The objective was to determine the causes of and mitigating factors for conflict between emergency physicians and other colleagues during consultations.
The objective was to determine the causes of and mitigating factors for conflict between emergency physicians and other colleagues during consultations.
From March to September 2010, a total of 61 physicians (31 residents and 30 attendings from emergency medicine [EM], internal medicine, and general surgery) were interviewed about how junior learners should be taught about emergency department (ED) consultations. During these interviews, they were asked if and how conflict manifests during the ED consultation process. Two investigators reviewed the transcripts independently to generate themes related to conflict until saturation was reached. Disagreements were resolved by consensus. The trustworthiness of the analysis was ensured by generating an audit trail, which was subsequently audited by an investigator not involved with the initial analysis. This analysis was compared to previously proposed models of trust and conflict from the sociology and business literature.
All participants recalled some manifestation of conflict. There were 12 negative conflict-producing themes and 10 protective conflict-mitigating themes. When comparing these themes to a previously developed model of the domains of trust, each theme mapped to domains of the model.
Conflict affects the ED consultation process. Areas that lead to conflict are identified that map to previous models of trust and conflict. This work extends the current understanding about intradisciplinary conflict in the clinical realm. These new findings may improve the understanding of the nature of conflicts that occur and form the foundation for interventions that may decrease conflict during ED consultations.
Determinar las causas y los factores atenuantes para los conflictos entre los urgenciólogos y otros colegas durante una interconsulta.
Se entrevistaron 61 médicos (31 residentes y 30 adjuntos de Medicina de Urgencias y Emergencias, Medicina Interna y Cirugía General) de marzo a septiembre de 2010, y se les preguntó sobre cómo deberían los estudiantes ser formados sobre las interconsultas del servicio de urgencias (SU). Durante estas entrevistas, se les preguntó si existe y cómo se manifiesta el conflicto durante el proceso de interconsulta en el SU. Dos investigadores revisaron las transcripciones de forma independiente y generaron temas relacionados con el conflicto hasta que se alcanzó la saturación. Los desacuerdos se resolvieron mediante un consenso. La confianza del análisis se aseguró mediante la generación de una auditoria, que se auditó posteriormente por un investigador no incluido en el análisis inicial. Este análisis se comparó con modelos de confianza y conflicto propuestos previamente en la literatura de las áreas de negocios y de sociología.
Todos los participantes recordaron alguna manifestación de conflicto. Hubo 12 temas negativos que producían conflictos y 10 temas que atenuaban conflictos. Cuando se compararon estos temas con un modelo previamente desarrollado de dominios de confianza, se mapeó cada tema para los dominios del modelo.
El conflicto impacta en el proceso de interconsulta del SU. Las áreas que conducen al conflicto se identifican al mapear los modelos previos de confianza y conflictos. Este trabajo amplía el entendimiento actual sobre los conflictos intradisciplinarios en el área clínica. Estos nuevos hallazgos pueden mejorar el entendimiento de la naturaleza de los conflictos que ocurren y la forma de crear intervenciones que pueden disminuir los conflictos durante las interconsultas en el SU.
Approximately 40% of emergency department (ED) visits result in some sort of consultation, defined as a request for admission or further consulting service management of a patient.[1-3] Consultation encounters create scenarios where physicians must collaborate on patient care and balance competing interests. Although collaboration or the ability to work with colleagues is a prominently featured competency in both the Canadian Medical Education Directions for Specialists and the Accreditation Council for Graduate Medical Education, the real-world milieu of the ED is a challenging environment for consultation, and conflict may arise between specialists when collaborative efforts fail.[4-6]
Saltman et al. have proposed that conflict is “a disagreement within oneself or between people that causes harm or has the potential to cause harm.” Learning how to avoid or manage conflict is particularly important in the ED setting, as human factors research suggests that collaborative and harmonious transitions of care are key to patient safety and error reduction.[8, 9] The Canadian Medical Protection Agency, which is a national organization that provides legal assistance and risk-management advice to nearly all Canadian physicians, endorses conflict management as a necessary physician competency, stating that “… [l]earning how to prevent or manage conflict effectively is an important clinical skill for physicians.”
Although the body of literature surrounding consultations in emergency medicine (EM) is growing, little is known about conflict in EM consultations or intraprofessional conflict management.[3, 10-19] Much of the published literature on physicians and conflict has focused on conflicts at the physician–patient level.[6, 20-22] To date, no studies that we found have explored the effects of conflict on the ED consultation process. Our previous research found that trust is a critical part of the consultation process. In light of those findings, and other findings from the conflict-resolution literature by Lewicki and Wiethoff, we hypothesize that conflict exists in the ED consultation process and that efforts to resolve conflict in the ED consultations are affected by interpersonal trust. Lewicki and Wiethoff propose that conflict is highly tied to interpersonal trust: when trust breaks down conflict arises, and rebuilding trust has a role in conflict resolution. The purpose of this study was to describe the manifestations of conflict as recalled by emergency physicians (EPs) and consulting physicians during the ED consultation process.
This was a qualitative study that used a grounded theory approach to examine transcripts from interviews and focus groups on the topic of EM consultations. We used a deliberate sampling of three specialties (EM, general surgery, and internal medicine), including both attending and resident physicians. These three specialties were chosen because they interact most frequently within the EDs of the four study hospitals.
The line of questioning was planned a priori and conducted as part of a larger study on ED consultations. We were granted approval by the Faculty of Health Sciences/McMaster University research ethics board. All participants provided written, informed consent.
We recruited our participants from all four academic hospitals that form the core teaching centers for McMaster University. Recruitment was via e-mail using a modified Dillman approach where program administrators e-mailed their faculty and residents three times, approximately 1 week apart. We capped recruitment once the maximum of 10 participants in each of the three groups was reached. A clerical error resulted in 11 EM residents participating.
Participants were questioned via either individual interviews (attendings) or focus groups (residents). The rationale for this difference in interview techniques was logistics. It was impossible to align the schedules of the attending physicians to create focus groups.
All residents had to be in their second year of residency or higher and had to have been actively seeing patients in the ED within the past 12 months to be eligible. Attending physicians were to have actively practiced in the ED setting within 6 months of the interview, to be eligible for enrollment in our study.
The initial interview template was based on input from three local content experts who had been recommended either because of their previous lecturing (JC) or because of research (DW, CKS). The survey was piloted on a representative sample of nonparticipating faculty members and residents. The template was revised according to their suggestions. No changes were made to questions pertaining directly to the topic of conflict; however, some clarification questions were added to nonrelated questions on the survey. This piloted interview template served as the basis for a series of semistructured questions for both the interviews and the focus groups. The complete listing of our questions has previously been published. The participants were only asked two structured questions directly pertaining to conflict (What are the impediments to a good consultation-referral encounter? How do you cope or compensate for these impediments?). They were free to comment during other lines of questioning on topics related to conflict, and our analysis involved both direct answers, and other mentions of conflict in the remainder of the questions. The same questions were administered during both interviews and focus groups.
All responses were recorded regardless of whether they answered the prompting questions. A single interviewer and focus group moderator (TC) conducted all questioning of participants and used unscripted probes to explore areas that required clarification and explanations. This interviewer was a junior EM resident physician at the time and had no power over any participating physicians or surgeons, but had previously worked with a minority of them.
Interviews were planned for about 40 minutes and focus groups were planned for 60 minutes, although our protocol allowed participants to speak until the completion of their thoughts. Data Supplement S1 (available as supporting information in the online version of this paper) includes the focus group instructions. Focus groups consisted of specialty-specific groupings of residents.
The interviews and focus groups were recorded and then transcribed by the lead interviewer and two hired research assistants (medical students). All transcripts were reviewed by the lead interviewer to ensure consistency of language and syntax. The transcripts were made anonymous prior to analysis.
The transcripts of the interviews were analyzed using grounded theory principles.[27-29] Transcripts were analyzed holistically and tagged throughout. Transcripts were independently reviewed by two investigators (FB, TC) to create a single registry of codes until a saturation point was reached and no new codes were generated. At the time of analysis, one reviewer was an EM resident (TC) and one was a senior medical student (FB). Each reviewer began with an independent open coding procedure. The separate codes were compared and merged, by consensus, into a single registry of codes. The two reviewers then separately analyzed the transcripts to identify themes in an axial coding procedure to interlink discovered themes to map them to major themes. Finally, the reviewers merged their analysis, by consensus, into a thematic framework. Using a constructivist approach, the common, final framework was considered in the context of other literature to better understand the results.
A third reviewer (DO) examined and audited the analysis and provided investigator triangulation. This reviewer was an EM resident. An audit trail was established between both initial reviewers. Additionally, all notes from the analysis meetings were reviewed. Suggestions highlighted by the third reviewer were appropriately amalgamated by three investigators (FB, TC, DO). There was one single theme (of 21 original themes) that was parsed further into two separate themes.
From March to September of 2010, a total of 61 physicians (31 resident physicians and 30 attending physicians) were questioned about the ED consultation process. The participant demographics are listed in Table 1. Interviews with attending physicians and surgeons ranged from 20 to 72 minutes in duration. Focus groups with residents ranged in duration from 80 to 96 minutes.
|Level of training|
|Residents||n = 31||11 EM, 10 GS, 10 IM|
|Attendings||n = 30||10 EM, 10 GS, 10 IM|
|Mean (±SD) age (yr)|
|Both groups||36 (±9.0)||35.2 EM, 38 GS, 35.5 IM|
|Residents||29.5 (±3.3)||30.5 EM, 39 GS, 29.2 IM|
|Attendings||42.7 (±8.0)||40.2 EM, 47 GS, 41.1 IM|
|Sex||41 M/20 F||62% M, EM; 70% M, GS; 65% M, IM|
|Residents||19 M/12 F||64% M, EM; 60% M, GS; 60% M, IM|
|Attendings||21 M/9 F||60% M, EM; 80% M, GS; 70% M, IM|
We identified some manifestation of conflict in all transcripts. Ultimately there were 12 negative themes and 10 protective themes. Table 2 lists these conflict-mitigating (positive) and conflict-producing (negative) themes. Data Supplement S1 shows exemplar quotes for each of these themes.
|Factors||Conflict-mitigating Themes||Conflict-producing Themes|
|Historical factors||Good reputation||Bad reputation|
|Good prior experiences||Unknown reputation|
|Doubt in other party's competence (based onprior experience)|
|Attitudes and values displayed||Empathy||Being disengaged|
|Actions taken||Agreeing with plan of care||Poor communication|
|Collaboration||Disagreeing with plan of care|
|Meeting expectations||Self-serving behaviors|
|Provision of expert care adjusting expectations||Providing inappropriate care|
|Failing to collaborate|
|Failing to meet expectations|
The 10 positive, conflict-mitigating themes were on a variety of topics. These themes could be sorted into three major thematic groups by axial coding: 1) historical factors, 2) attitudes and values, and 3) actions. The historical factors were good prior experiences and good reputation. Each of these themes was related to how participants recalled that their prior encounters or knowledge contributed to conflict during ED consultation encounters. Attitudes and values displayed by individuals included factors such as empathy, engagement, and professional behavior. These themes all concerned personal attributes that one party outwardly displayed to the other during the encounter. Actions were a grouping of themes that described activities participants felt they needed to perform during the consultation encounter to mitigate conflict. These were “providing expert care” (e.g., stabilizing very sick patients), “agreeing to the plan of care,” “adjusting expectations about the consultation,” “meeting expectations of others,” and “collaborating.”
The 12 negative, conflict-producing themes were similarly grouped under three major themes: historical factors, attitudes and values, and actions. One other theme was identified during the analysis: external stressors. Some examples of language used to describe maladaptive behaviors and conflict-producing themes were “being avoidant,” “ducking consult,” “getting angry,” “pulling rank.”
Curiously, the theme of “external stressors” could not be sorted in to the previous major themes, but was felt to be a separate major theme. An example of an external stressor that was prominently featured was the perceived time stress on the ED. This stressor was noted as a theme in transcripts by all parties (EPs, consulting physicians, attendings, and residents). Of note, we did not note a similar mirroring conflict-mitigating theme. Participants did not recall when external forces positively affected their ED consultation processes.
In examining our conflict-producing and conflict-mitigating themes, we noted that there was a fifth underlying axial theme (trust) that intersected all the other themes. This was not surprising, as it was the theoretical and conceptual framework that informed our original constructivist approach to analysis. Conflict recollections during our interview and focus group sessions seemed to show that attributes that caused conflict could similarly be seen as a breach of trust, while other trust-building activities might mitigate or resolve conflict.
Health care has evolved over the years into a team-based endeavor, and as such, conflict may occur during high-stakes, complex interactions. Therefore, it is not surprising that we were able to determine a number of conflict-causing and conflict-mitigating factors. Five major themes (historical factors, displayed attitudes and values, actions, external stressors, and trust) affected the consultation process. This adds a layer of complexity to existing basic models such as the 5Cs and PIQUED models that have been proposed for understanding the anatomy of the ED consultation.[20, 21] Our findings suggest that effectively performing a consultation may require extensive, more nuanced awareness of social forces at play.
Our previous work had shown that trust affects the interpersonal relationships between emergency and consulting physicians. In a study of residents in emergency and consulting services, we found that familiarity and trust were key themes that affected the interactions between housestaff. These factors aligned with the theory of interpersonal trust described by Giffin. By comparing our previous work to the themes found to produce or mitigate conflict, we suggest a link between the themes in this present study and the “domains of trust,” as well as Giffin's Theory of Interpersonal Trust (see Table 3).
|Domains of Trust||Mitigates Conflict||Produces Conflict|
|Reliability||Meets expectations||Failure to meet expectations|
|Good prior experiences|
|Shared interests||Adjusting expectations||Self-serving|
|Collaborating on patient care||Lack of or failing to collaborate/help|
|Agreeing with the patient care plan||Disagreeing regarding patient care issues|
|Personal attributes and conduct (e.g., being likeable or personable)||Empathy||External stressors|
|Professional behavior||Unprofessional behavior|
|Engagement||Being engaged||Being disengaged|
|Reputation||Good reputation||Bad reputation|
|Expertise||Providing expert care to the patient||Providing inappropriate care to the patient|
|Doubt in the other person's competence|
As originally discussed, we had based our study out of the work of Lewicki and Wiethoff's model for conflict negotiation. In this model, Lewicki and Wiethoff describe that, “As relationships are based on differing levels of trust, when that trust breaks down, conflict may arise between the parties involved.”
Mayer describes five major forces that cause conflict: history, emotion, structure, values, and communication. Our findings align with certain features of Mayer's model as well. For instance, having any sort of reputation (good, bad, unknown) is consistent with having a history with the other party. An individual's behavior due to external stressors and professionalism, as well as one's ability to empathize, will shape the emotional landscape of any encounter. Expertise, being engaged and collaborative with one's colleagues, or as a corollary, acting in a self-serving manner, may be a reflection of the values of parties involved. These may potentiate or mitigate conflict scenarios. Communication, especially if done poorly as we have identified, is most certainly a factor in causing conflict as well.
Conflict management has been frequently described in the medical literature, with translational work coming from the health management and medical education literature.[7, 32-34] This translational work has focused on management applications for conflict resolution[32, 33] and providing learners with conflict resolution skills.[7, 34] Examining high-risk interactions such as consultations or other transitions of care may bring new concepts to light. Using scenarios faced by EPs every day allows for more context specificity and helps translate concepts from other fields. This may be especially imperative for medical learners, as case specificity may be important understanding.
“… conflict resolution skills employed by most physicians mirror those that were modeled by parents or significant others … Not surprisingly, the outcomes of our attempts at work-related conflict management based on these models are frequently mediocre and sometimes disastrous.”
From our transcripts and analysis, we propose some strategies that may assist ED clinicians with consultations by reducing conflict (Table 4).
|Based on the analyses from this study, there are numerous immediate applications that clinicians might utilize. We theorize that the following three behaviors or attitudes may assist clinicians in any ED consultation process.|
|Know colleagues: Our study found that being unknown was a negatively impacting factor, similar to having a bad reputation. Remaining anonymous may make discussions more difficult, since the reservoir of trust and experience is not there to mitigate potential conflict.|
|Engage colleagues: Showing interest in a colleague's work and acknowledging their impact on a patient's care can go a long way to demonstrate respect, which builds trust and decreases conflict.|
|Display shared interests to colleagues: Try to display interest for common goals and highlight shared interests in helping to stamp out disease or alleviating the suffering for that shared patient.|
As this was a qualitative study, there may be some case specificity relating to the culture of an academic center that are not generalizable to other environments, such as community hospitals. Moreover, as with all qualitative studies, we would caution readers against generalizing the results, although themes may highlight certain occurrences common to other hospitals.
Moreover, our study was limited to a single academic center for logistical reasons as this was a resident research project and we had funding and resource limitations. We did attempt to sample from four different hospitals within the academic jurisdiction, to mitigate this problem.
The focus-group moderator and interviewer for this study was a junior EM resident, and this may have influenced the openness of the participants. The involvement of medical students and EM residents in the analysis may also have altered the analysis by influencing their metaposition from the data. Also, two-thirds of our participants were male, which may affect the forms of conflict that are described and the strategies described to mitigate conflict.
Conflict occurs within the ED consultation process. Our research reveals the factors that influence the development of intraprofessional conflict. There is a tie between conflict and trust, since distrust can lead to conflict around a patient care plan. The factors that we identify that increase trust between colleagues may possibly resolve or decrease conflict during ED consultations.
The authors thank Gerson Mobo and Sarah Compeau for their assistance with the transcription process. We also thank Drs. John Crossley, Douglas Wright, and Claire Kenny-Scherber for their assistance in determining our study design and survey. We are also indebted to Drs. Michelle Lin, Nikita Joshi, and Brent Thoma for their advice. Finally, we would like to thank Dr. Lalena Yarris and the Academic Emergency Medicine journal reviewers for their coaching and thoughtful suggestions.