like, when you should ask and is it safe to ask now. Is it right to ask in this context. (P#5)

Intrapersonal factors.

Intrapersonal factors refer to whether individuals perceived themselves capable of upholding their end of the “learning contract,” which seemed related to the participants’ expectations of themselves as learners. This sense of expectations in PMME and medical school was often established through social comparisons. Students elaborated that the less knowledgeable they felt compared with others, the less comfortable they felt and the more they focused on projecting an image of competency. One participant described this phenomenon by stating,

I mean if the group were full of people who knew so much more than me, even in the year below me, I, you know, I would probably still enjoy it but I wouldn’t feel as comfortable. (P#4)

This same participant contrasted the absence of social positioning experienced in the PMME context with his broader experience in medical training, saying,

Medical school in a lot of ways socially at least feels like high school and so you can kinda think of how, like in high school like I had this stupid sense of superiority as long as I was in a higher grade than people and vice versa. (P#4)

Consequences of safe vs unsafe environments

When the participants elaborated their sense of feeling unsafe, they described increased awareness of how they were performing relative to expectations. They often stated that they felt anxious, ashamed, or inadequate. Several participants described that they would be reticent to engage—by, for example, withholding questions, speaking out less, or expressing fewer personal needs—until they could understand the consequences of their actions. One participant explained this concept by stating,

I need time to test the waters, like, figure out where everyone is at. Who everyone is. And I’ll stay relatively reserved until I can confidently suss that out. Until I can confidently figure out where everyone is coming from and, like, how I fit into the group. (P#5)

In contrast, when the environment afforded PS, in PMME for example, the participants often described feeling that they could share where they were emotionally and intellectually with the group without concern of drawing scorn. One participant described this as follows:

[F]eeling free to ask questions and knowing that those questions will not be seen as stupid questions or something like that allows me to comfortably engage a lot more easily with the material. (P#8)

In essence, the students described that they were better able to focus on the present task at hand and more fully explore the learning opportunity when they were feeling a sense of PS.

Participants suggested that when there were consistently supportive interactions in PMME, this led to a self-reinforcing relationship that resulted in an increasingly tight-knit group. This interpersonal bond reinforced a sense of PS through closeness. For example, they suggested that failing in front of others with minimal consequences improved the interpersonal trust they had within the group. The sense that they could take risks without consequences allowed them to feel they had a support system that would mediate mistakes in the future. One participant elaborated on the concept by saying,

If you feel like you’ve failed in front of them or you’re vulnerable in front of them, getting those experiences as them being able to help you and assist you. And building that trust that okay, so even if society really seems to frown upon doctors making mistakes, I can find support with my fellow doctors or medical professionals. (P#7)

The participants also suggested that feeling safe with each other allowed them to broaden their conversations to focus on subjects beyond the clinical approaches covered. Subjects included, for example, medical career or residency matching processes. One mentee described how they benefited from informally drawing on the expertise of their mentors:

I’m comfortable now to send [MENTOR 2] and [MENTOR 1] an email, and I don’t do it all the time, but I’m comfortable enough to do that and say “hey look” you’ve committed to this group and you’ve politely, at least, I’m sure it’s more than just being polite, expressed that we can talk to you guys about other things than just the case we’re supposed to learn every week. (P#3)

These types of informal interactions appeared to strengthen the sense of rapport that the group members felt toward each other.

Discussion

Most medical education literature has described PS through reporting on unsafe environments characterized by social, verbal, physical, and sexual mistreatment.^1–7 This dominant focus on student mistreatment has failed to adequately operationalize broader notions of PS and learning. Our findings reveal dimensions in terms of what constitutes PS for learners. These novel findings suggest that exposure to a safe learning environment is interrelated with PS and may help mentees deepen the sense of flow, prolong intellectual engagement, and create cyclical reinforcement of group-level PS.

The findings demonstrated that a sense of PS is strongly affected by how much the students need to continuously assess themselves against what they feel is expected of them, whether it is determined through curriculum objectives, comparison with other peers, or internally derived standards. The safer the students felt, the more they could focus on the present moment and engage in learning. The students described how they experienced less need to present a competent image; thus, they were more able to express their thinking and emotions authentically.

These findings add nuance to the construction of PS as described in the current health professions education literature. The nursing and management science literature suggests that PS frees individuals to take interpersonal risks without fear of punishment.^27,30,34,35 However, our findings extend beyond notions of PS as “taking risks.” Rather, participants suggested that “safety” implied that they were not monitoring for “risks” to manage their images; they were engaging fully in the learning experience without a sense of risk. Therefore, we propose that PS in medical education be termed “educational safety,” which we define as the subjective state of feeling freed from a sense of judgment by others such that learners can authentically and wholeheartedly concentrate on engaging with a learning task without a perceived need to self-monitor their projected image. Key factors from our study that support educational safety are summarized in Figure 1.

Figure 1: Defining factors that affect educational safety.

Relationships and educational alliance

Our findings demonstrate the importance of relational support and fit well with the concept of the educational alliance. An educational alliance describes the quality of educational rapport between the educators and the student, emerging from “the learner’s liking, trusting and valuing of the preceptor and belief that these feelings are mutual.”⁴⁹ Theoretically and empirically, a stronger educational alliance increases the students’ uptake of the preceptors’ feedback.⁴⁹ In our context, interacting without judgment and feeling understood in a familiar interpersonal context with a mentor appeared to represent the presence of an educational alliance. The findings also depict how relationships seemed to foster increased informal interactions outside of the learning context. The increased frequency and possibly the caring, nonjudgmental quality of interactions outside of the learning sessions may further strengthen the relational dynamics of the members. This may deepen the sense of educational safety when the members return to the learning session. The possibility of establishing such a networked support cycle highlights the importance of bringing to our language around PS the construct of group-level educational safety.⁵⁰

Our insights could also be elaborated through the lens of attachment theory. Attachment theory was first developed to describe children’s relationship with their caregiver. When infants readily receive support, they feel securely attached to the caregiver and freely explore their surroundings.⁵¹ In contrast, inconsistent or threatening care can lead to poor attachment, evidenced by highly anxious behavior under stress, and even avoidance of the caregiver altogether.⁵¹ In adult workplace settings, leaders can foster secure attachments for employees by providing support during stressful times and encouraging exploratory behaviors.⁵² Secure workplace attachment improves the sense that others at work are trustworthy and leads individuals to take interpersonal risks.^52,53 Our participants’ descriptions of PS show striking resemblance to the phenomenon of secure attachment. The students described a sense of trust when they perceived that the environment was friendly, caring, and nonjudgmental. They also perceived that disclosing weaknesses would not result in penalization. As a result, the participants described feeling more comfortable exploring challenging topics, taking risks and experiencing failures, and acknowledging the limits of their understanding in front of the group. We propose that when the medical students start their training, they have little medical knowledge and experience to fend for themselves, akin to an “academic child.” When students feel safely attached to the mentors, they may more readily explore the “academic playground” and expand their zone of proximal development.⁵⁴ With educational safety playing such a key role in mentorship, these insights have relevance to mentorship and leadership within a medical education context. Future work could explore how safety is constructed within mentorship contexts.

Improved flow

The results further illuminated that when the participants felt safe, they were less preoccupied about projecting an image of competence. They shifted attention away from worrying about the consequences of their actions to focusing on the present moment instead. This enabled the participants to engage in more proactive learning. Their descriptions of increased concentration on the present moment with little regard for the self-preservation strongly resonate with the idea of flow. Nakamura and Csikszentmihalyi defined this psychological phenomenon as engaging in an activity with intense concentration such that nothing else seems to matter.⁵⁵ This detachment from ego into the present moment happens “because in that deep state of concentration consciousness is unusually well ordered. Thoughts, intentions, feelings, and all the senses are focused on the same goal.”^55(p41) Flow can help individuals move out of a self-conscious state, integrate the self,⁵⁵ reduce burnout,⁵⁶ and augment a sense of well-being.⁵⁷

Medical education implications

Establishing a supportive environment with longitudinal relationships free from judgments seemed especially important for the mentees. This nurturing environment appeared to foster a sense of educational safety, which may improve concentration, a detachment from self-consciousness, and the uptake of feedback.^49,58 Furthermore, the findings suggest that improved educational safety may diminish the burden to manage a competent image. This may in fact decrease the occurrence of a self-conscious imposter syndrome, “chronic feelings of self-doubt and fear of being discovered as an intellectual fraud,”⁵⁹ which has been linked to physician burnout. Overall, the positive effects of educational safety build a robust conceptualization and teaching language for PS that extends insights from the body of research that has examined PS as a consequence of mistreatments in medical school.^8–17 The findings further suggest the nature of group-level educational safety and the value of introducing peer-mentoring initiatives across the curriculum (e.g., the clinical learning environment and classroom-based learning). At an institutional level, educational safety can be fostered by decreasing the load of external assessment agendas; at an interpersonal level, by providing friendly community-building environments, mitigating the academic and social comparisons against others, and ensuring stronger supportive relationships; and at an intrapersonal level, by ensuring that the mentees’ expectations for their own learning are reasonable.

Limitations

The generalizability of our findings may be limited by our sampling methods. The rigorous process for admission to the PMME program may have selected for students who were interested in fostering mentoring relationships and therefore aware of mentor–mentee dynamics. The general population of medical students might be less sensitized to educational safety or have been unable to reflect on safe learning environments. Situating our research in the PMME context yielded rich insights into the nature of PS for medical learners. However, future studies are encouraged to explore PS in other contexts. Notably, our focus on PS may have limited our definition of educational safety. Finally, medical student abuse is unfortunately common and manifests through verbal, sexual, and physical forms.² However, only one participant described the importance of physical mistreatment as a consideration for PS. It will be important for future research to explore the interplay between psychological and physical safety as they shape students’ experiences of educational safety.

Conclusions

Within the bounds of these limitations, however, we believe that this study offers important insights into the framing of educational safety from a positive perspective. Our findings suggest that educational safety is a relational product and that when a sense of safety is sufficiently strong, learners are liberated from constantly worrying about the consequences of their actions. This enables learners to authentically and wholeheartedly concentrate on engaging with the learning task at hand without a perceived need to self-monitor their projected image. Learners’ sense of safety has a self-reinforcing nature that builds community and cycles back to a sense of group-level interpersonal safety. This emphasizes the importance of conceptualizing safety not as merely an individual or dyadic process but as a networked, interdependent product that interfaces with the larger learning environment.

Acknowledgments: We acknowledge the contributions of Dr. Michael Xu, Dr. Amy Tsai, Dr. Clarus Leung, Dr. Natalie Chan, Samantha Kennedy, Joshua Foley, Dr. James Yu, Dr. Jonathan Wong, Dr. Afrah Raza, Alex Suleiman, Ada Lo, Simran Lehal, Jessica Pettigrew, and Cameron Oliver throughout the development of PMME.