Rethinking the Difficult Patient: Formative Qualitative Study Using Participatory Theater to Improve Physician-Patient Communication in Rheumatology

Introduction

Background

Patient-physician interactions represent a pathway through which health differences manifest [1]. Interactions that make patients feel rushed, disrespected, or unheard reduce patient satisfaction and health outcomes [2], whereas those that demonstrate respect for patients increase their adherence to prescribed medication regimens, lead to more consistent follow-up, and improve their health outcomes [3]. Marginalized people (eg, those who identify with nondominant social identities such as race, gender, sexual orientation, gender expression, or class) are more likely to experience poor interactions [4].

Communication facilitates effective physician-patient interactions, enhances patient empowerment, improves patient understanding of health conditions, and enriches the therapeutic alliance between patients and practitioners. Effective communication has been linked to positive physical and mental health outcomes [2,5], thus making it a component of health care for all patients and is especially critical among those with chronic diseases because of the frequency of interaction between physicians and patients [6].

Patients with the chronic condition systemic lupus erythematosus (SLE) report challenges when communicating with their health care providers [7]. SLE is an autoimmune disease that follows an unpredictable course and disproportionately affects women of color [8]. Nonspecific symptoms and unpredictability of the syndrome create communication barriers with physicians [9]. In addition, medical mistrust among marginalized patients further strains physician-patient communication [10], because the health care professionals may not know how to best address it or how to recognize the social determinants of health (SDOH), which can contribute to what some have deemed contextual medical error (error in medical decision-making derived from overlooking patient context) [11]. Similarly, bias based on socioeconomic status exhibited by providers or other members of the health care system can also present challenges for patient-provider communication [12] by, for instance, relying on culturally produced stereotypes of individuals with low socioeconomic status [13,14]. These factors strain the therapeutic relationships between individuals with SLE and providers, thereby contributing to disparities in SLE outcomes [15].

Concordance reflects the degree to which patients and practitioners agree with the treatment plan [16,17]. A core component of concordance is empathy [18], which can be divided into 3 domains: cognitive, emotional, and behavioral. Although the cognitive and emotional domains are focused on understanding another person’s perspective (cognitive) and internalizing those feelings (emotional), behavioral empathy is the ability to express the (internally) experienced (cognitive and emotive) process [19]. The acquisition of these different components of empathy requires different teaching modes. For instance, instructional sessions focused on cognitive and emotional empathy may use patient narratives to teach recipients how to understand others’ situations or perspectives regarding living with a chronic disease. To improve behavioral empathy, educators might use patients who have been standardized to build physicians’ verbal and nonverbal communication skills to convey their understanding of patient experiences and viewpoints.

Interventions that address patient communication among physicians need to be tailored to the needs of specific vulnerable populations, both in the nature of disease and in the multitude of contextual factors that shape people’s abilities to engage with their health care providers [2]. Skills training for physicians often relies on information-sharing modules accompanied with observed practices with actors or patients [20]. The format of these sessions may hamper the incorporation of the patient context, the structural and social factors that influence that context, or how to address them during clinical encounters. To explore these topics, participants need a safe learning environment to ask questions about potentially sensitive topics, as well as opportunities to ask questions about the impact of their approach with the actual or standardized patient. Arts-based approaches to education are particularly well positioned to meet these needs because they facilitate active engagement among participants [21], particularly on social issues [22]. Given the relative novelty of arts-based interventions for physicians and medical trainees, a formative evaluation approach incorporating participant reflections is important for intervention refinement.

Objective

The purpose of this study was to develop, pilot, and conduct a formative evaluation of an interactive arts-based communication skills intervention for graduate-level medical trainees grounded in a narrative representative of the experience of patients with SLE. The results reported in this paper discuss the immediate impact of the intervention on participant perceptions of this mode of communication skills training and ways to improve the curriculum for future implementation.

Methods

Participants

We used a convenience sampling strategy for recruitment. All trainees in the rheumatology fellowship program at an academic medical center attended the session as a part of their required conferences (n=7). Internal medicine residents rotating through the specialty of rheumatology were also invited to attend (n=1). All studies were conducted in accordance with the relevant guidelines and regulations. Verbal informed consent was obtained from all the participants.

Conceptual Model

The intended outcome of the session was to alter physicians’ attitudes toward concordance and provide introductory skills training that might lead to better communication and concordance with patients during routine interactions. Through discussions with patients and health care providers and a review of the communication skills training literature, we identified 3 key concepts: SDOH, expression of empathy (behavioral empathy), and shared decision-making (SDM; Figure 1). Importantly, the SDOH module consisted of how providers may engage with various SDOH with their patients and group discussions on how to best facilitate connection to resources for patients when necessary.

‎

Intervention Development

We conducted a literature review to create interactive modules that align with the 3 components of the conceptual model (SDOH, empathy expression, and SDM) [23-27]. For delivery of the intervention content, we worked in partnership with the Metro Theater Company, a nonprofit educational theater organization. The research team collaborated with the members of the organization to develop a script for the initial scenario that depicted a physician-patient conflict and reflected a typical experience of living with SLE (eg, patients who were feeling that they were not being heard and physician were ignoring patient concerns). The actors playing the physician and patient had expertise in improvisation, which allowed for dynamic interaction with audience members. In addition, the research team assisted in preparing the actors for SLE-specific issues by providing relevant background information and feedback from both physicians and patients during rehearsals.

Ethics Approval

This study was approved by the Washington University Institutional Review Board (protocol #202010105).

Intervention Description

The intervention consisted of a 120-minute session with a mixture of didactic and interactive modules. Although originally designed for in-person learning, the intervention was adapted to web-based learning and delivered through videoconferencing software because of the COVID-19 restrictions. The videoconferencing environment functioned as the main “forum” of the forum theater. The primary adaptation made in the videoconferencing environment was that one member of the study team (JL) used the spotlight feature throughout the session so that only those speaking were visible to all (eg, actors during the actual scenes). All the participants were visible during the hot seating portion of the initial scenario. There were 5 phases of intervention, as described in Textbox 1.

Pedagogical Approach

We used an arts-based participatory theater approach to deliver the intervention. Participatory theater is grounded in the empowerment theory by Paulo Freire [28] and the initial Theater of the Oppressed by Augusto Boal [29]. This leverages the knowledge of disenfranchised people to transform oppressive structures by inviting members from the audience to participate in the performance and by asking them to envision ways to transform their current social realities. Audiences first engaged with performers and facilitators to represent the reality of their current experiences and then cocreate ways to modify these conditions using theater techniques as a vision to achieve change [29,30]. Actors play various parts outlined in the script in the actual application of this participatory theater approach. A facilitator engages the audience and allows for stoppage and replaying of scenes and prompts members from the audience to provide suggestions on how to redo interactions for different outcomes. Audience members are then asked to tap into the scene and take on the role of the characters (replacing the actors). This differs from standardized patients in medical education in that the trainees typically only play the role of health care provider and do not necessarily have in-built facilitation or structured observation [31].

Measurements

Participants shared feedback through 2 focus groups facilitated by members of the research team during the last 20 minutes of the session. Facilitators asked participants about the strengths and weaknesses of the intervention’s design and implementation as well as suggestions for improvements. Focus group questions were provided to the participants by the authors.

Analysis

We conducted a qualitative analysis using a constructivist paradigm. The entire session was audio recorded. Focus groups and discussions for each phase were transcribed verbatim and analyzed using focused coding techniques [32], in which multiple members of the research team independently assigned codes to transcripts based on the interview topics, discussed disagreements, and arrived at a consensus on the best way to address discrepancies. After the initial coding, codes were arranged into clusters with summary paragraphs describing key elements of each cluster, along with supporting quotations from the focus groups. The focus groups were audio recorded and transcribed. To maintain confidentiality, audio files and corresponding transcripts were stored on a password-protected computer. Complete anonymity could not be maintained with the audio recordings because voices could potentially be identified; however, the participants did not state their names during the focus groups. In addition, unique identifiers were generated to label the speakers in the transcript documents. Participants were not compensated for their participation.

Results

Overview

A total of 8 trainees participated in the workshop (Table 1). All the 8 participants also participated in the postsession focus groups (4 participants per group) and the demographic questionnaire. The participants’ ages ranged from 26 to 40 years, with the majority (6/8, 75%) being aged between 31 and 35 years, women (5/8, 63%), and non-Hispanic or Latinx (7/8, 87%).

Focus Groups

Overview

We grouped participant feedback into 3 categories as follows: value added by method of content delivery, suggestions for improvements in the method of delivery, and overall reflections.

Value Added Through Method of Content Delivery

Participants generally thought that the opportunity to deeply investigate the interaction between the patient and provider, rather than just observing it, added value to their learning experiences. They found value in being able to ask questions of both the patient and provider after the videotaped interaction, something they had not been able to do either in prior education settings or in actual interactions with patients (Table 2; quotation [Q] 1).

The ability to ask questions to both the patient and provider and have each respond while maintaining their role functioned as a type of confessional interview that allowed the respondents to hear the internal narrative of both the patient and provider. This was an important experience for providers who often walked out of a room and asked themselves what the patient took from the interaction but did not have the opportunity to ask the patient (Table 2; Q2).

In comparing the forum theater approach to their previous standardized patient experiences, participants indicated that the forum theater was better because they were able to test alternative strategies for changing the interaction and see how a patient might respond. For instance, the format of letting participants play either the provider or patient role while an actor played the other role was seen as more believable than previous experiences with standardized patients (Table 2; Q3). Participants noted that in prior experiences, they would practice certain scenarios but did not get to see someone role modeling the behaviors (Table 2; Q4). They would then be corrected by the instructor, but the standardized patient typically did not provide feedback or insight into their experience of the provider’s behavior (Table 2; Q5).

The desire to test strategies was further evidenced by the types of questions asked by the physician and patient actors. Questions for the physician actor centered on wanting to know more about why the physician did not ask about the patient being late and whether they thought the interaction would have gone differently had the patient been on time (Table 2; Q6). Participants went further by asking the patients how they would feel if they were asked to reschedule when they arrived at their appointment late (Table 2; Q7). The patient-actor response suggested that they would likely leave and be unlikely to reschedule unless the provider showed that they had taken the patient seriously (Table 2; Q8).

Participants appreciated when visual cues were built into the activities because they helped reinforce the primary message of that activity. For instance, the SDOH board game contains red chips, which are meant to represent an acquisition (or lack of) advantage in society. Participants commented that these chips were helpful in demonstrating the trends and accumulation of both health-promoting and deleterious impacts of SDOH (Table 2; Q9). The red chips provided an opportunity for critical reflection on the forces that led to the accumulation of certain resources that affected patients’ ability to manage their disease (Table 2; Q10).

An additional element of activities that participants found helpful in the training was the presentation of a variety of different tools for the same concept, as well as specific phrasing or verbiage suggestions. The use of a variety of tools enabled participants to begin making connections on how these tools could be used in their practice and future interactions with patients. (Table 2; Q11-12). Participants began to see that there is not just one way to address the issues, but a range of options they can choose from (Table 2; Q13).

Participants praised the interactivity of some parts of the session, particularly the opportunity to interact apply some of the strategies they had practiced throughout the session with the actors (Table 2; Q14).

Suggestion for Improvements in Method of Delivery

Although participants were enthusiastic about the training, they also noted that some parts of the session were not quite engaging. For instance, they mentioned that the presentation of some of the didactic components could have used more active ways to engage with written material (Table 2; Q15-17).

Participants also suggested ensuring multiple ways of engaging with information for different activities (ie, both visual and auditory elements). Having multiple methods of engagement would help the participants contribute to a deeper level of interaction during the discussion sections of the session (Table 2; Q18).

Suggestions were also made to make some of the activities more fluid. For instance, during role-playing exercises, participants were not sure what to say when role-playing as a patient, which could make the role-playing feel forced, especially because participants were not used to acting. Without cues or guidance, it was sometimes difficult for participants to tell whether they were fulfilling the goals of the session (Table 2; Q19). One suggestion was to provide actor notes that would guide how participants should try to act (Table 2; Q20).

Overall Reflections

In their reflections on the session, participants commented on how the 3 different pieces of the session (SDOH, expressing empathy, and SDM) complemented each other in ways that they had not previously considered (Table 2; Q21).

Participants noted that they were familiar with some workshop concepts. However, it was evident from their reflections that this session went further by building on the foundation that participants may have already had and provided additional context for these terms, such as SDOH, and how having this lens impacts their empathy with patients in the SDM process. The participants noted that this was particularly important given the impact SDOH has on the ability of patients with rheumatological diseases to manage their diseases (Table 2; Q22).

Specifically, participants saw the tools and activities as helpful by providing insight into what the patient and physician were thinking about the same topic. The activities appeared helpful in contextualizing physician-patient interaction and ensuring that physicians see patients as people who have medical problems and not just medical problems associated with the person. For example, physicians tend to focus on getting patients to take their prescribed medications without exploring why patients are not taking their medication or adhering to care recommendations (Table 2; Q23-25).

This reframing then encouraged participants to also think about their role as health care providers in taking action to address some of the factors outside of the medical condition, which may affect lupus (Table 2; Q26).

This synthesis was further evident in what the participants took away from the session. For instance, they considered framing the beginning of the interaction differently by spending more time with introductions and establishing a rapport with the patient. This underscores that one of the most important parts of the interaction is ensuring that the patient feels that their physician is listening to their concerns and connecting actions directly to those concerns (Table 2; Q27-28).

Participants also connected this rapport building to later points in decision-making with patients. Demonstrating that the physician provided early attention to patient concerns was helpful when delivering information that the patient was not completely agreeable to. Participants noted the importance of having a backup option for a treatment decision and that they would continue working toward an agreeable solution if the initial treatment recommendation was not working for the patient (Table 2; Q29).

Participants similarly saw rapport building as generally leading to more productive discussions in the context of patient life stressors and other SDOH. Although the participants noted that direct solutions to other life stressors cannot always be provided within the context of a clinical visit, attention to these life stressors by physicians is still meaningful if the patient feels heard and strategies are discussed to help the patient and provider get on the same page early (Table 2; Q30). Participants directly connected this early rapport building and acknowledgment of SDOH to being able to better emphasize important elements of lupus management such as the importance of consistently taking medications (Table 2; Q31).

Participants noted some obstacles to implementing the tools presented during the session given the reality of working in a clinical setting. For instance, some participants voiced concerns that asking how patients are doing may open the conversation to issues beyond what they can address. There was a concern that some of these strategies would invite patients to focus on self-catastrophizing behaviors or that some of the strategies would take more time than allotted during a clinical encounter (Table 2; Q32 and 33).

Others have noted that suggestions for interacting with patients might not be realistically implemented in a single patient encounter. Often, a provider may not be able to go through all steps in decision-making in one visit, so it can be challenging to cover each of the SDM processes and examine the patient in a time-limited clinical visit (Table 2; Q34 and 35).

Discussion

Principal Findings

We presented formative evaluation data from an interactive, arts-based communication skills workshop for rheumatology trainees. Our qualitative data suggest that participants responded positively to the level of interactivity of some session components, the overall cohesiveness of the separate phases, and the unifying approach of theater-based performance. The design of the session enabled participants to integrate the different components of the sessions, relative to the previous types of workshop training (often compared with standardized patients).

The novelty of this session was the integration of structural and social factors as a context for patient encounters. This was achieved through the intentional sequence of theory-grounded activities built upon each other (Figure 1) and facilitation, which guided participants in linking activities together. This approach allowed participants to move from simply receiving information about communication behaviors to applying that knowledge in a dynamic environment that incorporated patient lived experience. This was evidenced by participants in their generation of strategies: participants discussed the importance of asking how patients were doing at the beginning of the encounter (SDOH activity or empathy expression), which they saw as helping with rapport building and ultimately facilitating the space for cooperative SDM to occur. Although discussions on structural factors and SDOH have permeated medical education and are increasingly large parts of the curricula [33], the integration of these SDOH in communication-specific modules is less common [2]. Our findings suggest that this integration is critical in the uptake of information and skills in both SDOH and communication domains.

The incorporation of the structural context of health inequities was greatly facilitated by the participatory theater approach. The unifying scenario, presented at the beginning and developed in conjunction with an educational theater group, was an important grounding for participants, as they interacted with actors throughout the session. This unifying scenario was received positively by participants, who affirmed that the application of arts-based pedagogy was suitable for graduate medical education. Our findings align with the application of forum theater in other medical settings including general medical education [22] and specific discussions around social factors such as race [34]. In line with these studies, we observed the advantage of forum theater in the exploration and generation of ideas among participants rather than prescribing a particular solution. This was evident in our findings, as participants praised the dynamic interactivity of the theater components, which appeared to be more engaging than their previous experiences in medical education settings with standardized patients.

Limitations

This study has several limitations. First, we performed only a single session with 8 rheumatology trainees. Although we would have ideally conducted additional sessions, our intention was to pilot materials that were specifically designed for a rheumatology setting to assess what activities did or did not work well. Moreover, within this narrow population, we were able to include all rheumatology trainees at the study institution. An additional limitation is that we relied on a single script in our study. This may limit the scope of information received by participants; therefore, future iterations of this type of intervention may consider developing multiple scripts for the same topical area. Future iterations of this session may also consider designing multiple sessions to track changes in attitudes and integrate evaluation metrics with participant observations in actual clinical settings. In addition, the intervention was conducted via videoconferencing because of the COVID-19 pandemic restrictions, but the intervention may have been received differently had it been in person. Finally, focus group methods were limited (eg, not able to engage in member checking) because of funding, but future versions of this work may consider additional recruitment strategies for more rheumatology trainees (ie, multisite intervention) to obtain a larger sample to better gauge data saturation.

Conclusions

Our findings from this pilot test of communication modules suggest that participatory theater is an effective method for framing physician education through a health equity lens. Feedback through qualitative focus groups revealed several considerations for the further exploration of this approach and framing. First, a critical insight from participant feedback was the need to acknowledge the workload of the physicians. Participants raised concerns about the reality of working in fast-paced clinical environments and the difficulty of implementing some of the tools. This aligns with the current understanding of the increasing physician loads, more time spent on administrative tasks or electronic medical record [35,36], and pressure to see high volumes of patients while also providing optimal care management [37]. The prospect of adding additional tasks to an encounter, such as spending time discussing social factors with patients, is daunting. This suggests that education sessions such as the one presented here need to balance the realities of clinical practice with the new knowledge from these sessions. These insights also suggest that complementary sessions designed to educate patients about the demands of physicians are necessary, using similar tools and approaches as those described here. This suggestion is in line with work that has pointed to the disproportionate emphasis of interventions on physician behavior, but which misses an important half of the physician-patient dyad [38].

In addition, our intervention incorporates elements of “structurally competency,” which Metzl and Hansen [39] define as the “ability to discern how a host of issues defined clinically as symptoms, attitudes, or diseases also represent downstream implications of a number of upstream decisions about such matters as health care and food delivery systems, zoning laws...” Structurally competent approaches to medical education prioritize training medical professionals to recognize the ways in which structural (upstream) factors shape the medical encounter and patient outcomes (downstream). The evidence of the basis of this work, supported by our formative findings here, suggests that these structurally competent approaches are effective at integrating SDOH education. Moreover, it suggests that this approach can also facilitate concordance between physicians and patients, a bedrock for a productive therapeutic alliance, and improvement of health outcomes.