Abstract
As health care delivery moves toward more complex, team-based systems, the topic of medical teamwork has gained considerable attention and study across disciplines. This systematic review integrates empirical research on teamwork and health care to identify broad trends. We identified and coded 1,818 relevant, English, and peer-reviewed journal articles using a teamwork processes rubric. Several themes emerged. The health care teamwork literature has grown substantially over the past 20 years. Approximately half of the studies were descriptive (rather than interventional or psychometric); the majority relied on quantitative methods. Health care teamwork was also studied in thematically distinct manners. Interpersonal processes were most commonly studied across fields. Of all disciplines, medicine focused most on transition processes, whereas those from team science centered more highly on action processes. There were also finer grained disciplinary differences in content areas of communication and collaboration. Interprofessional journals represent a potential area for interdisciplinary efforts. Implications and future directions are discussed.
Keywords
Today’s health care systems are increasingly relying on complex, team-based methods of care delivery. Modern advances in technology, training, and organizational structures have driven changes to the way health care is performed and delivered. Indeed, researchers estimate that a single primary care physician caring for Medicare patients is linked, on average, to 229 other physicians annually (Pham, O’Malley, Bach, Saiontz-Martinez, & Schrag, 2009). These touch points are only projected to increase, given the trends evinced over time.
Accordingly, research and practices concerning teamwork processes in health care have gained considerable traction. Recent qualitative and quantitative reviews have established the importance of teamwork in effective health care (e.g., A. M. Hughes et al., 2016; Manser, 2009). However, as multiple domains of scholars, clinicians, and administrators move into the teamwork space, they have failed to do so in a uniform or collaborative fashion—resulting in highly specialized and siloed realms of knowledge, across clinical science, behavioral and social science, engineering, and other fields.
This review aims to unite interdisciplinary findings and systematically establish insights into the nature of research in health care teamwork. First, we briefly review the study of health care teamwork through theoretically and empirically based frameworks. Second, we explore the landscape of research by bibliometrically analyzing the quantity and quality of research across disciplines, including tracking its emergence and describing its expression in the scientific literature. Finally, we synthesize findings to characterize the interdisciplinary study of teamwork processes and project into the future.
Background for the Review
Health care is a critical context within which teamwork processes emerge—but this has not always been the case. The delivery structure of health care has evolved rapidly over time, replacing the solo practitioner model with complex networks of providers and professionals (Mitchell et al., 2012). In the United States, the typical Medicare patient now interfaces with an average of two primary care clinicians, five specialists, and innumerable ancillary providers annually (Bodenheimer, 2008). The increasing complexity of modern health care systems has led to the interdependence of providers and reliance upon interdisciplinary and interprofessional teams (Borrill, West, Shapiro, & Rees, 2000).
Medicine is not alone in this trend; work across many contexts has been increasingly performed by teams of individuals. This universal pattern has resulted in the rise of team science, in which the primary unit of study is the team: “a distinguishable set of two or more people who interact dynamically, interdependently, and adaptively toward a common and valued goal/objective/mission” (Salas, Dickinson, Converse, & Tannenbaum, 1992, p. 4). Researchers in this area may stem from a number of disciplines, including the social and behavioral sciences and engineering. They can further specialize in the study of teamwork, or the funneling of interdependent actions of individuals toward a common goal (Marks, Mathieu, & Zaccaro, 2001).
The quality of teamwork often depends on important precipitating factors, as well as the specific consequences being examined. To this end, team effectiveness has been conceptualized using the input–mediator–output–input (IMOI) heuristic (Ilgen, Hollenbeck, Johnson, & Jundt, 2005). Input antecedents (including at the individual, group, and environmental levels) are influenced by mediators, generating output in group performance and member reactions, which, in turn, provide cyclical feedback to inputs. Understanding these linkages enables researchers to draw conclusions about whether teams are effective, the types of teams that are effective, what they are effective at, and the conditions under which they are effective (Lemieux-Charles & McGuire, 2006). Notably, research within health care has often focused on the inputs and outputs of team effectiveness, given these factors’ more quantifiable nature and the phenomena-driven approaches of many disciplines. Below, we describe how medicine has studied these stages of teamwork.
Inputs
A large body of work has arisen on understanding and optimizing team input. For example, several reviews in health care have examined multi-level predictors of team effectiveness. Previous research has modeled task type (e.g., project, health care delivery), task features (e.g., interdependence, autonomy, guidelines/procedures), team composition (e.g., size, age, gender, tenure, discipline), and organizational context (e.g., goals/standards, structure, rewards, resources, training environment) as affecting teamwork processes, which, in turn, affect team outputs (Lemieux-Charles & McGuire, 2006). Another review further elaborated on antecedent conditions important to effective teamwork, including: clear purpose, appropriate culture, specified task, distinct roles, suitable leaders, relevant members, and adequate resources (Mickan & Rodger, 2000). For example, leadership has been explicitly identified as critical in optimizing critical care outcomes (Künzle, Kolbe, & Grote, 2010). Personality variables of team members such as conscientiousness and extraversion have also been shown to predict contextual performance in team settings (Morgeson, Reider, & Campion, 2005). Indeed, several streams of research focus on these predisposing factors of team effectiveness.
This has also naturally led to a body of work that focuses on the development and implementation of interventions, such as medical education and team trainings and simulations (e.g., Barrett, Gifford, Morey, Risser, & Salisbury, 2001; Clancy & Tornberg, 2007; A. M. Hughes et al., 2016; Manser, 2009; Tanco, Jaca, Viles, Mateo, & Santos, 2011; S. J. Weaver et al., 2010). For example, reviews of team training in health care revealed that simulation and classroom-based interventions can improve nontechnical skills (Buljac-Samardzic, Doorn, van Wijngaarden, & van Wijk, 2010), teamwork processes, and patient safety outcomes downstream (S. J. Weaver, Dy, & Rosen, 2014). These findings were further underscored by a recent meta-analysis, which demonstrated that provider team training not only positively influences trainee reactions but can also improve patients’ length of stay and mortality (A. M. Hughes et al., 2016). Altogether, this line of research aims to optimize the inputs that lead to successful teamwork in health care, including by bolstering mediating factors.
Mediators
Mediators intervene and transmit the influence of inputs to outputs, typically consisting of bonding, adapting, and learning components (Ilgen et al., 2005). The IMOI model includes emergent states and teamwork processes as mediators (Marks et al., 2001), and also addresses how these factors can interact and influence one other (Ilgen et al., 2005). This distinction between teamwork processes and emergent states, as well as the articulation of how they can affect each other, are among the many reasons for the widespread adoption of the IMOI model within the team sciences.
Emergent states
Although interest in emergent states has grown, existing research on the topic is still in its infancy (Coultas, Driskell, Burke, & Salas, 2014). Emergent states are constructs that characterize dynamic team properties and describe a team’s cognitive, affective, and motivational state (Wiggins & Crowston, 2010), such as trust, respect, and cohesion (Jehn, Greer, Levine, & Szulanski, 2008). These states have a temporal aspect, such that they evolve in a group based on the interactive teamwork processes over time (Marks et al., 2001). They both emerge and, in turn, affect teamwork processes (Seeber, Maier, & Weber, 2014). For example, trust within a team is an emergent state that is influenced by group social processes such as conflict (Jehn et al., 2008). A related concept, psychological safety (the shared belief that a team is safe for interpersonal risk taking; Edmondson, 1999), has been studied extensively in medical contexts, given highly hierarchical cultures present in medicine. Although emergent states typically relate to developing and existing conditions and cultures of teams, the other class of mediators refers to the actual behaviors that directly develop and maintain teamwork.
Processes
Teamwork processes are operationalized herein as “members’ interdependent acts that convert inputs to outputs through cognitive, verbal, and behavioral activities directed towards organizing taskwork to achieve collective goals” (Marks et al., 2001, p. 357). The difficult methodological issues and less accessible motivations surrounding the study of teamwork processes have historically resulted in a black box of factors, relatively neglected not only in medicine but also throughout teams’ research at large (Mohammed & Hamilton, 2007). However, over the past 20 years, researchers have sought to establish theoretical and empirical bases of teamwork processes. In particular, Marks et al. (2001) and LePine, Piccolo, Jackson, Mathieu, and Saul (2008) theoretically developed and empirically refined three major dimensions of teamwork processes, respectively.
Transition processes involve the actions that teams execute between performance episodes, including mission analysis, goal specification, and strategy formation (LePine et al., 2008). Given increasingly multi-team systems and rapidly changing structures in medicine, these processes must be revisited as they apply to health care contexts. For example, in one systematic review, Nagpal, Vats, Ahmed, et al. (2010) found that breakdowns in information transfer and communication in interprofessional teams often led to adverse events in surgery. However, transition processes such as planning can buffer against such errors (Castelao Fernandez, Russo, Riethmüller, & Boos, 2013). Thus, practices such as handoff and debriefing during rounding and closed-loop communication should be maximized to improve transitions between clinical performance episodes.
Action processes reflect four types of activities that occur as teams work toward their goals: monitoring progress, system monitoring, team monitoring, and coordinating (LePine et al., 2008; Marks et al., 2001). Such processes are especially critical in ensuring continuity of care (Haggerty et al., 2003) and consequent patient safety and satisfaction. For example, health care teams must monitor performance, anticipate team members’ workloads, provide support, troubleshoot difficulties, and adjust the sequence and timing of events (Fernandez, Kozlowski, Shapiro, & Salas, 2008). Literature reviews have found that communication and coordination are critical in supporting effective teamwork (Manser, 2009) and treatment outcomes (Castelao Fernandez et al., 2013) in high-stakes clinical contexts. These critical behaviors, which are deployed during performance episodes, comprise the base foundation for successful completion of tasks.
Finally, interpersonal processes involve the management of relationships between team members, requiring consideration of conflict management, motivation, and affect management (LePine et al., 2008; Marks et al., 2001). For example, studies in health care have shown that nonpunitive, inclusive, and emotionally supportive behavior can enhance organizational culture and quality improvement efforts (Edmondson, 1999; Nembhard & Edmondson, 2006). Social behaviors are especially important given the particulars of clinical settings today, including the detrimental influence of time pressure in high-stakes contexts and sociocultural shifts in the patient–provider relationship. Indeed, creating healthy interpersonal relationships raises baselines of well-being, buffers against negative events and challenging care parameters, and allows individuals to perform optimally.
In this review, we primarily focus on teamwork processes as opposed to emergent states for several reasons. First, there is a robust literature surrounding teamwork processes, resulting in structured taxonomies and arguably more nuanced understanding. Such clear operationalization is critical to a systematic review, especially one of such a large and dispersed literature. Second, some emergent states may be couched under processes and are thus captured in our review. For example, Marks et al. (2001) identified cohesion as an emergent state, given that it characterizes a property of a team rather than an interdependent team property. However, in our review of teamwork processes (e.g., coding materials from LePine et al., 2008), cohesion was so closely related to interpersonal processes that it was included in our coding rubric. We describe our translation of the teamwork processes literature into systematic search and review procedures in further detail in the “Method” section. Altogether, a focus on teamwork processes allowed us to build out a review that was theoretical, systematic, and comprehensive.
Notably, however, these preexisting organizational models have not successfully crossed the divide between behavioral science and health care disciplines. For example, in a review of critical care units, Reader, Flin, Mearns, and Cuthbertson (2009) developed their own teamwork framework of communication, leadership, coordination, and decision-making processes. Although Ilgen et al.’s (2005) formal teamwork processes taxonomy has not always been universally applied to research in medicine, its temporal bases provide a useful and relevant structure for the study thereof. Moreover, as links between teamwork and its outputs (i.e., performance and health outcomes) become stronger, adoption of the theory- and evidence-based science behind its mechanisms will likely follow suit.
Outputs
Team outputs refer to “team outcomes associated with productivity, as well as to the capability of team members to continue working cooperatively” (Barrick, Stewart, Neubert, & Mount, 1998, p. 377). They can include consequences such as individual and group performance, attitudinal changes, organizational effectiveness, and financial costs. Team outputs are often multi-dimensional and poorly conceptualized, such that comparisons across studies are difficult (Lemieux-Charles & McGuire, 2006). However, a review of health care team literature identified three levels of objective outcomes related to team effectiveness: patient (e.g., functional status, satisfaction), team (e.g., clinical quality of care), and organization (cost-effectiveness; Lemieux-Charles & McGuire, 2006).
In clinical settings, patient health outcomes are typically the most immediate and tangible barometer of success. Thus, medicine has often based teamwork-related research on studies that directly concern output-oriented dependent variables (e.g., patient safety, procedural success; Baker, Gustafson, Beaubien, Salas, & Barach, 2005; Barrett et al., 2001; Smith & Cole, 2009; S. J. Weaver et al., 2010). For example, in a review of dynamic domains of health care (i.e., operating rooms, intensive care, emergency medicine, and trauma and resuscitation teams), Manser (2009) found that teamwork plays an important role in the causation and prevention of adverse events, including through its impact on provider well-being. Furthermore, meta-analytic results also indicated that rapid response teamwork in hospitals is associated with a reduction in rates of cardiopulmonary arrest outside the intensive care unit for adults and children (Chan, Jain, Nallmothu, Berg, & Sasson, 2010). Given outcome-based findings such as these, the research surrounding the importance and effectiveness of teams is highly emphasized within health care.
Notably, outputs have a synergistic, looping relationship with teamwork inputs. The IMOI model invokes the notion of a cyclic causal feedback to explain how outputs, in turn, influence inputs (Ilgen et al., 2005). For example, team performance, typically viewed as an output, can be conceptualized as an input to future teamwork processes and emergent states. Indeed, previous research has found that performance feedback contributes to effective learning and group decision making (Mo & Xie, 2010).
The Current Review
Overall, researchers have made substantial progress investigating how teamwork operates in health care, adding to our understanding of specific team inputs, mediators, and outputs. However, this abundance of research has also become a double-edged sword in certain regards. There are many diverse perspectives on teamwork that have yet to be meaningfully synthesized. This review responds to this need in the literature, making a contribution in three ways.
First, this review uses a strong theoretical basis, derived from the team sciences, to systematically understand the study of health care teamwork processes. Specifically, it draws upon Marks et al.’s (2001) and LePine et al.’s (2008) model to parse various teamwork behaviors. Previous research collaborations have mapped these theories onto medical contexts, although they have either focused on specific teamwork processes or particular environments. For example, Fernandez et al. (2008) used teamwork processes models to describe teamwork within emergency medicine. Benzer and colleagues (2016; Benzer et al., 2015) have also used Marks et al.’s (2001) approach to thematically analyze primary care team performance over time. In contrast, we use the teamwork processes model beyond a single medical setting, to broadly survey its applications in health care.
This touches upon a second point. This review examines health care teamwork across disciplines. To our knowledge, no interdisciplinary reviews of the holistic teamwork processes underlying health care effectiveness currently exist. Researchers have argued convincingly why teamwork is important in health care. For example, a systematic review of teamwork processes in health care found that, in each of the 28 included studies, teamwork processes were significantly connected to at least one intervention or performance variable, with typically medium to high effect sizes (Schmutz & Manser, 2013). However, less is understood about how teamwork as a whole is being studied, particularly when there are many scientists, clinicians, administrators, and policymakers in the space. Indeed, previous works synthesizing these relationships have been smaller in scope (e.g., by only considering articles focused on patient safety from the medical literature; Schmutz & Manser, 2013). By meta-cognitively studying and uniting the dispersed literatures, this review seeks to summarize the state of health care teamwork research both within medicine, where discrepancies exist between specialties, and across disciplines including basic science, education, human factors, and management.
Finally, this review is conducted with an eye toward areas of potential contribution for team scientists. There remains work to be done in terms of combining cross-disciplinary insights on teamwork processes. For example, in an empirical review of shared knowledge structures within a team, Burtscher and Manser (2012) found that health care research had not yet studied team mental models despite its relevance for clinical care. This suggests that theory and innovation from scholarly and interdisciplinary literatures do not always fully translate into medical realities. In conducting this comprehensive review, we have paid special attention to identifying similar opportunities—clarifying ways in which health care teamwork can be aided and abetted by more collaborative and translational efforts.
The health care setting warrants unique consideration as it is characterized by a number of unique features that do not necessarily converge in other settings. As previously mentioned, medicine has moved from historically independent practices to team-based care; thus, the field is still learning and incorporating lessons from teamwork science. In comparison, other team-dependent industries such as aviation and the military have a longer and deeper investment in teamwork research, driving the science given their reliance on crews (Burke, Salas, Wilson-Donnelly, & Priest, 2004). For example, Burke et al. (2004) emphasize the extensive research on, and practice of, closed-loop communication in the military, which is still being applied to medical settings (e.g., El-Shafy et al., 2018).
Despite its relatively recent emergence in medicine, however, teamwork has been quickly adopted as a critical characteristic of care; professionals are expected to collaborate effectively in multi-disciplinary teams and multi-team systems today (Fiscella & McDaniel, 2018; Rosen et al., 2018). Moreover, patient-centered care has come to the forefront in medicine, in which a team’s “central member is the patient” (Davis, Schoenbaum, & Audet, 2005, p. 956). The engagement of patients and families in care also inherently introduces unpredictable expectations, roles, and power dynamics into team functioning. Thus, relative to other contexts, the historical precedents and current trends in medicine have created idiosyncratic conditions under which teamwork emerges.
Furthermore, the parameters of medical specialties can differ significantly and influence the expression of teamwork. For example, in the intensive care unit, teams work in emotionally and physically challenging environments, where temporal instability often impedes team dynamics (Ervin, Kahn, Cohen, & Weingart, 2018); stakes may be lower and patient care less demanding in lower acuity settings. Unlike in other highly structured industries, teamwork principles must consequently be adapted to the needs of the medical context. Thus, health care in teamwork is distinguished by its growing adoption in practice, the unique interaction between patients and providers, and the diverse factors that characterize subfields. Although similar features may also exist in other contexts, the extent to which they coincide within health care emphasizes the importance of teamwork therein and underscores the value of understanding how teamwork processes unfold in the health care context.
To this end, we conducted a systematic, integrative, and bibliometric review of the teamwork and health care literatures. We sought to answer the question: How do different disciplines discuss health care? Given that we elicit many voices from diverse fields, it was necessary to establish a common language in discussing teamwork processes. Ilgen et al.’s (2005) synergistic IMOI model served as an organizing framework through which teamwork was investigated. Given that we examined multi-disciplinary literatures, their theory provided a lens through which we were able to observe and describe teamwork processes.
Method
We first identified peer-reviewed journal articles that discuss teamwork processes in health care contexts. In April 2018, we conducted a literature search across several interdisciplinary academic databases, including Academic Source Complete, Business Source Premier, PsycINFO, and MEDLINE. We used a Boolean operator keyword search of the titles and abstracts of articles, using the term teamwork with lexemes representing health care (health care, medicine, clinical). We used wildcard operators to ensure that the search engines captured all potentially related articles. This returned 7,975 articles for review.
Following this, we reviewed abstracts for possible inclusion based on thematic relevance and empiricism. First, the work had to be a peer-refereed journal article written in English. Second, the article had to examine health care contexts. Health care was herein identified as any environment in which clinical care is delivered, including all domains of medicine, dentistry, optometry, psychology, nursing, and other allied health and health care professions. We also included environments within which medical providers were professionally trained (e.g., medical school, simulation centers), given that these are associated with downstream provision of care. Third, the study had to collect primary data. Finally, the article had to relate to teamwork, as it was defined earlier in the “Background for the Review”. Specifically, the work had to discuss teamwork as a central topic, such that it was an independent or dependent variable (for a quantitative study) or primary research question or outcome (in qualitative work).
The authors then coded each article to identify and synthesize themes throughout the literature, applying two types of analyses. First, we bibliometrically reviewed the state of the literatures, given our goal of characterizing the emergence of health care teamwork research across disciplines. Specifically, one of the authors coded each document based on the following variables: publication year, publication discipline, study approach (i.e., descriptive studies, trainings and simulations for health care providers, interventions at the organizational level, psychometric or measure evaluation studies, or descriptive/none of the above), and data collection method (i.e., qualitative, quantitative, or mixed methods/hybrid). Following this, 15% of total articles were double-coded by another author to ensure consistency of analyses, with an agreement rate of 93.4%. Any discrepancies were resolved through additional review of materials.
Publication discipline was coded independently of the other variables and conducted in two passes. The initial round of coding translated the discipline as closely to the journal topic as possible; following that, we reviewed all the codes and subsumed them into thematic categories based on mutual content area and critical mass of constituent articles. Journals that ostensibly fit into two disciplines were sorted by the more specialized category; that is, pediatric critical care and intensive care nursing studies would both be sorted into the same acute care bin. In our review, we refer to publication discipline when discussing discipline at large, given that the research outlet directly relates to the researchers’ and readers’ field of expertise. Alongside the study approach and data collection methods, these markers were used to describe how various disciplines have studied teamwork in health care.
Second, we used software to analyze qualitative content areas related to teamwork across disciplines. This began with the development of a rubric of teamwork key terms. We drew from the language in LePine and colleagues’ (2008) meta-analytic model article and coding guide to generate key terms under Marks et al.’s (2001) categories of transition, action, and interpersonal processes. Herein, we report the key terms as complete words, but used lexemes to capture all potential forms in our search; for example, we applied the search phrases “hand-o” or “hando” to identify all occurrences of words related to handoff and handover. Transition lexemes included adapt, debrief, handoff, handover, huddle, reflexivity, and strategy. Action lexemes included monitor, coordinate, support, and backup. Interpersonal lexemes included affect, attitude, cohesion, conflict, cooperation, emotion, and motivation. If an article abstract included one or more key terms within a category, they were coded positively for that teamwork process.
Notably, there are instances in which a single teamwork-related term can fit multiple processes, given that teamwork dynamics do not exist in isolation but interact with each other synergistically. We thus included relevant key terms independent of teamwork process categories, generated from the literature and the current review’s documents themselves. In the case of the former, we reviewed LePine et al.’s (2008) model for terms that could not be assigned to a specific teamwork process. For the latter process, we loaded the titles, abstracts, and keywords of each article into ATLAS.ti qualitative analysis software (Version 8, Scientific Software Development, 2018). We then used its native word counting tool, which identified the most frequently occurring words across abstracts, and selected teamwork-related terms for inclusion in our keyword rubric. Ultimately, the key terms generated by these two procedures included communication and collaboration.
Once the rubric was established, we loaded the abstracts into Microsoft Excel. We then used a combination of functions to identify and code the presence of key terms within each article abstract. That is, software formulas were used to scan each abstract’s text and detect whether or not it included any key terms (i.e., whether the abstracts used the language of transition, action, interpersonal processes, and/or other teamwork-related content areas). The program dummy-coded abstracts for the presence of key terms within the appropriate teamwork process or content area. The authors then manually reviewed all coding to ensure comprehensive and accurate application; for example, codes were removed if they were not related to a teamwork variable in the study or added if they had not been detected in an abstract due to permutations of language. After data cleaning, we were then able to parse apart results by discipline.
This computer-assisted method allowed us to quantify qualitative results, given that we could generate numerical indices of how frequently teamwork processes and content areas emerged. We visualized results using a number of methods, generating line and bar graphs in Microsoft Excel and ternary plots with Hamilton’s (2018) ggtern and Wickham’s (2016) ggplot2 packages in the software program R. Finally, we used these results to guide closer readings of the articles. This allowed us to generate deeper insights into the literature.
Results
We found 1,818 articles eligible for inclusion, as shown in Figure 1; these references are available upon request from the first author. Figure 2 illustrates the total output of health care teamwork research from 1981 (e.g., the publication of the first article eligible for this study) through April 2018, by discipline. Medical journals have published more than any other field on teamwork in health care, constituting over half of literature in the area (50.3%). This was followed by health care services (16.8%) and interprofessional (12.3%) disciplines. Given the substantial proportions of literature derived across areas, they were further broken down into subclusters based on content area and volume.
Article inclusion chart.
Line graph of volume of empirical research articles related to health care teamwork, by discipline.
Medicine was partitioned into acute care, including emergency medicine and critical care; general medicine, for clinical journals with a broad focus; primary care, including family practice, obstetrics and gynecology, and general pediatrics and geriatrics; surgery, including all surgical subspecialties; medical specialties, for those consultant journals that did not fit into the previous categories; and medical education, for journals that center on training providers. Health care services was divided into three streams: general health care management, public health, and quality improvement, including patient safety efforts. The interprofessional literature was bisected into allied health, for those journals that were specifically devoted to dentistry, pharmacy, social work, or other professions distinct from medicine and nursing; and explicitly interprofessional publications, focusing more generally on the integration of collaboration of health and social care providers. The social sciences were categorized into education and evaluation, human factors and informatics, management and psychology, and general social science, including communications, sociology, and other fields not captured in the aforementioned groups. Nursing and basic science remained standalone fields, given their distinct content areas. The remaining results will be reported using both the original disciplinary clusters and these more detailed subcategories.
Once these finer grained distinctions were made, the greatest contributors to the health care teamwork literature specifically have been medical education, nursing, primary care, and nonsurgical medical specialty journals. The supplemental material includes a table summarizing articles by discipline and variables of interest.
Given that we examined the empirically based contributions of each discipline to the teamwork and health care oeuvre, we also parsed apart results by scientific study design. Specifically, we looked at the approaches of studies and data collection methods most often used by different disciplines. In this section, we quantify and characterize (a) the trends across literature and (b) differences of output in terms of proportion within each discipline. We examine and describe these findings in greater depth in the “Discussion” section.
Study Approach of Reviewed Articles
As shown in Figure 3, nearly half (49.9%) of the reviewed studies were descriptive in nature, meaning that they explored currently existing phenomena (e.g., provider attitudes toward collaboration and perceptions of patient safety principles). In particular, a disproportionately large portion of the team science disciplines (63.8%) focused on this area.
Study approach by discipline.
Approximately one third (34.0%) of articles tested trainings or simulations administered directly to health care providers. Perhaps unsurprisingly, medical education had the greatest proportion of studies dedicated to examining trainings (63.3%).
Of all articles, 10.7% examined organizational interventions, such as the implementation of surgical checklists or changes to patient flow procedures. Here, human factors and informatics and quality improvement fields devoted a greater-than-average proportion of their research to organizational interventions: 22.0% and 20.5%, respectively.
Finally, the smallest percentage of studies (5.3%) were validation studies, evaluating methods of measurement related to health care and teamwork, with fields comprising at most 15.6% and 11.0% of their publications to psychometrics (education and evaluation and health care management, respectively).
Data Collection and Analysis Methods
In Figure 4, a ternary plot demonstrates each discipline’s relative use of quantitative, qualitative, and mixed methods. That is, each point represents a discipline; the proximity of the point to a corner represents the degree to which it incorporates that specific method, relative to the other two methods (corners). As a whole, the health care teamwork literature leans on numeric data collection (53.2%); this is followed by qualitative (32.8%) and mixed methods (13.9%). This pattern is observed, more or less, across most disciplines.
Ternary plot of study methods by discipline.
However, quantitative data were collected at a relatively higher rate by quality improvement (67.8%) and allied health (67.3%) disciplines; in the case of the former, this also resulted in a lower rate of qualitative research (18.5%) compared with other fields. Conversely, 64.7% of general social sciences articles utilized qualitative methods. Mixed-methods and hybrid designs were also used in a comparatively large percentage of public health (21.9%) and interprofessional (19.7%) articles, while being less applied in human factors and informatics (7.3%) and basic science (6.9%) fields.
Teamwork Processes in Health Care
Next, we examined how teamwork is being studied within the separate literatures, as shown in Figure 5. As in the previous ternary plot, the location of each discipline’s point, relative to the three corners, represents how that discipline distributes its focus among the three processes. However, each discipline also discusses teamwork processes at different overall rates; we thus represent this through size of point. For example, the basic sciences discussed teamwork processes in only a fraction of their articles. As a result, its point on the plot is relatively small, corresponding to 17.2% of the maximum size. Within the basic sciences, the largest percentage of articles (10.3% of all basic science articles) mention transition, as opposed to action or interpersonal (0.0% and 6.9%, respectively), processes; thus, the point’s location on the ternary plot places it closer to the transition corner than any other. Through location and size of point, this ternary plot thus provides us with a summary-at-a-glance of teamwork processes by discipline.
Ternary plot of teamwork processes by discipline.
Overall, 19.3% of articles discussed transition processes. Of the disciplines reviewed, two groups in particular studied transition processes at relatively high proportions: acute care services (i.e., critical care and emergency medicine; 37.5%) and surgical specialties (31.2%). Moreover, of all the key terms related to transition processes, strategizing was the most frequently mentioned (55.7%), followed by debriefing (24.9%) and adapting (14.0%).
As a whole, action processes were mentioned by 24.5% of articles. Nonmedical fields set themselves apart from medicine by studying these processes more exclusively. In particular, management and psychology journals studied action at a relatively high rate (45.9%). In order, the most cited action processes were supporting (60.3%), coordinating (26%), and monitoring (12.2%) behaviors.
Of all reviewed articles, 35.6% touched upon interpersonal processes. Here, several trends were apparent. Interprofessional and allied health fields researched interpersonal processes at much greater rates of 60.7% and 59.4%, relative to other fields. We also observe substantive contributions from the social sciences, including general social sciences (52.9%) and education and evaluation (50%). Across the board, attitudes were the most frequently mentioned (52.0%) dimension of interpersonal processes; cooperation (16.2%) and conflict (9.9%) were also discussed.
Teamwork Content Areas
In our last set of analyses, we explored empirically based content areas related to health care teamwork, as shown in Figure 6: communication and collaboration. Communication was cited by 44.2% of all journals. Among them, surgical articles mentioned communication at the highest rate, 60.2%, followed by education and evaluation (59.4%) and quality improvement (56.2%). Interestingly, management and psychology journals had the lowest rate of communication citations (18.9%).
Percentage of articles studying communication and collaboration by discipline.
About a quarter of journals, 24.5%, identified collaboration in health care teams, with interprofessional (54.9%) and general social science (52.9%) journals devoting the major share of studies to such research. The lowest proportion of articles studying collaboration was observed in human factors and informatics (2.4%) and basic science (6.9%) journals.
Discussion
This systematic review gives a high-level view of the health care teamwork literature. Below, we describe several trends that emerged from this literature. Overall, our findings suggest that, despite health care teamwork being an inherently interdisciplinary area, there is relatively little cross pollination in this area. To this end, we highlight ways in which team scientists, in particular, can bolster the science in this area, given the demonstrable demand for greater teamwork expertise in medicine.
Trends in the Greater Health Care Teamwork Literature
Over the past 20 years, there has been a marked increase in teamwork literature across disciplines. This reflects what Mathieu, Hollenbeck, van Knippenberg, and Ilgen (2017) characterized as a “marked upward curve” (p. 453) in their review of the greater teams literature. Interestingly, although they observe the beginning of the said curve in the 1990s, our study found a slightly delayed uptick in health care–specific literature starting in the early 2000s. This time period coincides with the publication of two notable pieces in the separate health care and teamwork literatures. First, The Institute of Medicine (2000) released their landmark report, To Err Is Human, in which the authors recommended improving teamwork to curb the overwhelming number of fatal medical errors in the United States. Shortly thereafter, Marks et al.’s (2001) seminal piece on teamwork processes was published, marking a moment wherein teams’ theory had formalized and progressed to the point of taxonomy development. These two publications represent how clinicians and academicians simultaneously arrived at similar conclusions on the essentialness of teamwork in health care.
Medical journals have been the primary contributor to the health care teamwork literature, both over time and in terms of quantity. Indeed, in a recent review on foundational teamwork literature, Driskell, Salas, and Driskell (2018) note that medicine has specifically embraced the importance of group research. However, there have also been substantive contributions from a variety of disciplines, particularly the fields comprising team science. We frame much of the discussion based on this dichotomization of health care teamwork; that is, juxtaposing research from medicine versus team science. The historic Institute of Medicine (2000) and Marks et al. (2001) papers themselves demonstrate the dual spheres of influence that have driven research in health care teamwork.
Given the number of multi-disciplinary approaches, we observe diversity in conceptualizations of health care teamwork. Teamwork in health care has been studied and subsumed under numerous guises across disciplines: crew resource management, interprofessional collaboration, nontechnical skills, patient safety culture, problem-based learning competencies, safety-relevant topics (e.g., patient safety, safety climate attitudes), and so on. This area is examined as a health services issue and studied alongside critical outcomes such as patient safety. The rise of the interprofessional literature itself suggests that there is increasing interest in the conceptualization and practice of group-based care in medicine.
Importantly, the relatively small and stable volume of articles derived from the team sciences and other nonmedical concentrations suggest that the goal of interdisciplinary research has not yet been fully realized. As described in the “Background for the Review” section, team science can refer to research across a number of disciplines, though we found that social sciences were the primary nonmedical contributors in the health care teamwork space. Regardless, medicine’s demand for teamwork research appears to be increasing, yet is unmet in scale by the team sciences. These findings may be driven by several possible phenomena.
First, we consider the issue of journal publication. That is, team scientists may be placing their health care teamwork manuscripts in medical journals rather than their own disciplinary publications. However, our review of these literatures suggests that this does not satisfactorily explain the disparate rates of health care teamwork study publication between medical and social science journals. We randomly sampled 100 articles within the medical disciplines and found that a majority did not include social scientists as first authors, nor did they include author groups with any nonmedical institution affiliations. However, we did see evidence for collaboration. Author lists often included individuals with various types of disciplinary training and degrees. Nonetheless, our analyses suggest that lead researchers often published in journals within their own fields; that is, scholars in medicine are driving a large volume of investigations on health care teamwork. Overall, there appears to be ample opportunity for multi-disciplinary expertise in the development, implementation, and evaluation of health care teamwork studies.
Notably, however, closer readings revealed that a substantial number of medical articles did draw upon teamwork theory from the team sciences, as evidenced by their bibliographies. This brings us to a second important, potential explanatory factor. Team scientists may focus more heavily on producing theory than medical researchers when studying health care teamwork. Given our goal of understanding the empirical landscape of health care teamwork, our systematic review captured only articles that collected and analyzed data. As a result, this may have excluded the substantive theoretical and review-based contributions from the team sciences in the health care teamwork space. Relatedly, we note that the sole quantity of articles published does not necessarily translate to a greater overall impact. That is, the team sciences may not be as highly represented in terms of total output of empirically derived publications, but their contribution to theory and knowledge should not be undersold. We discuss this further in the “Limitations and Future Directions” section.
Ultimately, this suggests that team scientists have unique insight into the health care teamwork space, yet are underutilized in the actual empirical deployment of such research. To advance the collective research in this area, team researchers can draw upon how health care teamwork is being studied in other disciplines to identify potential areas of contribution. In the following sections, we identify distinguishing characteristics of the literature that can help foster and strategize these partnerships.
Trends in study approach
There are diverse study designs and data collection methods being employed in the empiricism of health care teamwork. Across disciplines, researchers have favored observational research, such that anywhere between 25% and 80% of studies within sub-disciplines (e.g., acute care and primary care within medicine) are descriptive. These designs are particularly prevalent among the team sciences, comprising greater than 80% of the general social sciences and more than 70% of management and psychology. Some examples of descriptive studies within the team science disciplines include those that identified team behaviors that affect simulation performance (Shetty, Cohen, Patel, & Patel, 2009) and evidence-based intervention design (Aveling, Stone, Sundt, Wright, & Singer, 2016). Conversely, studies from medicine have investigated coordination patterns related to high performance in medical teams (Manser, Harrison, Gaba, & Howard, 2009) and identified factors causing stress in the operating room (Sami et al., 2012). These examples suggest that researchers are examining the impact of teamwork at a number of levels within health care.
Notably, medicine has moved toward incorporating and evaluating teamwork through training. This is true not only in the pertinent medical education journals but also in those of allied health providers, interprofessional providers, and surgeons. Indeed, a recent meta-analysis identified 129 studies on team training in health care (A. M. Hughes et al., 2016), conveying the investment of medical educators and administrators in practical application. Moreover, 41 articles in this review mentioned the use of popular team-based training TeamSTEPPS™, which is the product of a large-scale collaboration between the Agency for Healthcare Research and Quality, the Department of Defense, and numerous interdisciplinary organizations (Castner, Foltz-Ramos, Schwartz, & Ceravolo, 2012; Cooke, 2016; Keebler et al., 2014; A. L. Weaver, Hernandez, & Olson, 2017). These findings underscore the demand for teamwork; the need for its immediate translation into practice may be greater than previously understood.
In this regard, we turn to ways in which team scientists can more fully contribute to health care teamwork—not only being a part of the conversation but by informing its practice. In addition to individual training, scientist–practitioners can also focus on systems-level changes. In particular, human factors scientists have contributed an outsized proportion of its health care teamwork output in the area of organizational interventions. For example, human factors intervention studies found that a telemedical robot increased task-based communication and did not negatively affect team trust, psychological safety, or transactive memory systems during rounds (Lazzara et al., 2015). Studies within this discipline have also found that resuscitation teamwork is improved with the presence of situation displays (Parush et al., 2017). Indeed, there are several mechanisms, including novel engineering solutions, that can help boost teamwork in the medical domain.
Finally, we observe that health care management has examined measurement at greater rates, relative to other fields. Studies in this field include those that evaluate the psychometric properties of safety culture and climate surveys, such as the Hospital Survey on Patient Safety Culture (Bodur & Filiz, 2010) and the Safety Attitudes Questionnaire (Bondevik, Hofoss, Hansen, & Deilkås, 2014), as well as surveys relating to burnout (Profit et al., 2014), work–life climate (Sexton et al., 2017), and informational role self-efficacy (Chiocchio, Lebel, & Dubé, 2016). In addition, health care management studies have contributed toward developing reliable and valid checklists for documenting team and collaborative behaviors (Henneman, Kleppel, & Hinchey, 2013). This focus on validation reflects the inherent need of health care management to clearly assess outcomes. This trend, taken with the other aforementioned aspects of study design, suggest that researchers can and should turn their attention from describing phenomena to applying it to real-world issues.
Trends in data collection and methods
We found that the health care teamwork literature has been primarily quantitative, as evidenced by the cluster of disciplines near the quantitative axis of the ternary plot in Figure 4. However, the general social sciences form an outlier, with a greater proportion of its articles devoted to qualitative and mixed-methods studies. Indeed, the other social science disciplines (education and evaluation, human factors and informatics, and management and psychology) also skew slightly more toward non-quantitative methods, relative to other fields. This aligns with the previous characterization of social sciences articles as more theory-driven, given that qualitative research is a basis for theory building (Charmaz, 2006).
This also suggests two strategies for productive methodological collaboration in the health care teamwork space. First, team scientists can augment quantitative investigations with rigorous qualitative methods of collection and analysis. For example, a large number of medical studies used numeric indices to measure the utility of trainings (e.g., Gardner, Kosemund, & Martinez, 2017; K. M. Hughes et al., 2014; Nielsen et al., 2007). Future research in this space could also incorporate qualitative evaluation components, including observations, interviews, or free response survey items. Such data, provided their rigorous collection and analysis, would provide deeper understanding of health care team training outcomes. This would also boost the number of mixed-methods studies in this area, of which there were relatively few.
Conversely, team scientists can build off of their existing qualitative work and develop and implement more quantitative studies. This would be particularly beneficial, given the reliance of other disciplines on quantitative data. For example, many studies published in the social sciences used qualitative methods to describe the emergence of teamwork processes within health care teams (e.g., Anonson et al., 2009; Hilligoss, 2014; Lohuis, Sools, van Vuuren, & Bohlmeijer, 2016). Future studies on these topics can further explore findings through quantitative markers, measuring dimensions of qualitative codes or developing measures around emergent themes.
Importantly, both of these recommendations assume that there is continuity within the health care teamwork space. Indeed, given that research is an iterative process, studies should build off of prior work. However, in our review, we found an often fragmented body of work on health care teamwork, with a lack of strong unifying elements. Thus, we suggest that researchers draw upon the full extent of the health care teamwork literature when designing research. There is a marked need for studies that replicate or continue existing lines of work. Conversely, innovative projects should incorporate preceding health care teamwork theory. In the remaining subsections, we attempt to guide this process by describing content areas studied in the health care teamwork literature.
Trends in Teamwork Processes
Our systematic review identified clear trends within the content areas of the disciplines studying health care and teamwork. Overall, over a third of studies included some component of interpersonal processes, making affective and motivational dynamics the most common operationalization of teamwork. Moreover, applying our rubric of teamwork processes uncovered several notable contrasts between fields. In broad strokes, we were able to identify how specific teamwork processes were dominated and disproportionately studied by particular fields.
Trends in transition processes
Of all the disciplines, medical subfields leaned most toward transition processes. These include practices of strategizing, as well as interventions directly targeting debriefing and huddling (e.g., Krych et al., 2005; Martin & Ciurzynski, 2015). In particular, medical education and surgery journals published a large proportion of articles specifically on debriefing; this not only suggests that such strategies are critical to particular specialty environments, but that it is also being applied increasingly as part of provider training. Indeed, medical research has identified the criticality of transitions of care (i.e., when two or more workers exchange mission-specific information, responsibility, and authority for an operation; e.g., Lardner, 1996). Given the high rate of issues arising from shift changes and handoff (e.g., Cheung et al., 2010; Joint Commission, 2002), the Agency for Healthcare Research and Quality has prioritized research on transitions of care (Clancy, 2006). In this regard, health care has already taken steps to look to interdisciplinary research. For example, medical researchers previously identified how aviation practices, as developed by human factors, can contribute to improved handoffs (e.g., Cheung et al., 2010; Lyndon, 2006). Team scientists working within health care have also singled out debriefing as a particularly effective group behavior (Salas, Zajac, & Marlow, 2018). However, our review suggests that there is still room for continued collaboration in this area.
Interestingly, we observed a shared interdisciplinary focus in the transition process of strategy. Specifically, interprofessional and social science journals both published a relatively high volume of work that optimizes health care teamwork through planning behaviors. For example, Shaw, Kearney, Glenns, and McKay (2016) examined palliative care teams in an interprofessional journal, whereas Solheim, McElmurry, and Kim (2007) looked to multi-disciplinary primary health care practice through an applied social science lens. This is only one example of a larger overlap between interprofessionalism and the social sciences; we return to and discuss this trend and its implications later in the section.
Trends in action processes
Conversely, the social sciences have focused relatively highly on action processes. In particular, management and psychology journals discussed performance episode behaviors at a strikingly high proportion, compared with other fields. For example, one organizational psychology study used quantitative data to understand backup behaviors and other processes that buffer against the negative effects of job demands on quality of care (Costa et al., 2014). This emphasis on action processes thus highlights an area in which scholars in management and psychology can apply their expertise and contribute to health care teamwork.
There may be several reasons why medicine has focused relatively highly on transition, whereas management and psychology have centered more on action processes. Transition processes have successfully translated into several concrete practices as previously mentioned (e.g., debriefs, huddles). This is, in part, due to the fact that transition episodes have temporal space during which recommended behaviors can be implemented. Conversely, action processes appear to be relatively less applicable to the health care context. One potential explanation for this could be that the language and operationalization of action processes has not been employed by clinical science. That is, medical researchers and practitioners do explore teamwork behaviors during performance episodes but do not use the same terms (e.g., monitoring, coordinating, supporting).
Another reason could be that the action processes may be difficult to distill into clear and specific behaviors. For example, backing up and coordination are essential components of teamwork but can be comprised of many different practices. However, researchers have developed exercises, such as those involving closed-loop communication, that can enhance action processes (e.g., El-Shafy et al., 2018). The identification of such practices does not address the second barrier to action process implementation—that is, real-time, real-world execution. Given the stakes and unpredictability of health care environments, it can often be challenging to implement or evaluate action process–based behaviors during actual performance episodes. In light of these characteristics, as well as the fact that transition episodes are especially sensitive periods in the delivery of care, medical researchers and clinicians may have shifted their attention away from action to transition processes. However, team scientists may serve the field of health care teamwork by bolstering action processes, developing more distinct and feasible behaviors.
Interestingly, when we look at how each discipline studied specific action processes, we observe fine-grained differences. In particular, medicine considered monitoring behaviors at higher proportions than did the team sciences. For example, surgical studies examined monitoring as it related to situations (Capella et al., 2010), safety (Leeper et al., 2018), and across the team (Paige et al., 2009). However, among team sciences, scant articles studied these behaviors as they were empirically applied to the workplace (e.g., only Sonesh et al., 2015, and Burtscher, Kolbe, Wacker, & Manser, 2011, empirically examined situation monitoring). Rather, the team sciences focused more on coordinating and support behaviors (e.g., Chiocchio, Rabbat, & Lebel, 2015; Ogbonnaya, Tillman, & Gonzalez, 2018). These findings suggest two potential ways for team scientists to apply their research and further health care teamwork. First, they could study variables that incorporate monitoring behaviors. Second, team scientists can connect coordinating and support behaviors to others that are currently prioritized in health care. Hearkening back to the synergism of the IMOI model, a behavior such as coordination can help bolster teamwork processes that occur during transition periods and mutually strengthen monitoring during performance episodes.
Trends in interpersonal processes
Finally, interpersonal processes were studied at proportionally high rates by the interprofessional sciences, including both allied health and interprofessional subfields. This finding emphasizes that successfully crossing medical specialty and practice boundaries requires the bolstering of affective and motivational aspects of team function. Among these were both interventions aimed at enhancing values of openness (e.g., James, Page, & Sprague, 2016; Weinstein et al., 2018) and ethnographic studies describing existing attitudes (e.g., Goldman, Kitto, & Reeves, 2017; Hoffman & Redman-Bentley, 2012). As health care delivery continues to incorporate other consultants, specialists, providers, and ancillary staff in team-based care, researchers and administrators should ensure that attitudinal competencies are developed.
Interestingly, the general social sciences matches interprofessional journals in terms of its connection to interpersonal processes, with over half of its articles exploring these themes. This includes studies that directly investigated team cooperation and cohesion in hospitals and other medical settings (e.g., Wang, Chen, Lin, & Hsu, 2010). This overlap in content areas is thus a second example of how interprofessional and team science research can mutually benefit from collaboration—specifically on the emotional aspects of teamwork.
In terms of specific behaviors within interpersonal processes, attitudes were most frequently studied across nearly all disciplines. However, the team sciences focused relatively highly on cooperation. For example, Shetach and Marcus (2015) quantified its effects among medical and nursing teams; Lewin and Reeves (2011) also discussed cooperation as it is impacted by impression management within the health care team. Team scientists are thus able to apply strong methods and theories to systematically understand the emergence and maintenance of these interpersonal processes.
Finally, our review using the teamwork processes rubric also characterized certain disciplines in self-evident manners. The basic sciences and human factors and informatics often cited teamwork-related key terms the least, aligning with both fields’ emphasis on more technical aspects of group functioning. Although these findings may seem intuitive, they had not yet been thoroughly quantified or substantiated. Moreover, despite the fact that these disciplines may not be studying specific teamwork processes in great detail, their inclusion still suggests that health care teamwork is still emerging as an important context of study across fields.
Trends in teamwork content areas
Our review revealed differences in the content areas studied in the medical versus team sciences literatures. Although communication and collaboration are vital to the health care teamwork literature, the emphasis on these processes in medicine is not reflected in team science journals. Indeed, the discrepancy between the team science and medical literature is highest for medical sub-disciplines that rely most heavily on these processes. In fields like surgery, where coordination is paramount, communication processes are discussed thoroughly and have been linked to patient morbidity outcomes (Davenport, Henderson, Mosca, Khuri, & Mentzer, 2007). Similarly, collaboration seems to be of particular interest in interprofessional journals, where understanding how individuals interact across specialties is especially germane.
Ironically, among team scientists, management scholars and organizational psychologists contribute the least to these areas of inquiry. Team scientists, and particularly those interested in studying organizations, have an opportunity to bridge this gap by providing insights from our broad base of research on communication best practices. In medicine, the discussion of communication and collaboration often ends at TeamSTEPPS™ and debriefing (Salas et al., 2008). However, researchers in the team sciences have advanced our understanding of communication best practices that could be instrumental in the health care sphere (see Marlow, Lacerenza, Paoletti, Burke, & Salas, 2018, for a meta-analytic review).
Relatedly, our data suggest that the onus on studying communication and collaboration has fallen on nursing and interprofessional investigators, who produce the bulk of research in these areas. This suggests two paths forward. First, we recognize that these skills are critical not only to these particular populations; communication and collaboration are crucial to the team function of nearly all frontline providers in and across the health care spectrum. Team scientists have an opportunity to drive the conversation around these skills in other medical disciplines.
Second, team scientists may also choose to focus more closely on interprofessionalism as an outlet for partnership and publication. As a whole, our results suggest that interprofessional research has strongly contributed to work on teamwork processes and content areas. Indeed, the discipline itself represents a meeting ground where medicine encounters team science. Inherently, teamwork is essential to the performance of nurses and allied health professionals, as well as the delivery of care under the systems that employ them. As a result, team scientists can and should look to the discipline of interprofessionalism to maximize impact. Our review alone identified two areas of overlap between team science and interprofessionalism: (a) their mutual study of strategizing within transition processes and (b) an overall emphasis on interpersonal processes. Thus, researchers from these domains should look to one another, both in terms of building collaborations and disseminating findings in interdisciplinary journals and outlets.
Overall, health care teamwork is characterized by a number of interesting trends, many of which distinguish it from team research in other domains. For example, research on teamwork in military teams has often involved action processes such as monitoring, backup, feedback, communication, and coordination (see Salas, Bowers, & Cannon-Bowers, 1995, for a review). Similarly, in construction, research on processes related to team performance (e.g., preventing and managing errors, avoiding accidents) has focused on mutual performance monitoring, feedback, and backup behaviors (Mitropoulos & Memarian, 2012). This contrasts with the health care context, given that, as a whole, the literature has explored a number of teamwork processes with an overall leaning toward the interpersonal dimension (as shown in Figure 5). The National Aeronautics and Space Administration (NASA)’s research on astronaut teamwork also focuses heavily on interpersonal processes because astronauts work within teams in high-stress environments for long durations. NASA focuses on teamwork processes for astronaut training including self-care and team care, as well as constructive conflict, cooperation, and support of teammates (Landon, Slack, & Barrett, 2018). This overlap suggests there may be some similarities and lessons to be learned from other contexts, particularly aerospace.
Notably, we did not locate comparable reviews on the milieu of teamwork processes in other contexts (i.e., with similar aims of summarizing holistic teamwork processes within or across disciplines). Thus, bibliometric or systematic reviews on teamwork processes in other settings may be useful, not only to summarize the literature in the appropriate domain but also to compare the content and nature of research conducted across environments.
Limitations and Future Directions
There are several limitations to this research, which may be addressed in future research. Our inclusion criteria restricted the types of research eligible for review. This work only included primary data, given that we wanted to focus on the empirically based study of teamwork in health care. A large number of studies (as shown in Figure 1) were excluded based on study approach, such as commentaries and summaries of teamwork as they occur in the field. These potentially valuable sources of information can be explored in future research, particularly as they may reveal information about perspectives toward teamwork. Moreover, we identified articles that related not only to teams but also to more specifically teamwork. That is, our search terms honed in on studies that examined the processes and mechanisms of group-based care, filtering out those that discussed health care teams in passing or without more specific inquiry into teamwork. Going forward, studies may seek to be more inclusive by incorporating work that involves teams at an even broader level.
Furthermore, we used a computer-assisted approach to help parse the large number of included studies. This system was adopted to ensure that analyses were performed objectively and uniformly across the volume of data reviewed in our study. It also enabled us to identify general patterns across a vast, interdisciplinary literature. This overarching view allowed us to develop insights on ways in which team scientists can better contribute to this burgeoning area of research. However, such broad synthesis of thousands of studies also inherently and necessarily involves loss of detail, including potentially not capturing uncommon permutations of the same themes. Traditional qualitative methods (e.g., those which involve iterative coding and line-by-line readings of the literature), in combination with subset analyses of this literature (e.g., comparing specific disciplines with one another), could lead to finer grained outcomes and answers to more specific research questions. We thus encourage researchers to expand upon trends identified in this literature, incorporating alternate methods of review and comparison.
Relatedly, this systematic review focused on exploring teamwork research through descriptive coding and analyses. Although we quantified many markers, these were not meant to test hypotheses or differences; rather, we sought to provide an overview of the current state of the literature. In the future, meta-analyses can statistically measure variables describing these bodies of work and provide further evidence of interdisciplinary phenomena and differences.
These parameters notwithstanding, this systematic review contributes to the literature in several ways. To our knowledge, this work is among the first to systematically integrate across disciplines the health care teamwork literature, which synthesizes findings from industrial and organizational psychology, human factors, informatics, public health, communications, medicine, organizational behavior, allied health, and other clinical and behavioral sciences. Historically, previous reviews of teams in health care have typically limited scope to dynamic specialties (e.g., only emergency medicine or surgery; Manser, 2009) or examined specific processes (Castelao Fernandez et al., 2013; Nagpal, Vats, Ahmed, Vincent, & Moorthy, 2010). However, our search included various health professions and specialties, as well as diverse study designs and data collection approaches, painting a comprehensive picture across domains.
Second, this breadth of analysis was made possible through the novel, computer-assisted application of a key term rubric. Future studies may find this methodology useful in systematically mining through large volumes of research. Importantly, key term rubrics should be applied carefully. As in our study, output should be reviewed to make sure that coding was accurate. Moreover, these analyses return a description of the data, but should be appropriately augmented by closer readings of the material. In this study, we used the quantitative findings to point us toward potentially interesting trends that could be investigated further through deep dives of the material.
Third, by clarifying how different disciplines study teamwork, this research also informs researchers on past and potential directions of this area. It quantifies and establishes the foothold of teamwork research in health care, illustrating the many parties invested in this topic. It also allowed us to match the emergence of health care teamwork research with events in scholarly literature and medical reports.
This review helps clarify the status quo, and thereby opens up potential areas of inquiry. Indeed, scholars from various disciplines can follow several streams based on our findings. For example, researchers from dynamic domains of medicine may choose to continue focusing on popular topics, such as those within the transition processes—or, given the relative gap in interpersonal teamwork research, they may instead explore the less frequently examined dynamics of emotions and attitudes therein. This work also clarifies how some disciplines may mutually benefit from collaboration. Quality improvement and human factors and informatics demonstrated a higher-than-average investment in organizational interventions but apply methodological approaches differently. Scholars in these two fields may thus consider bringing their varied perspectives and expertise into the multi-disciplinary development and evaluation of health care initiatives. Overall, this systematic review helps delineate the teamwork processes associated with specific health care contexts, clarifying new lines of research and encouraging collaboration by identifying overlap and contributions across disciplines.
Conclusion
Overall, teamwork in health care has become an internationally recognized affair. Authors, journals, and health care industries across the globe, from highly complex health care systems to the developing world, have demonstrably invested in research in this area, per our results. These expanding bodies of work are reflections of the realities of health care. This research demonstrates that awareness is growing, the literature is evolving, and that its translation into practice via trainings and interventions has been gaining traction.
It also illustrates how many different sectors of academia and clinical practice have been actively investing in building this important literature. The importance of teamwork is readily discussed in diverse health care contexts, from nursing and community health care settings (e.g., Dressel et al., 2017) to disaster zones (e.g., King, Larkin, Fowler, Downs, & North, 2016). Although teamwork’s relationship to serious health outcomes was highlighted in many articles, it was even discussed in tandem with more mundane topics, including escape room team-building activities (Wu, Wagenschutz, & Hein, 2018; Zhang et al., 2018) and the playing of music in operating theaters (Weldon, Korkiakangas, Bezemer, & Kneebone, 2015). Indeed, it appears that the medical and scientific world has acknowledged the integration of teamwork into innumerable aspects of health care provision.
Our research tracks the bibliometric evolution of teamwork in health care, demonstrating how teamwork has entered the lexicon of scholars across disciplines since the early 2000s. Given the exponential growth of health care teamwork research, it is useful to examine how it is being empirically studied across disciplines. Our research identified several trends, including the dominance of descriptive and quantitative approaches on health care teamwork, as well as interdisciplinary differences in methodological and study design. It also clarified how different schools of thought approach health care teamwork through specific processes and content areas. Although interpersonal processes are the most commonly studied teamwork process, medicine’s focus on transition processes and the team sciences’ emphasis on action processes can contribute to fuller understanding. Indeed, the area of health care teamwork would be best served by scholarly teamwork. This underscores what Salas et al. (2018) identified: “Clinical and nonclinical professionals (e.g., organizational health, public health, cognitive engineering) are capable of transforming health care by working together” (p. 359). Overall, this review identifies ways to expand knowledge, particularly through the incorporation of team science expertise, and lead to a more robust understanding of health care teamwork.
Supplemental Material
Supplement_Study_summary – Supplemental material for Cross-Disciplinary Care: A Systematic Review on Teamwork Processes in Health Care
Supplemental material, Supplement_Study_summary for Cross-Disciplinary Care: A Systematic Review on Teamwork Processes in Health Care by Julie V. Dinh, Allison M. Traylor, Molly P. Kilcullen, Joshua A. Perez, Ethan J. Schweissing, Akshaya Venkatesh and Eduardo Salas in Small Group Research
Footnotes
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) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This material is based upon work supported by the National Science Foundation Graduate Research Fellowship Program under Grant No. 1450681. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.
Supplemental Material
Supplemental material for this article is available online.
Author Biographies
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
