Team-based practice—where responsibility for the delivery of patient-centered care is distributed across a team of health professionals working collaboratively—is increasingly becoming the norm in health care, yet practitioners and learners on clinical teams often have difficulty “getting on the same page” to provide optimal patient care.1 Developing a common understanding of both the roles of team members and the structure of the work is called developing a shared mental model (SMM).2 Several empirical studies, both within and beyond health care, have demonstrated the value of SMMs in supporting teamwork.1 , 3 It follows then that health professions learners be trained to recognize, adapt, and align their mental models with those of their health care team members to create an SMM related to patient care. However, we need to know what an SMM is and how it can be developed and assessed before we can advocate its use in the context of health professions education (HPE). In our study described here, we explored the SMM construct as it relates to clinical teamwork among health professions learners.
In cognitive psychology, mental models are cognitive representations of the environment, including objects, activities, situations, or people.4 , 5 These organized knowledge frameworks allow individuals to understand phenomena, develop inferences, and make predictions.6 When the organized mental representations of individual team members overlap, they are said to have an SMM. SMMs encompass declarative, procedural, and strategic knowledge (i.e., content) as well as the organization of that knowledge (i.e., the knowledge structure or relationships among concepts).2 , 7 , 8 (For a clinical example illustrating this distinction between SMM content and structure, see Supplemental Digital Appendix 1 at http://links.lww.com/ACADMED/A496.) SMMs fall into two interdependent content domains—task-related and team-related mental models.2 , 3 , 5 , 7 , 9 Task-related mental models include goals and performance requirements; team-related mental models focus on interpersonal interactions and team member skills.5 , 7
SMMs also have two distinct properties—similarity and accuracy.7 Similarity is the extent to which team members share organized knowledge. This “sharedness” refers to the degree of overlap among team members’ mental models2 and may range from low to high. Accuracy reflects the degree to which team members’ mental models are consistent with reality10 or what is considered by expert consensus to be the ideal mental model.3 , 4 Though multidimensional,7 , 11 many simply describe SMMs as shared understandings among team members or as members being on the same page (see Figure 1).7 , 12
SMMs are considered one of the key coordinating mechanisms of effective teamwork, along with closed-loop communication and mutual trust.9 They support team members’ ability to (1) predict each other’s needs, (2) identify changes in the team or task, (3) adjust strategies, and (4) coordinate behavior.3 Empirical evidence suggests that highly similar and accurate mental models among team members support team functioning, yielding improvements in team processes and performance.1 , 3 , 5 , 10 , 13 Empirical studies outside of HPE have also shown that a range of team interventions may effectively facilitate SMM development in teams.6 , 7 Given the importance of the SMM construct in the teamwork literature and its relevance to health professions training specifically and health care generally, we believe that the potential utility and impact of SMMs in education warrant a comprehensive review and synthesis of the existing literature.
Considering that our goal was to explore the SMM construct as it relates to clinical teamwork in the context of HPE in a comprehensive and inclusive manner and that we discovered few empirical studies of SMMs in health professions learners in our initial PubMed search, we felt that a scoping review was the appropriate approach for our study. Colquhoun and colleagues14 defined a scoping review as a “[form of] knowledge synthesis that addresses an exploratory research question aimed at mapping key concepts, types of evidence, and gaps in research related to a defined area or field by systematically searching, selecting and synthesizing existing knowledge.” The scoping review methodology supports less restrictive inclusion criteria than the systematic review methodology and also allows for the inclusion of information from disparate sources.14 , 15
There are four primary purposes for conducting a scoping review, including (1) to examine the extent, range, and nature of research activity in a given area; (2) to determine the value of undertaking a full systematic review; (3) to summarize and disseminate research findings; and (4) to identify gaps in the existing body of literature.14 , 15 The primary objective of this scoping review was to conduct a broad investigation of the SMM construct as it relates to clinical teamwork in the context of HPE, to identify gaps in the current literature, and to disseminate these findings to the HPE community.
Following the five required steps outlined in Levac and colleagues’15 refined methodological framework for scoping reviews, we (1) identified the research questions; (2) identified relevant studies; (3) selected studies to be included in the review; (4) charted the data; and (5) collated, summarized, and reported the results.15 While we did provide our local educational research community with opportunities to critique the study design and to review an early draft of this article, we did not feel that this engagement rose to the level of a stakeholder consultation (the optional sixth step). The methods we used in each step are detailed below.
Identifying the initial research questions
The initial step of the scoping review process is to develop research questions to guide the review.15 We generated research questions that would allow for a broad exploration of the SMM construct in the context of clinical teamwork in HPE, including definition, application, interventions, and measurement approaches. We refined our questions during several research team meetings and finalized them as:
- How has the SMM construct been defined and applied in relation to clinical teamwork involving health professions learners?
- What educational interventions are used to develop health professions learners’ SMMs related to clinical teamwork? What impact do these interventions have on SMM development and related outcomes?
- How are SMMs measured in clinical teams with health professions learners, and what are the findings?
Identifying relevant studies
Following an initial pilot search of PubMed, using the search terms “shared mental model” OR “shared mental models,” to identify synonyms and to locate the entry of this construct into the HPE literature, our reference librarian (E.W.) generated specific search terms, which encompassed the SMM construct (e.g., “shared mental model” and “shared mental models”). And, because we were most interested in trainees (rather than established clinical practitioners), keywords relating to undergraduate and graduate health professions learners were also included (e.g., “clinical training” and “nursing education OR pharmacy education OR medical education OR dental education”). After reaching consensus with the team regarding the search terms, the librarian (E.W.) developed a database-specific search strategy intended to identify the relevant literature from a broad array of English-language, academic, and gray literature sources (see Supplemental Digital Appendix 2 at http://links.lww.com/ACADMED/A496 for details regarding the search terms and strategy).
Using this approach, we conducted two separate searches of the CINAHL, EMBASE, ERIC, Scopus, Web of Science, PubMed, and PsycINFO databases (see Figure 2). The initial PubMed pilot search retrieved references related to the SMM construct as it applies to teamwork and simulation. This first comprehensive literature search, conducted in December 2015 by our reference librarian (E.W.), spanned January 2000 through December 2015. As data analysis proceeded from January to April 2016, we discovered new search terms that could potentially both expand and refine the search. And, because the scoping review methodology supports an iterative approach to searching the literature as new ideas or search terms are generated during the review process, a second search (conducted by E.W. and L.C.F.) was conducted in May 2016. This search included an expanded list of search terms that were encountered in those articles reviewed following the first search (e.g., additional terms related to the SMM construct including “team mental model(s),” “taskwork,” and “teamwork”; more specific terms related to trainees such as “resident(s),” “internship and residency,” “fellow(s),” and “fellowship”; and terms that attempt to capture “interdisciplinary” and “interprofessional” teams) (see Supplemental Digital Appendix 2 at http://links.lww.com/ACADMED/A496 for details regarding this expanded search strategy). We (L.C.F. and D.D.) also searched the reference lists of all included articles by hand to identify additional articles for review.
Selecting studies for review
Two authors (L.C.F. and D.D.) independently reviewed all titles and abstracts for eligibility using a screening tool that allowed for direct comparison of each reviewer’s recommended action (i.e., include in the primary analysis, include as a background paper, or exclude) and rationale for eligibility (e.g., mentioned an SMM or team mental model [TMM] in the context of HPE and included learner categories). After this initial screening, we read the full texts of the articles deemed eligible for inclusion. Eligible articles included empirical and descriptive studies, conceptual papers, letters or communications, commentaries, perspectives, meta-analyses, systematic reviews, abstracts, and poster presentations. Inclusion criteria were developed by the team based on our guiding research questions and required that articles (1) use the term “shared mental model(s)” or “team mental model(s)”; (2) pertain to undergraduate-level (i.e., medical, pharmacy, nursing, physical therapy students, etc.) and/or graduate-level (i.e., residents, fellows) health professions learners; and (3) take place in a real or simulated clinical setting. We excluded articles not in English and those not related to clinical teamwork or clinical training. Additionally, during the full data extraction, we found that several of the articles that had initially met our inclusion criteria treated the construct of SMMs or TMMs in a cursory manner (n = 8) or contained only a brief mention of the education of undergraduate- or graduate-level health professions trainees (n = 8). We deemed these articles “superficial” and excluded them from the review (see Figure 2). Any disagreements regarding article inclusion were resolved through discussion (L.C.F. and D.D.).
Charting the data
Two authors (L.C.F. and D.D.) developed a data collection form to collect all the information necessary to answer our research questions. Data categories included author, year of publication, study design (descriptive, experimental, qualitative, quantitative), educational setting, learner characteristics, focus of article, description of intervention (if applicable), SMM definition, SMM content and properties, application of the SMM construct, SMM measurement methods, and key outcomes/findings. We (L.C.F. and D.D.) piloted the data collection form by each extracting data independently from five articles. The high degree of consistency between our extracted data sets supported the utility of the collection form. Our research team identified a few pieces of missing information, so we added article type (program, empirical, conceptual, opinion/position, summary), specific study aims, and target group (i.e., study population) to the form. Using the refined collection form, one author (L.C.F.) revisited the first five articles to extract data relevant to the newly added categories, then extracted data from the remaining 18 articles. Another author (D.D.) then reviewed all of the extracted data for accuracy and completeness. Discrepancies were resolved through discussion. Two authors (L.C.F. and B.C.O.) reviewed all extracted data independently, discussed the findings, and ensured that the extracted data would help us best answer our research questions.
Collating, summarizing, and reporting findings
One author (L.C.F.) reviewed and then analyzed the extracted data using both narrative and numerical description. The narrative summaries, combined with the numerical analysis, were intended to highlight the most relevant findings related to each of our three main research questions, including (1) the proportion of studies that defined the SMM construct, how SMMs were characterized in each definition (i.e., shared knowledge, knowledge organization, SMM properties, etc.), and the nature of the application of the SMM construct in HPE (i.e., clinical setting, learner characteristics, etc.); (2) the categories of interventions related to SMM development in HPE (i.e., teamwork curricula, team training, or teamwork supportive tools); and (3) approaches to measuring SMMs taken by researchers in the context of HPE. After several in-depth discussions among the research team, we finalized the data summaries.
Our database and reference list searches retrieved a total of 1,273 records (see Figure 2), and, after removing duplicates, 853 records remained. We screened all titles and abstracts and excluded 753 records on the basis of our eligibility criteria. The full texts of 100 articles were read, and, in the end, 23 articles met our inclusion criteria and were included in our review (see Tables 1 and 2 and Supplemental Digital Appendix 3 at http://links.lww.com/ACADMED/A496 for the data associated with our research questions and Supplemental Digital Appendix 4 at http://links.lww.com/ACADMED/A496 for the complete data we extracted from all the included articles).
We address the topic of each research question in turn: (1) definition and application, (2) interventions, and (3) measurement.
How has the SMM construct been defined and applied in relation to clinical teamwork involving health professions learners?
Less than two-fifths of the articles (9/23) explicitly defined the SMM (or TMM) construct (see Table 1).16–24 All definitions characterized an SMM as a cognitive construct encompassing knowledge shared across team members.16–24 Definitions characterized mental models as “shared,”17 , 18 , 22 “common,”21 , 24 or “overlapping”23 among team members. One referenced mental model similarity.19 Another implied that mental models “held by members” are shared, but the language was not explicit.20 Two-thirds of all definitions (6/9) referenced knowledge structure,16 , 18–20 , 22 , 23 either explicitly19 , 20 or by mentioning “organized”16 knowledge or cognitive “representation(s).”18 , 22 , 23 Four definitions differentiated between the content domains of task- and team-related knowledge.20–23
All articles discussed SMMs in the context of hospital-based care, most often in interprofessional teams (14/23) and high-intensity settings (i.e., surgery, trauma) (14/23).16–23 , 25–30 Most articles included graduate-level physicians (21/23)16–36 within clinical teams (18/23).16–23 , 25–34 Three articles involved undergraduate-level medical students.29 , 33 , 37 Nonphysician learners included junior nurses,36 nursing and physical therapy students,37 , 38 and other unspecified health professions learners.33
Most articles discussed SMMs as an outcome (e.g., of team interaction, team training, or curricular interventions) (12/23).17 , 18 , 20 , 22 , 23 , 26 , 28 , 30 , 34 , 36–38 Others discussed SMMs as a prerequisite for effective teamwork or performance (5/23)16 , 19 , 24 , 29 , 31 or as both an outcome and prerequisite (5/23).21 , 25 , 27 , 33 , 35
What educational interventions related to SMMs are described, and what impact do they have?
Interventions designed to foster SMMs (6/23) included teamwork curricula/training28 , 35 , 37 and teamwork supportive tools20 , 24 , 30 (see Table 2). Most interventions focused on taskwork such as resuscitation,28 developing treatment plans,30 crisis care,34 and rounding.20 Others focused on teamwork skills35 and team-based behaviors.37 Interventions occurred in both simulated clinical settings and with in situ clinical teams involving graduate-level (5/6)20 , 28 , 30 , 34 , 35 and undergraduate-level learners (1/6).37 Most interventions focused on the clinical team20 , 28 , 30 , 34 rather than on the individual learners directly.35 , 37
Teamwork curricula/training programs.
Carbo et al35 described a case-based, team training curriculum intended to develop residents’ teamwork skills, including their SMMs; they found that learners’ knowledge of key teamwork skills nearly doubled, increasing from 35% pre training to 67% post training. Hicks et al28 proposed a simulation-based, emergency department team training program based on the Crew Resource Management principles to develop SMMs of resuscitation processes. Garbee et al37 discussed a case-based, simulation curriculum with postcase debriefing to support SMM development in health professions undergraduates. Participant and observer SMM subscale scores increased significantly post intervention.
Teamwork supportive tools.
Wu et al34 described an interactive, large-screen display and tablet, which promoted crisis care team dialogue to support the development of SMMs. Leykum et al30 reported on their design, implementation, and planned evaluations of a structured communication tool to improve pre- and postround briefings. Xie et al20 developed a checklist to support an SMM of family-centered rounds; though highly used, it was applied inconsistently, prompting further team training on checklist items. The impact of these tools on SMM development was not directly measured.
How are SMMs measured in clinical teams with health professions learners, and what are the findings?
Researchers measured SMMs qualitatively25 and quantitatively16 , 18 , 19 , 22 , 23 , 25 , 37 (7/23) (see Supplemental Digital Appendix 3 at http://links.lww.com/ACADMED/A496). Three studies focused on taskwork, including anesthesia induction,16 intensive care unit (ICU) handoffs,18 and pediatric intensive care unit (PICU) patient care.25 Two focused on teamwork, including medicine team members’ roles and responsibilities19 and general ICU teamwork.37 Two others quantified the degree of similarity among team members’ mental model content and structure16 , 18; one also measured SMM accuracy.16
Custer et al25 organized verbal fragments from interviews into themes to elucidate SMM content related to complex PICU patients. SMMs facilitated longitudinal care across handoffs, but variable interpretations of a patient’s condition negatively impacted SMM development.25
McComb et al19 developed a seven-point, Likert-type survey to investigate the similarity of nurses’ and physicians’ mental models related to roles and responsibilities on general medicine wards. Participants rated the professional they believed to be responsible for a specific role (i.e., diagnosis, administering medicines, etc.). Practitioners’ mental models were significantly different for 14 of 22 roles. Garbee et al37 used a three-item, Likert-type SMM subscale within an overall teamwork scale to measure team-level performance. The authors reported significant improvements in SMM subscale scores post simulation and debrief as rated by both participants and observers. Burtscher et al16 employed concept mapping to investigate the TMM of anesthesia induction. Residents and nurses arranged 30 task-related concept cards (e.g., ventilate patient, hand intubation set) by sequence and role to create individual maps. Maps were compared (1) within each team to assess TMM similarity and (2) with maps produced by experts to assess TMM accuracy. When TMM accuracy was high, TMM similarity was positively related to performance.16 Similarly, Nakarada-Kordic et al23 developed a computer-based card-sorting tool to measure SMMs in operating room teams that were comprised of three subteams (surgery, anesthesia, and nursing). Before each of two simulated laparotomies, team members sorted 20 key tasks by sequence and subteam responsibility. For more than half the tasks, the authors found mental model similarity across team members for task sequence but poor agreement for subteam responsibility. Mamykina et al18 analyzed speech fragments from ICU team members during handoffs to generate a Shared Mental Model Index (SMMi), which represented the weighted proportion of overlapping statements. Work rounds supported the alignment of individuals’ mental models around patient care. In another study, Mamykina et al22 analyzed critical care ICU teams’ verbal handoffs. They reported higher SMMi scores for statements related to patient presentation and those reflecting past events, as well as an association between SMMi score and a team coherence measure.
We conducted this scoping review to explore the construct of SMMs as it is applied to clinical teamwork and health professions learners. Few articles explicitly defined the SMM construct, interventions to foster SMMs were rare, and few studies measured SMMs. On the basis of these findings and our review of the literature outside HPE, we offer the following recommendations to enhance education and research related to SMMs: (1) carefully define the SMM construct to promote consistent application, (2) improve both the design and evaluation of interventions that support SMMs, and (3) measure key aspects of SMMs in clinical teams with health professions learners. We also discuss challenges related to SMM definitions, interventions, and measurement as well as additional considerations related to SMMs in the context of clinical teamwork.
Defining the SMM construct
Several authors described the lack of clarity in the definition of an SMM.6 , 7 McComb and Simpson11 noted that, in health care, authors generally provide superficial definitions of SMMs and often fail to articulate the dimension of the SMM under study, making it difficult to apply the construct consistently in practice and research. Our review corroborated these findings.
A clear, detailed definition allows researchers and educators to accurately characterize the SMM construct. We suggest that such a definition include three key components, based on definitions proposed by Canon-Bowers and colleagues,2 Mathieu and colleagues,3 and Klimoski and Mohammed.6 First, capturing knowledge content (concepts) as well as structure (relationships among concepts) in the definition differentiates the SMM from other common team cognition constructs (i.e., group learning, situation awareness, and strategic consensus)7 and acknowledges the centrality of knowledge structure to the SMM construct.7 Second, specifying that mental model “sharedness” connotes commonality in cognitive representations adds precision to the definition.6 Third, characterizing the SMM as an individually held knowledge structure that teammates have in common highlights that measurement of this team-level construct requires aggregation of data across individuals.
To clarify the meaning of the SMM construct in the context of health care and to promote its consistent use and application across HPE, we developed an operational definition, adding common characteristics of the definitions we identified in our review16–24 and situating the construct in the context of teamwork among health care professionals. From this synthesis, we propose the following definition of a mental model that is shared among health care team members:
A shared mental model is an individually held, organized, cognitive representation of task-related knowledge and/or team-related knowledge that is held in common among health care providers who must interact as a team in pursuit of common objectives for patient care.
Two content domains characterize SMMs—task- and team-related knowledge. Task-related knowledge encompasses task goals, procedures, strategies, and relevant equipment. Team-related knowledge includes role interdependencies, responsibilities, and communication patterns as well as team members’ knowledge, skills, attitudes, and preferences. To address the two dimensions of an SMM—concepts and knowledge organization—we included the term organized to refer to knowledge structure or the relationships among concepts. Considering the two properties of an SMM—similarity and accuracy—it is important to recognize that the term common in our proposed definition signifies a degree of similarity that will vary in intensity from team to team, ranging from low to high. We excluded the term accuracy because team members may have highly similar mental models that are accurate, inaccurate (i.e., the SMM neither reflects the true state of the world nor overlaps with an expert’s mental model), or indeterminate (i.e., the situation or task is ambiguous or uncertain).
Throughout the review process, we debated several challenging elements of the SMM construct. We pondered how to characterize the relationship between task-related knowledge and team-related knowledge in health care teams. For example, is it possible to have team-related knowledge without task-related knowledge? Because the team and its task are inextricably connected (i.e., the health care team gathers to do a job related to patient care, not just to socialize), we struggled with the conventional separation of the task- and team-related knowledge content domains.7 Because this separation of the SMM content domains is prevalent in the broader literature, as is the understanding that team members hold multiple SMMs simultaneously7 (e.g., task requirements and responsibilities), we aligned our definition with common uses of this construct to both gain conceptual clarity and promote standardized use across HPE.
We also debated the SMM properties of similarity and accuracy. For example, we discussed instances where team members’ mental models might have minimal overlap. While this overlap might technically generate an SMM, little is known about the ideal level of similarity in the clinical context, and the question remains,22 “What are the functional consequences of a barely existent SMM?” We also discussed whether or not to include accuracy in our definition. Though similarity and accuracy of team members’ mental models are desirable,3 for an SMM to exist, mental models only need to be shared. There is no requirement that they reflect reality or align with an expert’s mental model. A team with an inaccurate SMM of clinical task priorities might actively pursue secondary goals, negatively affecting team performance5 and patient care. Therefore, we excluded the term accuracy from our definition, allowing for the real possibility that team members might have highly similar but inaccurate SMMs. Another reason to exclude accuracy was that, absent an expert mental model, it is impossible to determine SMM accuracy.
Ultimately, our proposed definition aims to provide a coherent conceptual framework for the SMM construct and to guide health professions educators and researchers in the practical application of SMMs in health care teams rather than to serve as an absolute truth.39
Applying the SMM construct
An SMM can function as both a dependent and independent variable in education. Educational interventions may support SMM development, and a team’s SMMs can impact learning and performance. We found SMMs that were described as expected outcomes of interventions as well as prerequisites for improved team performance. We suspect that this dual use contributes to what Mohammed and colleagues7 characterized as “a fair amount of conceptual confusion surrounding [SMMs]” in research and practical application. While both uses are acceptable, achieving conceptual clarity requires researchers and practitioners to explicitly define how they are using the term.
The prevalence of high-intensity health care teams described in the articles we reviewed is consistent with the broader SMM literature, where the construct has been applied frequently to teams in high-risk environments, such as cockpits and military combat.2 , 5 , 40 , 41 This focus reveals an important gap in the literature because most health care occurs in lower-intensity, outpatient settings.42 Though clinical teams practicing in lower-intensity settings (e.g., ambulatory care) would not be expected to encounter the same emergent situations as those teams in higher-intensity settings (e.g., the ICU), where the need for immediate coordination is generally great,25 , 31 they do face unique communication and organizational challenges as members of complex, “virtual,” distributed health care teams that provide care in an asynchronous fashion.43 , 44 Whether or not accurate SMMs among these health care team members—with respect to their collective task (i.e., goals of care for a specific patient), their respective roles and responsibilities related to that patient’s care, or their attitudes toward patient safety—would benefit team performance and improve patient care and safety warrants further study.12
Designing educational interventions to facilitate SMM development
Although the broader literature offers a wide range of interventions that facilitate SMM development (i.e., team training, planning, leadership, and reflexivity),6 , 7 we found few interventions focused on clinical teamwork among health professions learners.
Despite the complexity of operationalizing and measuring SMMs in teams, educators might use the SMM construct to design interventions to improve team performance outcomes or enhance knowledge of teamwork principles. We offer a few examples of such interventions used within and beyond health care that could be adapted for use in clinical, team-oriented education.
Team training has been studied extensively across fields; it may support either the pursuit of general teamwork outcomes or the development of SMMs7 , 45 and includes computer-based training to develop general teamwork competencies; team-interaction training, where teams are trained to coordinate their actions; and cross-training, where team members learn about the tasks, roles, and responsibilities of other team members.7 Computer-based training has improved team knowledge, communication, and skills45 and increased both the similarity and accuracy of TMMs.13 Team-interaction training and cross-training have led to improvements in team outcomes46 and promoted SMM development.41 Team-interaction training during an inpatient rotation, for example, might include a case-based curriculum focused on effective team communication. Cross-training in HPE might provide learners with opportunities to shadow other team members and to see teamwork from various perspectives, such as a medical resident shadowing a nurse during nursing rounds.
Other opportunities to use SMMs include during team huddles—to develop team goal setting, coordination, and communication skills47—and during team coaching, team performance monitoring, and group and individual reflections. For example, a group reflection exercise, implemented in July as new residents arrive, could allow new residents to reflect on their individual and shared expectations related to team functioning and processes in their new surroundings. Alternatively, an SMM-focused intervention might serve as a team diagnostic tool to encourage team members to explore how the lack of an SMM might have contributed to a near miss during a patient encounter.
The complexity of assessing and representing cognition at the individual and group levels has been characterized as a “thorn in the side” of this field of research.48 Several factors contribute to the complexity of measuring SMMs and to the limited empirical progress in SMM research.48 The SMM construct lacks a common definition4 and is inherently complex with two content domains (task- and team-related knowledge), two dimensions (concepts and organization), and two properties (similarity and accuracy).
Because SMMs are “organized knowledge structures,”2 their measurement requires that the content of each individual team member’s mental model be elicited and the structure of their knowledge elucidated.48 Then individual mental models must be evaluated and aggregated to determine the degree of similarity or “sharedness.” Accuracy may be determined by comparison with an ideal mental model (if one exists) that is derived by aggregating mental models from subject matter experts.3 , 7
Though the challenges with measuring SMMs are well documented,7 the literature suggests that they are not insurmountable.4 , 8 , 48 The four main measurement techniques described in the literature include paired comparisons, card sorting, concept mapping, and qualitative analysis.7 , 8 To choose a method to measure an SMM, the purpose and setting of the investigation must be considered.7
We considered the limitations of measuring SMMs, from instrument development to application. Most instruments are context dependent and lack generalizability.49 Logistical difficulties in administering these instruments include the substantial time needed for completion, the difficulty of conducting in situ team SMM measurements, and the analytic expertise required to analyze the data. In light of these challenges, SMM measurement may not be feasible for many educators and may limit the practical application of SMMs as a diagnostic tool to assess team performance in the workplace or learning environment.49 Alternatively, direct measurement of SMMs may not be necessary if the outcomes expected, such as team processes (e.g., coordination, communication) or measures of team effectiveness (e.g., performance metrics), can be determined.15 , 49 However, without direct measurement of mental model similarity and accuracy, it would be impossible to tie any team performance improvements directly to SMM development.
Our analysis revealed that the complexity of the SMM construct, in combination with the myriad measurement issues previously identified and substantiated in several of the articles we reviewed, may limit the wide applicability of this construct in HPE. In addition, we believe that, while some researchers have discussed the benefits of distinctive perspectives,2 the general emphasis in the literature on team members’ mental model “convergence”50 may lead to a biased view in favor of greatly overlapping mental models. And, the overt promotion of SMMs in clinical teams, without the creation of a safe team atmosphere where alternate issues or solutions to problems are welcomed and expected, may have unintended negative consequences such as promoting “groupthink.”51 Groupthink may prevent the potentially productive divergence of opinion,44 result in lower quality team decision making,52 or instantiate the status quo that is in need of change. For these reasons, some researchers feel that members of a team must be given the opportunity to bring their diverse knowledge and perspectives forward for the team’s consideration50 and that mental model complementarity (i.e., where team members’ mental models are related to one another in a complementary fashion) might be as important as mental model similarity among team members in improving team performance.22 , 50
First, because shared and mental model(s) were defining keywords in our literature search, articles that applied the SMM construct but did not contain these terms were not captured. And, though our search included databases that capture gray literature sources (EMBASE, Scopus, PsycINFO, and Web of Science), our search of the gray literature was limited, and we may have missed relevant information. Next, our review was based on a small set of articles that met our inclusion criteria, which speaks to the limited number of publications in the field and perhaps to the limited utility of the SMM construct in HPE. However, the included articles accurately reflected the published literature focused on SMMs to support clinical teamwork among health professions learners.
Through this scoping review, we explored how the SMM construct has been applied to clinical teamwork involving health professions learners. The gaps we identified in this review revealed opportunities for refinements and further research. We recommend that, if health professions educators and researchers choose to use the SMM construct, they (1) consistently apply a clear definition of the SMM construct; (2) design and evaluate interventions to support SMM development in a variety of clinical environments; and (3) practice methodological rigor in measuring SMM content, structure, similarity, and accuracy. Following these recommendations can expand our understanding of the ways in which SMMs can empower team members, including health professions learners, to get on the same page and more effectively collaborate to deliver optimal team-based clinical care.
Acknowledgments: The authors wish to thank Pat O’Sullivan, EdD, the Research in Medical Education group, and the Utrecht University–University of California, San Francisco health professions doctoral students for their helpful suggestions.
1. Westli HK, Johnsen BH, Eid J, Rasten I, Brattebø G. Teamwork skills, shared mental models, and performance in simulated trauma teams: An independent group design. Scand J Trauma Resusc Emerg Med. 2010;18:47.
2. Canon-Bowers JA, Salas E, Converse S. Shared mental models in expert team decision-making. In: Individual and Group Decision-Making: Current Issues. 1993:Hillsdale, NJ: Lawrence Erlbaum Associates; 221–246.
3. Mathieu JE, Heffner TS, Goodwin GF, Salas E, Cannon-Bowers JA. The influence of shared mental models on team process and performance. J Appl Psychol. 2000;85:273–283.
4. Langan-Fox J, Code S, Langfield-Smith K. Team mental models: Techniques, methods, and analytic approaches. Hum Factors. 2000;42:242–271.
5. Lim B-C, Klein KJ. Team mental models and team performance: A field study of the effects of team mental model similarity and accuracy. J Organ Behav. 2006;27:403–418.
6. Klimoski R, Mohammed S. Team mental model: Construct or metaphor? J Manag. 1994;20:403–437.
7. Mohammed S, Ferzandi L, Hamilton K. Metaphor no more: A 15-year review of the team mental model construct. J Manag. 2010;36:876–910.
8. Wildman JL, Salas E, Scott CP. Measuring cognition in teams: A cross-domain review. Hum Factors. 2014;56:911–941.
9. Salas E, Sims DE, Burke CS. Is there a “big five” in teamwork? Small Group Res. 2005;36:555–599.
10. Edwards BD, Day EA, Arthur W Jr, Bell ST. Relationships among team ability composition, team mental models, and team performance. J Appl Psychol. 2006;91:727–736.
11. McComb S, Simpson V. The concept of shared mental models in healthcare collaboration. J Adv Nurs. 2014;70:1479–1488.
12. Burtscher MJ, Manser T. Team mental models and their potential to improve teamwork and safety: A review and implications for future research in healthcare. Saf Sci. 2012;50:1344–1354.
13. Smith-Jentsch KA, Mathieu JE, Kraiger K. Investigating linear and interactive effects of shared mental models on safety and efficiency in a field setting. J Appl Psychol. 2005;90:523–535.
14. Colquhoun HL, Levac D, O’Brien KK, et al. Scoping reviews: Time for clarity in definition, methods, and reporting. J Clin Epidemiol. 2014;67:1291–1294.
15. Levac D, Colquhoun H, O’Brien KK. Scoping studies: Advancing the methodology. Implement Sci. 2010;5:69.
16. Burtscher MJ, Kolbe M, Wacker J, Manser T. Interactions of team mental models and monitoring behaviors predict team performance in simulated anesthesia inductions. J Exp Psychol Appl. 2011;17:257–269.
17. Janss R, Rispens S, Segers M, Jehn KA. What is happening under the surface? Power, conflict and the performance of medical teams. Med Educ. 2012;46:838–849.
18. Mamykina L, Hum RS, Kaufman DR. Patel VL, Kaufman DR, Cohen T. Investigating shared mental models in critical care. In: Cognitive Informatics in Health and Biomedicine: Case Studies on Critical Care, Complexity and Errors. 2014:London, UK: Springer London; 291–315.
19. McComb SA, Lemaster M, Henneman EA, Hinchey KT. An evaluation of shared mental models and mutual trust on general medical units: Implications for collaboration, teamwork, and patient safety [published online ahead of print February 24, 2015]. J Patient Saf. doi: 10.1097/PTS.0000000000000151.
20. Xie A, Carayon P, Cox ED, et al. Application of participatory ergonomics to the redesign of the family-centred rounds process. Ergonomics. 2015;58:1726–1744.
21. Tourgeman-Bashkin O, Shinar D, Parmet Y, Zmora E. Inter-observer agreement among medical professionals in critical care of neonates and children. J Healthc Eng. 2010;1:357–366.
22. Mamykina L, Jiang S, Collins SA, et al. Revealing structures in narratives: A mixed-methods approach to studying interdisciplinary handoff in critical care. J Biomed Inform. 2016;62:117–124.
23. Nakarada-Kordic I, Weller JM, Webster CS, et al. Assessing the similarity of mental models of operating room team members and implications for patient safety: A prospective, replicated study. BMC Med Educ. 2016;16:229.
24. O’connor P, O’dea A, Lydon S, et al. A mixed-methods study of the causes and impact of poor teamwork between junior doctors and nurses. Int J Qual Health Care. 2016;28:339–345.
25. Custer JW, White E, Fackler JC, et al. A qualitative study of expert and team cognition on complex patients in the pediatric intensive care unit. Pediatr Crit Care Med. 2012;13:278–284.
26. Duthie EA. Recognizing and managing errors of cognitive underspecification. J Patient Saf. 2014;10:1–5.
27. Gonzalo JD, Moser E, Lehman E, Kuperman E. Quality and safety during the off hours in medicine units: A mixed methods study of front-line provider perspectives. J Hosp Med. 2014;9:756–763.
28. Hicks CM, Bandiera GW, Denny CJ. Building a simulation-based crisis resource management course for emergency medicine, phase 1: Results from an interdisciplinary needs assessment survey. Acad Emerg Med. 2008;15:1136–1143.
29. Dickerson EC, Alam HB, Brown RK, Stojanovska J, Davenport MS; Michigan Radiology Quality Collaborative. In-person communication between radiologists and acute care surgeons leads to significant alterations in surgical decision making. J Am Coll Radiol. 2016;13:943–949.
30. Leykum LK, Lanham HJ, Provost SM, McDaniel RR, Pugh J. Improving outcomes of hospitalized patients: The Physician Relationships, Improvising, and Sensemaking intervention protocol. Implement Sci. 2014;9:171.
31. Reader TW, Flin R, Mearns K, Cuthbertson BH. Developing a team performance framework for the intensive care unit. Crit Care Med. 2009;37:1787–1793.
32. Thomas EJ, Sexton JB, Lasky RE, Helmreich RL, Crandell DS, Tyson J. Teamwork and quality during neonatal care in the delivery room. J Perinatol. 2006;26:163–169.
33. Weller J, Boyd M, Cumin D. Teams, tribes and patient safety: Overcoming barriers to effective teamwork in healthcare. Postgrad Med J. 2014;90:149–154.
34. Wu L, Cirimele J, Card S, Klemmer S, Chu L, Harrison K. Maintaining shared mental models in anesthesia crisis care with nurse tablet input and large-screen displays. In: Proceedings of the 24th Annual ACM Symposium Adjunct on User Interface Software and Technology. UIST ’11. 2011:71–72. Available at: hci.stanford.edu/publications/2011/icogaid/icogaid-uist2011.pdf
. Accessed November 2, 2017.
35. Carbo AR, Tess AV, Roy C, Weingart SN. Developing a high-performance team training framework for internal medicine residents: The ABC’s of teamwork. J Patient Saf. 2011;7:72–76.
36. Weller JM, Barrow M, Gasquoine S. Interprofessional collaboration among junior doctors and nurses in the hospital setting. Med Educ. 2011;45:478–487.
37. Garbee DD, Paige J, Barrier K, et al. Interprofessional teamwork among students in simulated codes: A quasi-experimental study. Nurs Educ Perspect. 2013;34:339–344.
38. Senette L, O’Malley M, Hendrix T. Passing the baton: Using simulation to develop student collaboration. Clin Simul Nurs. 2013;9:e39–e46.
39. Eva KW. What’s in a name? Definitional clarity and its unintended consequences. Med Educ. 2017;51:1–2.
40. Mathieu JE, Rapp TL, Maynard MT, Mangos PM. Interactive effects of team and task shared mental models as related to air traffic controllers’ collective efficacy and effectiveness. Hum Perform. 2009;23:22–40.
41. Marks MA, Sabella MJ, Burke CS, Zaccaro SJ. The impact of cross-training on team effectiveness. J Appl Psychol. 2002;87:3–13.
42. Bernstein AB, Hing E, Moss AJ, Allen KF, Siller AB, Tiggle RB. Health Care in America: Trends in Utilization. 2003.Hyattsville, MD: National Center for Health Statistics.
43. Demiris G. The diffusion of virtual communities in health care: Concepts and challenges. Patient Educ Couns. 2006;62:178–188.
44. Lingard L, Sue-Chue-Lam C, Tait GR, Bates J, Shadd J, Schulz V. Pulling together and pulling apart: Influences of convergence and divergence on distributed healthcare teams [published online ahead of print January 23, 2017]. Adv Health Sci Educ Theory Pract. doi: 10.1007/s10459-016-9741-2.
45. Nelson S, White CF, Hodges BD, Tassone M. Interprofessional team training at the prelicensure level: A review of the literature. Acad Med. 2017;92:709–716.
46. Salas E, DiazGranados D, Klein C, et al. Does team training improve team performance? A meta-analysis. Hum Factors. 2008;50:903–933.
47. Shunk R, Dulay M, Chou CL, Janson S, O’Brien BC. Huddle-coaching: A dynamic intervention for trainees and staff to support team-based care. Acad Med. 2014;89:244–250.
48. DeChurch LA, Mesmer-Magnus JR. The cognitive underpinnings of effective teamwork: A meta-analysis. J Appl Psychol. 2010;95:32–53.
49. Johnson TE, Lee Y, Lee M, O’Connor DL, Khalil MK, Huang X. Measuring sharedness of team-related knowledge: Design and validation of a shared mental model instrument. Hum Resour Dev Int. 2007;10:437–454.
50. Mohammed S, Dumville BC. Team mental models in a team knowledge framework: Expanding theory and measurement across disciplinary boundaries. J Organ Behav. 2001;22:89–106.
51. Janis IL. Groupthink. Psychol Today. 1971;5:84–90.
52. Hart PT. Groupthink, risk-taking and recklessness: Quality of process and outcome in policy decision making. Polit Individ. 1991;1:67–90.