“Patients Are the People Who Teach Me the Most”: Exploring the Development of Communication Skills During Internal Medicine Residency

The importance of physician-patient communication is widely accepted, and the development of expert patient-centered communication skills is expected of all medical trainees.^1,2  High-quality physician-patient communication is associated with positive health outcomes such as lowering blood pressure, reducing anxiety, reducing morbidity, and improving quality of life.^3-5  Communication training is a burgeoning field of research, yet medical professionals still often lack the mastery of communication that is expected of them.^6-9 

Existing evidence supports the use of a variety of communication training strategies. For example, the use of standardized patients (SPs) is ubiquitous, due to their capacity to teach communication techniques, the opportunity for direct feedback, and widespread acceptance of efficacy.^6,10-14  Video recording patient interactions for the purpose of asynchronous feedback is gaining popularity, although primarily in outpatient settings.^7,13-17  Numerous other strategies, such as role modeling, communication checklists, and limited communication courses, have also demonstrated effectiveness.^{7,14,15,18-21}  However, despite robust evidence to support these modalities, the concern remains that expertise in communication is not achieved and that deficiencies exist in the communication training of medical learners.^6-9,22  Additionally, evidence suggests that learners' patient-centeredness may decrease during clinical training.^18,23-26  Several explanations have been proposed to explain why seemingly sufficient communication training falls short, including few opportunities for observation and feedback,^19,22,27  a concern that non-expert instructors limit learners,^7,27,28  and little opportunity for deliberate practice.^6,7,29 

One approach to untangling these discrepancies is to explore learners' perspectives on why communication training may be inadequate. Literature on this topic is scant and focuses predominantly on outpatient settings. Key findings are that learners report that their interactions with patients are rarely observed, that feedback is often not instructive or focused on communication skills, and that skills developed through SP training do not translate to real clinical settings.^16,27,30  Role modeling of communication by senior physicians is highlighted as a significant source of development by learners, but it is also frequently cited as a hindrance, due to modeling by non-expert communicators which can lead to the adoption of poor communication techniques.^27,31,32  There are a few studies that explore family medicine residents' perspectives on communication skills, but they are limited to ambulatory care.^33,34  This gap is important to address, as specialties where residents primarily train in inpatient settings, such as internal medicine (IM), surgery, anesthesia, and emergency medicine, receive less communication training than their outpatient colleagues and prioritize efficiency over communication.³⁵ 

To address the existing gap between the importance of and emphasis placed on communication training and the imperfect communication skills of physicians, we sought to better understand how IM residents develop the communication skills that they use in their everyday clinical encounters, and to elucidate their views on existing communication training modalities.

A constructivist grounded theory approach was used for data collection, analysis, and interpretation, given its ability to explore social phenomena for which we require a deeper understanding and new theoretical insights.³⁶  In constructivist research, the emphasis is on how individuals perceive and create meaning from their experiences.^37,38  As researchers this means knowledge and meaning are co-constructed between ourselves and our participants.

Recruitment emails were sent to all 225 postgraduate year (PGY)-1-3 residents at the University of Toronto IM program in November and December 2020. Participants were selected to achieve diversity with respect to level of training and gender.³⁹  Fifteen volunteers were chosen: 5 junior residents (5 PGY-1s) and 10 senior residents (8 PGY-2s and 2 PGY-3s). Participants provided informed consent prior to participation and were given a $15 gift card.

Between November 2020 and January 2021, one-on-one semi-structured interviews were conducted virtually via Zoom. Interviews ranged in length from 27 to 63 minutes and averaged more than 45 minutes. Participants were asked a series of questions to explore how IM residents develop communication skills, the role of formal communication training, and the scope of communication development that occurs during IM residency (see online supplementary data for the interview guide). Consistent with constructivist grounded theory, probing questions asked in each interview varied depending on the responses of interviewees and were intended to obtain rich, reflective responses.⁴⁰  Audio recordings were transcribed by a third party service. Transcripts were anonymized by G.B. prior to being made available to the research team for analysis.

Analysis and data collection occurred iteratively to allow for identified themes to be better explored in later interviews.⁴¹  One author (G.B.) read each interview transcript and independently created open codes using NVivo version 12 (QSR International), which were reviewed with the research team throughout the data collection process in order to refine the codes as data collection progressed. Constant comparison was used to refine coding as transcripts were added to the data set. Axial coding was used to group codes into meaningful categories and explore the relationships between them. The research team repeatedly revisited the data to construct meaningful themes that were shaped into an evolving framework. Once we reached 15 interviews, we determined that we had enough sufficiently rich data to understand and describe the phenomenon under study. We then constructed a model that represents the core themes and their interrelationships.

A PGY-2 IM resident at the time of data collection, G.B. conducted all interviews given his shared understanding of residents' experience and because his insider status was anticipated to facilitate more honest responses. S.G. is a senior attending physician and education researcher, who observes and assesses residents' communications skills on the inpatient teaching units. L.M. is an early career attending physician and researcher with an interest in workplace-based assessment. At all stages of the study, we openly discussed and reflected on our own experiences in communication skills training and assessment, being careful to ensure that we did not overinterpret the data.

Ethics approval was received from the University of Toronto Research Ethics Board.

We identified 5 themes that represent how IM residents develop communication skills: the modalities and value of formal training, the role of staff physicians, the limitations of directly observed settings, the value of working independently with patients, and the evolution that occurs during residency.

Regardless of where they attended medical school, participants described communication training deviating little from a common formula consisting of didactic lectures accompanied by practical experience with SPs. Emphasis was placed on patient-centeredness via strategies such as asking open-ended questions, affording patients time to digest information, confirming patient understanding, and thoroughly answering patients' questions. Most residents attributed a degree of comfort and some foundational communication skills to this training.

Some pointed out that not all strategies taught in medical school were translatable to IM settings. One resident (R6) brought up the SPIKES methodology of breaking bad news (an acronym for Setting, Perception, Invitation, Knowledge, Empathy, and Summary), and recalled how effective it was in objective structured clinical examinations (OSCEs). However, they noted that “in the real world” they “don't think with SPIKES” because they lack the requisite privacy and time; instead, they rely on more basic techniques such as active listening and making eye contact.

Additionally, participants felt that learning with SPs had limitations. Some took issue with the scripted nature of SP responses, which caused them to alter their interviewing style. One resident explained that it “just wasn't natural” because in real life “you would have an open-ended conversation” and “a real patient would volunteer” information. (R13)

But for an SP, open-ended questioning would soon transition to very closed-ended. “You need to ask a specific question, and if you don't, I am not going to tell you the answer you want.” (R13)

Other participants expressed a lack of trust in feedback they received from SPs, as explained by R4: “SPs have taken liberties to say how I should do something because ‘this is how the majority of patients would feel,' and there's no way to prove which one of us is right.”

Formal communication training during residency was felt to be limited. Participants reported having a small number of academic half days on communicating with certain populations or conducting goals of care discussions, but little formal training beyond these lectures.

Staff physicians were felt to contribute significantly to communication skills development. However, participants ascribed this contribution primarily to observing staff communicate, rather than to receiving feedback or instruction. One resident observed that some attendings communicate well and some “don't communicate so well,” so they “take little bits from both” and “try to incorporate that into the way that I communicate.” They went on to reflect: “I wouldn't say I've ever really gotten any feedback that's drastically changed the way in which I communicate. I think it's more from observing attendings.” (R10)

This lack of impact of feedback was partially attributed to a sense that most feedback on resident communication is generic, positive, and lacks actionable features, which frustrated some participants. “So it gets a bit tiring at times when people are like, ‘Yeah, you're great, keep doing what you're doing,' and you're like, ‘Okay, but give me more than that. What can I do next time? How can I change things? How can I make it better?' And I very rarely get that side of it.” (R5)

Other participants noted rare occasions when feedback was specific and constructive, for example, being told that “I use my hands too much when I talk” (R12) or to be wary of “upspeaking when having a goals of care discussion” (R11), which were thought to be helpful.

One major hindrance to feedback is the limitations of directly observed settings. First, participants noted that observed interactions were infrequent: “Sometimes in the morning, as we're rounding on the new patients, ‘Okay, [resident], just explain the plan to [the patient],' and so in front of the team, I'll explain the plan for the day. And that doesn't happen all that often. The other times that staff observe me communicating would be in family meetings. Again, doesn't happen all that often.” (R12)

Critically, almost all participants described altering their communication style when observed, to appear more professional or to avoid saying something that might be disapproved of by an observer. These stylistic modifications were described in different ways, including “I definitely get more into ‘presentation mode'” (R12), “I might try to mimic [the staff's] style” (R2), and “it's [like] driving your car on a normal day versus driving with a driving instructor” (R3). This phenomenon of communicating in a manner that was less authentic was a recurrent theme: “I'm more deliberate with what I say, and I am more cognizant about what I'm saying and how I'm saying it because I know there's somebody else in the room. So I'm just more mindful when I'm speaking to patients in the presence of staff…I'm more careful.” (R7)

These changes were not just about participants putting their best foot forward. Some felt that being observed forced them to withhold aspects of their natural communication style that they considered strengths, such as using a joke to build rapport with a patient; this in turn prevented them from receiving feedback on their true communication style. This is described by R1, who noted that they develop rapport with patients by taking a relaxed, collegial approach, and try to make patients laugh, which “is what gets [patients] to trust me very quickly.” Yet they “probably would not employ those kinds of communications strategies if I was being evaluated or observed, despite how effective I find them.” The deliberate suppression of apparently effective communication strategies while being observed robs R1 of the opportunity to receive feedback that may help her fine tune this approach.

Participants consistently expressed that patients were the best resource for communication skills development: “How I learn is from actually having conversations with real patients… patients are the people who teach me the most.” (R5)

Patient interactions were felt to be so valuable because of the inherent authenticity of a patient's reactions, “When a patient tells you something then you know that that's actually how you made them feel” (R4), and the informal feedback they provide, “you pick up a lot of feedback from the patients.” (R14)

Every participant expressed that communicating directly and independently with patients played a significant, if not the largest, role in the development of their communication skills. As R6 explained, “Twenty percent of my communication skills is what I see staff do. Eighty percent is the trial and error of interacting with patients.”

Emphasis was placed on the self-directed nature of this development, with phrases such as “It was more just self-reflection” (R10), “We never really have people watching us talk to patients” (R15), and “Experience is your biggest teacher” (R3), being commonplace when participants discussed their communication skills. This highlights that IM residents believe their communication practices develop independently, with infrequent supervision or guidance: “I'd say it's mostly, if not all, from my own experience and learning myself and from my interactions, because I don't think that we really get much teaching or coaching on how we're speaking with our patients.” (R8)

During patient interactions, IM residents described self-monitoring and adapting their communication behaviors based on their interpretation of the patient's reactions: “We can pick up non-verbal cues from people, like how comfortable they feel when you're speaking with them or if they're getting agitated…That kind of makes you take a step back and maybe go slower, maybe change your tone, focus on a different topic.” (R14)

While real-time adjustment in response to patient behavior is an important and taught component of communication, this process relied on IM residents' ability to accurately interpret patients' reactions and adjust their communication, which some residents noted could be subject to error: “I think there is definitely a bit of guesswork involved, but again, that comes down to the complexity of human interaction and communication.” (R9)

Many participants commented that during training they expected improvement in their skills and comfort with specific communication tasks, such as conducting goals of care discussions. However, residents' development spanned a broad spectrum. Some became more patient-centered: “My place is to give [patients] the information. Make sure they have all the right facts to come to decisions that are right for them. That is my role here.” (R5)

Yet others gravitated toward more paternalistic approaches: “I'm more prescriptive...When it comes to things like code discussions, I'm going to tell you what I think is best and I don't let the family or the person run the conversation.” (R1)

A recurrent finding was that residents believed they needed to become faster with their communication. They described a myriad of adaptations to improve efficiency, such as “I am, in general, more succinct” (R11), “You just have to cut people off sometimes” (R12), and “Not tiptoeing around things as much” (R5). Most participants described brevity being necessitated by their workload: “On GIM, sometimes I'm carrying 8 [patients], I'm carrying the pager, I'm co-signing for med students, I can't spend 10 minutes with the patient.” (R6)

Further stressing efficiency, many participants perceived that productivity was prioritized by their attendings over patient-centeredness: No one cares about your therapeutic relationship, unfortunately. They care about, “Did you see your patients and how are you managing their problems?” (R13)

While some participants felt they could balance efficiency with effective communication, others described a potential tradeoff of, and tension between, efficiency and patient-centeredness: “I don't want to say sacrificing compassion, because I try not to do that, but sacrificing, maybe, I guess time...I think that's good from an efficiency and workflow standpoint. I don't think it's gotten to the point where…I haven't perceived myself as being rude to patients or losing that compassion, but yeah, I'd like to think it's more efficient. Maybe with a little bit less empathy. Well no, actually, not empathy. No, I take that back. I definitely still try to be empathetic, but…yeah, I don't know.” (R9)

These findings suggest that in the inpatient setting, residents' development of communication skills primarily occurs through self-reflection on unsupervised patient interactions, with smaller but significant contributions coming from modeling of communication by staff physicians and formal communication training. Relatively little was attributed to staff observation and direct feedback. We constructed a model to visually represent these findings by depicting each learning modality in different sizes proportional to their contribution to communication skills development (see the Figure).

Participants in our study entered residency equipped with communication skills they were taught in medical school, which were aimed at making them patient-centered communicators.^1,2  However, many felt that the tools they were taught, and that were emphasized on practical examinations, did not translate well to real inpatient settings. Modeling by staff played a role in communication skills development, but our participants described the majority of their development occurring through self-reflection on unsupervised patient interactions. As a result, residents developed personalized communication styles and strategies without direct guidance from supervising physicians, impacted by prioritizing efficiency over patient-centeredness due to time constraints on busy services. When observed, residents altered their communication styles to appear more professional, and withheld behaviors they feared attending physicians would not approve of, even if those behaviors had been successful for them in the past. This change in style, coupled with a perceived paucity of observation, reduced the impact of observation and feedback on their authentic communication skills.

Teaching residents to communicate with patients is of paramount importance to medical education and is guided by the CanMEDS and Accreditation Council for Graduate Medical Education frameworks, which outline competencies in communication required of all physicians.^1,2  Existing literature describes the important role of self-reflection and patient feedback in communication skills development but mainly through observation and collaboration with attendings.^31,33,34  Our findings suggest that residents in inpatient settings may develop their communication skills more independently, without guidance from attendings or residency programs. In the absence of appropriate guidance residents may be developing communication habits that they find anecdotally effective but lack patient-centeredness and do not align with educational frameworks. Some research has reported that communication skills remain stagnant during residency, patient-centeredness decreases during training, and staff physicians and residents lack true communication expertise.^6,8,18,24-26  Given our findings, these problems may be partially explained by 2 issues: current training modalities falling short and residents subsequently filling the gap by teaching themselves.

The phenomenon of self-directed communication development is likely due, in part, to the well described limitations of direct observation and feedback.⁴²  Residents' proclivity to alter their behavior when observed renders feedback they receive on these interactions less relevant to their authentic communication style. These changes in behavior also affect assessment, which relies primarily on observation.^6,14,17,43  In essence, participants display what are perceived to be “optimal” behaviors while being assessed, that they do not normally use when unsupervised.⁴²  These findings beg the question: What role does observation play in teaching and evaluation if observed interactions predominantly demonstrate what learners believe is expected of them and not their authentic communication practices?

Consistent with other studies, our findings demonstrate that the discrepancy between how medical learners are taught to communicate with patients and how residents in inpatient settings actually communicate with patients may be partially explained by busy inpatient services leading residents to prioritize efficiency over patient-centered communication.³⁵  Lack of sufficient time is a barrier to effectively breaking bad news,⁴⁴  prevents residents from trying new communication techniques,¹⁶  limits thoroughness of communication at time of discharge,⁴⁵  and prevents practicing and maintaining taught communication strategies.^7,27  The perceived need for efficiency to the detriment of patient-centered communication may contribute to the well documented potential decrease in residents' patient-centeredness that occurs over the course of training.^18,24-26  This discrepancy between what is learned vs what is enacted is an example of the influence of the hidden curriculum,⁴⁶  and is supported by our finding that participants believe communication techniques they learn in classroom teaching environments, such as with SPs or in OSCEs, often do not translate well to real inpatient encounters. If we are teaching IM residents to be patient-centered via communication strategies that they do not have time to implement on the wards, or which cannot be used in a loud emergency department or ward rooms devoid of privacy, perhaps we are setting them up for failure. Our current communication training may inadvertently prepare learners to perform well on OSCEs but not for the realities of inpatient practice.

Our findings suggest possible new ways forward for communication training. For example, curricula could be developed that purposefully incorporate time constraints or distractions as a way to better translate formal communication training to inpatient settings where these limitations are a reality. Additionally, given residents' predilection to learn from patients and the large volume of unobserved interactions on inpatient services, perhaps feedback from patients can be incorporated into residency education.⁴⁷ 

One limitation of our study is that participants were all residents at one academic institution and from one inpatient specialty, although they came from multiple medical schools. It is possible that IM residents at other institutions or from different specialties have experiences that differ from those reported here, due to variations in clinical practices and different residency communication curricula. Additionally, although we purposefully selected residents to achieve diversity, all were volunteers and may be more interested in communication skills than non-participants. Finally, this study was conducted during the COVID-19 pandemic and the PGY-1 participants would only have had a few months of residency training, all done during the pandemic, at the time of their interviews.

Our findings suggest that IM residents believe they predominantly develop communication skills independently, from unsupervised patient interactions, without much guidance from their attendings. When they are observed, many residents alter their behaviors to reflect communication styles they believe will be judged favorably. The perceived need for efficiency sometimes results in less patient-centered communication that is not in keeping with communication techniques taught to medical learners.