Abstract
Background
In situ simulation holds promise to enhance the effectiveness of performance. However, a dearth of
Aim
The aim of this study was to examine
Method
Twenty-four health care professionals within an
Results
Participants reported significant
Conclusion
Our results suggest that
Keywords
Patient care is a team sport (Salas et al., 2008). Caring for patients is an inherently multidisciplinary task that involves the interaction of professionals from highly diverse backgrounds regarding discipline, training, and experience. For example, a trauma team may consist of several nurses, allied health professionals, a surgeon, an emergency physician, a respiratory therapist, and several specialized physicians. Although these individuals have received extensive training in their discipline, they often have not received formal training concerning how to interact with one another. Furthermore, caring for patients in trauma situations is often characterized by high-stakes outcomes, elevated stress, and intense time constraints. Thus, identifying factors that influence team cohesion and performance in these settings is essential to better understand how teams impact patient care. The current study borrows socio-cognitive theory from psychology to help explain how medical simulation training might be able to enhance team performance. Specifically, two trauma scenarios were created to investigate how simulation impacts the development of transactive memory systems and how such systems relate to observer-rated team performance.
Transactive Memory Systems
As noted by others (e.g., Crookall, 2011; Dieckmann, Gaba, & Rall, 2007; Gardner & Rich, in press; Issenberg, Ringsted, Ostergaard, & Dieckmann, 2011; King, Holder, & Ahmed, 2013; Ostergaard, Dieckmann, & Lippert, 2011), the application of theories outside of medical domains is a necessary next step to expand the knowledge base of simulation-based education. In response to this call for research, the current study examines how transactive memory systems (TMS) may impact team performance on trauma-based medical simulations.
A TMS is a combination of the knowledge possessed by each individual and a collective awareness of where knowledge resides within the team (Wegner, 1987). However, TMSs describe more than just understanding others’ roles and responsibilities. Teams with developed TMSs are aware of others’ specialized knowledge in such a way that they can coordinate actions according to members’ unevenly distributed knowledge and task structure. According to TMS theory, members of a team divide the cognitive labor for their tasks, with members specializing in distinct domains. Group members then rely on one another to be responsible for specific expertise such that collectively they possess all of the information needed for their tasks. This shared dependence frees members to develop more specialized expertise in their respective areas, while also permitting access to others’ task-relevant information.
Teams with developed TMSs are able to efficiently process information, plan actions, and anticipate and respond to the actions/needs of their teammates (Kozlowski & Ilgen, 2006). For example, in a trauma resuscitation with a compromised airway, airway management duties need to be delegated to the member with the most experience. Simultaneously, other critical life-saving tasks need to be performed. The team may need to perform tasks such as a needle decompression, intubation, chest tubes, venous access, and medication administration in an appropriate and timely fashion. Teams need to be able to dynamically assign priorities and manage different aspects of the resuscitation while working toward the collective goal of patient resuscitation. Thus, delegating responsibilities to those with appropriate expertise and trusting that team members are adequately performing their tasks allows individual members to perform their own respective duties successfully. In short, TMSs deepen the development of expertise, reduce cognitive load, and enhance performance among teams (Wegner, 1987).
In order to successfully integrate their distributed skills and expertise, however, teams must first interact with one another to develop accurate cognitive structures (Pearsall, Ellis, & Bell, 2010). Indeed, shared knowledge does not develop automatically; it instead requires team members to engage in purposeful interactions (Lewis, Lange, & Gillis, 2005). Cognitive interdependence, the perception that personal outcomes are dependent on the knowledge of others and that others’ outcomes are dependent upon personal knowledge, is a precondition for the development of transactive memory and has been posited to be instilled in teams through simulations (Manthous, Nembhard, & Hollingshead, 2011). Simulation-based trauma scenarios and debriefing are well-suited to facilitate such processes. By working together on a patient scenario, group-trained members are able to acquire an understanding of member-specific expertise and weaknesses, delegate responsibilities accordingly, and focus on performing personal tasks at an optimal level. Furthermore, when additional support is needed, team members know from whom to seek assistance. In this way, simulation offers opportunities for teams to gain the necessary competencies to optimally leverage the distributed expertise of team members.
In addition, debriefing sessions conducted after a hands-on simulation can help establish and maintain these transactive memory systems. During debriefing, participants are encouraged to reflect upon the scenario regarding personal performance, the performance of team members, and the overall management of the simulated scenario. This facilitated discussion helps to illuminate important insights from the learner’s performance and helps to connect the instruction and guidance from the simulated exercise to future actions in the clinical environment. As debriefing is coined as the “heart and soul” of the simulation experience (Rall, Manser, & Howard, 2000), we believe that it is also an integral part of the formation of transactive memory systems. It is during this time that participants are able to critically analyze their performance and develop integral metacognitive structures (Ostergaard et al., 2011).
The current study used in situ simulation (simulations performed in the actual clinical environment) to achieve two main aims. First, we investigated whether team-based simulation promotes the development of transactive memory systems among team members. Second, we sought to examine how theoretical concepts and empirical findings concerning transactive memory in nonmedical domains apply to simulation-based health care settings. Specifically, we assessed the extent to which transactive memory relates to team performance during two trauma simulations.
Materials and Method
Design
Questionnaires were distributed for completion before and after the simulation. No identifying information was collected from the study participants, though pre- and post-scenario questionnaires were linked by unique identifiers to allow comparison and performance evaluation. The Institutional Review Board (IRB) classified this study as Quality Improvement rather than primary research and, therefore, informed consent was not applicable to the exempt project.
Setting and Population
In situ training was chosen because of its many advantages (Ostergaard et al., 2011). Specifically, in situ allows participants to train in operational settings with equipment frequently used and also allows access to a variety of ED roles. All simulation participants were employees or contracted providers of the institution and had roles in patient care or diagnostic testing in the ED, representing a variety of disciplines and experience levels that routinely function in this setting. Simulation participants were invited by their department leadership to voluntarily participate in the mock trauma sessions.
Study Protocol
Study participants were approached to complete a pre-survey prior to the start of the simulation sessions. The pre- and post-surveys included 15 items assessing transactive memory (Lewis, 2003). As the participants were part of intact teams that frequently worked together prior to the simulation, they were asked to complete the pre-survey keeping in mind teams with whom they typically worked. Participants were asked to think about the team they just worked with when responding to the post-measure.
The cases developed for the in situ simulation involved two multi-system trauma victims occurring in adjacent trauma bays. Specifically, the first scenario involved an intoxicated 46-year-old male (patient simulator) found on scene ejected from his vehicle in a motor vehicle collision. He was rushed in to the trauma bay with cervical and longboard spinal immobilization by confederate paramedics. He presented with a decreased level of consciousness, no chest rise on the left, agonal breathing, and an open fracture to the right leg. This patient’s pre-determined traumatic pathology was a critical tension pneumothorax, respiratory failure, hypotension from extremity blood loss, and a closed head injury.
The second case involved a 48-year-old female (patient simulator) who arrived via EMS after a 20-minute vehicle extrication in which her vehicle exhibited extensive front-end damage and a starred windshield. When she arrived to the trauma bay, she was unconscious, apneic, had diminished pulses, and an open fracture to the left forearm. Confederate paramedics delivered her with cervical and longboard spinal immobilization and intubated in the field. This patient’s pre-determined traumatic pathology was a hepatosplenic abdominal bleed resulting in hypotension, moderate head injury with a Glasgow Coma Scale (GCS) of 9, and an open extremity fracture.
All participants were expected to be proficient in the clinical skills needed to manage the cases, as they were all certified in ATLS® (Advanced Trauma Life Support), TNCC (Trauma Nursing Core Course), or ITLS (International Trauma Life Support). However, they were not specifically familiar with the presented cases. In both trauma scenarios, the primary objectives were to perform a standard ATLS evaluation with the appropriate interventions to identify and resolve clinical and task management issues arising during the care of the simulated patients.
In addition, both scenarios were purposely designed to facilitate behaviors consistent with the development of TMSs including role delegation, communication, relying upon others, and delegating tasks within teams. For example, the teams encountered a large cognitive and procedural load that needed immediate delegation for timely completion; team members had to effectively communicate success and difficulties with tasks; team members had to provide feedback regarding the patient’s status; teams had to recognize the critical nature of patient injuries and had to identify and utilize the expertise of its members to successfully resuscitate both patients; and teams had to re-establish team goals throughout the scenario.
After final completion of the 30-minute scenarios and debriefing, participants completed a post-survey (assessing transactive memory) and participated in a group debriefing session. The debriefing was performed in a standard approach using the 4E method (Mort & Donahue, 2004) by faculty debriefers from Emergency Medicine and Surgery. As a result of the complexity of the scenarios and size of the teams, an advocacy and inquiry approach (Rudolph, Simon, Dufresne, & Raemer, 2006) was utilized to ascertain the participant’s self-perception of their performance in a debriefing with good judgment process. Participants were asked to reflect upon and discuss aspects of the scenario that prompted them to consider other’s expertise, trust one another, communicate, access information from others, specialize, share tasks, coordinate behaviors, delegate, and set team goals. These topics were specifically mentioned to help us identify whether our scenario design successfully prompted these processes, and also to encourage participants to reflect upon these teamwork principles. The simulated scenarios were video recorded for performance review, but these videos were not used in the debriefing session.
Measures
Participants reported their discipline from the following: attending physician (surgery), attending physician (emergency medicine), resident physician (surgery), resident physician (emergency medicine), physician assistant, medical student, nurse, nurse practitioner, EMT/paramedic, respiratory therapist, EKG technician, radiology technician, or other. Transactive memory was measured via pre- and post-simulation self-report questionnaires (Lewis, 2003). This scale contains 15 items designed to assess three dimensions of transactive memory: specialization, coordination, and credibility. Specialization refers to the acknowledgment of distributed expertise within the team; coordination refers to the ability of team members to work together efficiently with greater cooperation, less confusion, and misunderstandings; and credibility refers to the degree to which group members trusted one another’s task expertise. Each item was scored on a Likert-type response scale ranging from 1 (strongly disagree) to 5 (strongly agree).
Performance was measured through videotaped review of the sessions. The recordings were subsequently reviewed by three emergency medicine attending physicians using a trauma performance checklist consisting of 22 graded and 6 timed items (Holcomb et al., 2002; see the appendix for descriptions of each scale.) Raters had been previously trained on the instrument by watching performance episodes, rating performance, and discussing discrepancies to achieve consensus on identified performance levels.
Data were analyzed using the SPSS version 18 (SPSS, Inc.; Chicago, IL) and a significance level of p < .05 was chosen. Cronbach’s alpha was used to assess the reliability of the transactive memory scale. Interrater reliability was assessed for the performance ratings. Intraclass correlation coefficients (ICC) representing the reliability of ratings across raters was .73 (95% CI = [0.06, 0.92], p < .001), warranting the use of mean ratings of performance. Paired-samples t tests were used to compare changes in transactive memory systems before and after the simulation. A Pearson correlation analysis was used to measure the association between transactive memory and performance ratings.
Because individuals were nested within teams in this study, both variables of interest are aggregated to the team level. The resuscitation performance ratings of individuals require aggregation as individuals from the same group received the same performance score and were thus not independent of one another. To avoid potential unclear interpretation of a cross-level correlation, transactive memory values are also reported at the aggregated team level, and the group-level relationship is examined. To ensure that we had sufficient evidence to warrant aggregate individual-level data to the team level, we computed agreement indices. The extent to which members within a single team agreed with one another at a level greater than chance was acceptable, rwg(j) = .71. In addition, the ICC was significant (p < .001), indicating that between-group response variance was greater than within-group variance. Thus, our data supported the use of averages in the group-level analyses, although the data were collected at the individual level.
Results
A total of 24 staff members participated in one of the four scenarios. The final sample was composed of emergency medicine resident physicians (29%), surgery resident physicians (25%), nurses (20%), and other allied health personnel (26%).
Pre- and Post-Simulation Comparisons
Mean pre- and post-simulation survey scores for the variables of interest were evaluated (Table 1). Changes in transactive memory for each group are illustrated both by dimension (i.e., credibility, specialization, and coordination) and as a combined mean. As shown, credibility and coordination increased significantly after participating in the simulation for all four groups. Specialization, however, only exhibited a significant increase for two of the groups, although the change in these groups approached significance (p < .10). When examining the TMS mean including all three dimensions, all four groups exhibited a significant difference.
Team-Level Transactive Memory Values Before and After Simulation.
Note. Boldfaced values are significant at p < .05. Pre cred M = team credibility rating before simulation; Post cred M = team credibility rating after simulation; d = Cohen’s d; Pre spec M = team specialization rating before simulation; Post spec M = team specialization rating after simulation; Pre coord M = team coordination rating before simulation; Post coord M = team coordination rating after simulation; Pre TMS total M = Mean of credibility, specialization, and coordination ratings before simulation; Post TMS total M = Mean of credibility, specialization, and coordination ratings after simulation.
The relationships among variables were also examined (Table 2). Bivariate correlations revealed that the post-simulation transactive memory was positively related to observer-rated performance (r = .79, p < .01). A deeper examination of this relationship within the dimensions indicates that team performance is significantly related to post-simulation credibility (r = .79, p < .01) and coordination (r = .88, p < .01), but not specialization. In addition, both post-simulation coordination and credibility were related to each other (r = .65, p < .01), but neither were significantly related to specialization. As expected, the TMS total mean was related to all dimensions (for specialization, r = .64, p < .01; for credibility, r = .79, p < .01; for coordination, r = .87, p < .01).
Intercorrelations of Study Variables.
Note. Post cred M = team credibility rating after simulation; Post spec M = team specialization rating after simulation; Post coord M = team coordination rating after simulation; Post TMS total M = mean of credibility, specialization, and coordination ratings after simulation.
p < .01.
Discussion
As a whole, participants in this study exhibited significant increases in transactive memory after being involved in an in situ trauma simulation. The results of this small pilot study suggest that during the designed trauma simulation, the participants developed a greater sense of coordination (ability of team members to coordinate their work efficiently) and team credibility (trusting teammates’ expertise). However, two groups had values that only approached significance for the specialization dimension. The non-significant change in specialization aligns with prior studies evaluating anesthesia teams and transactive memory (Michinov, Olivier-Chiron, Rusch, & Chiron, 2008). The instability of this relationship may be explained by a closer examination of the items on the specialization scale; these items asked respondents to judge the extent to which team members have different areas of expertise and specialized knowledge. As these were intact teams, members may already have had adequate knowledge about tasks roles and areas of expertise. Alternatively, a brief 30-minute simulation may not have been powerful enough to produce widespread differences in knowledge of specialties and role delineations, although it did demonstrate improved perceptions of credibility and coordination processes. Thus, simulation may be most effective in producing important cognitive processes within trauma teams that involve more “soft” processes such as trust, confidence, and working together.
Another important finding from this study involves the relationship between TMSs and team performance. Our results found that trauma teams with stronger transactive systems had higher performance ratings as evaluated by external subject matter experts using a standardized trauma checklist. These results suggest that medical simulation may be an effective pedagogical approach for the development of transactive memory systems within trauma teams. A high-fidelity simulated environment provides a valuable venue for the essential and specific form of understanding and communication that must be employed for the development of TMSs, namely, information regarding the knowledge, expertise, and relevant experience of other individuals in the team (Hollingshead, 1998). The quick formation of teams that was required in these scenarios prompted members to consider who was most qualified for assignments, the level of expertise of individuals in certain roles, the knowledge deficits of members in certain roles, and the requirements for the successful completion of critical tasks (Hollingshead, 1998). Thus, our results offer a potential explanation as to why simulation is effective for team training: It stimulates the development of transactive memory systems within teams. Future research using an experimental design should be conducted to more accurately identify the causal relationship between these variables.
These findings bring to light the importance of purposefully designing simulations to facilitate the development of transactive memory processes among medical teams, as team performance will also increase. Given the dynamic nature of health care teams and current health care challenges including duty hour restrictions for resident physicians, workforce shortages, and shift work, simulation may catalyze the development of TMSs within teams. The following basic guidelines might help simulation designers ensure that training scenarios are set up to facilitate these processes. First, team members should be encouraged to get together in the beginning moments of the scenario to delineate “who knows what” and “who will do what.” This will help set the stage for this knowledge structure to appropriately develop. Once this information is encoded, trainees must also be encouraged to communicate directly to appropriate group members in order to share relevant patient information and status updates. In addition, strategically designing scenarios in which teams must identify each other’s strengths and weaknesses and trust one another’s skills can help ensure that TMS will emerge. Finally, implementing team-based rewards and encouraging discussion of shared experiences may help instigate these processes.
Limitations
This study had various limitations. First, the absence of a comparison group who took the pre- and post-measures, but did not participate in the simulation, serves as a boundary condition of our findings. The observed improvement would be more compelling if still found compared with a group who completed the same instrument twice.
Another limitation concerns potential selection bias of those who participated in the simulation. The voluntary nature of the study allowed for some staff members to only observe the simulations. The 24 participants who chose to participate may be systematically different from those departmental staff who did not partake in the simulation. Future study designs should incorporate random assignment to participation and observation roles.
Finally, the composition of teams could impact conclusions extrapolated from this study. We did not obtain information regarding how familiar participants were with other members within their group, previous experiences, and so on. Although participants had likely participated in a number of trauma situations with their fellow team members, we did not collect any data concerning these relationships and experiences. It is reasonable to expect that less experienced physicians and trauma staff may benefit more from such team simulations.
Conclusion
This study demonstrates that empirical findings about team learning from other domains are also relevant to health care. In a recent meeting on simulation-based education, scholars (Issenberg et al., 2011) encouraged researchers to use simulation to validate learning theories from other nonmedical domains in a health care context. Our findings validate this recommendation by expanding the domain to which transactive memory is applicable.
In sum, simulation-based research needs to be grounded in a theoretical or conceptual framework. We incorporated socio-cognitive theory to help understand why simulation is effective. Our results suggest that in situ simulation is effective because it prompts transactive memory systems within teams to develop. TMS helps shed light on how team knowledge is created, shared, and integrated to leverage distributed expertise, ultimately impacting team performance.
Footnotes
Appendix
Acknowledgements
We wish to express our appreciation and indebtedness to our reviewers for their integral role in improving this article.
Author Contributions
All authors contributed equally to this article. AKG and RAA conceived, designed, and implemented the activity. AKG performed the statistical analysis. AKG and RAA wrote the final manuscript.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
Author Biographies
Contact:
Contact: