Abstract
A team’s ability to coordinate and adapt their performance to meet situational demands is critical to excellent patient care. The goal of this article is to identify common coordination characteristics that enable health care action teams to ensure effective patient care and to discuss specific examples of adaptive coordination within the health care setting. Task analyses were conducted to identify situational demands, in three different clinical settings: cardiac anesthesia, pediatric sepsis simulation, and trauma resuscitation. Each task analysis identified specific coordination requirements for pertinent tasks. The research team compared these task analyses, identified emerging themes, and agreed on core coordination characteristics common across all three environments by consensus through iterative abductive analysis. Findings across these diverse clinical settings showed that expert action teams (a) continually appraise their dynamic environment, (b) identify and define points of coordination, and (c) respond to the demands of nonroutine events by making coordination highly explicit. Specific examples of adaptive coordination within the health care setting are discussed, and implications for training are articulated. Findings are also pertinent outside of health care and may contribute to the understanding of coordination behaviors within action teams across multiple settings.
Introduction
In health care, teamwork is repeatedly cited as a critical aspect of performance (Schmutz & Manser, 2013), and it is now included in interprofessional competencies within the foundation of health professional education (Interprofessional Education Collaborative, 2011; Weaver et al., 2010). Although progressively better-defined competencies are evolving for specific clinical domains (Eppich, Brannen, & Hunt, 2008; Gaba, 2010; Leasure et al., 2013; Yule, Flin, Paterson-Brown, Maran, & Rowley, 2006), most are based on generic team competencies (Salas, Sims, & Burke, 2005) and assume that teamwork requirements remain constant across time, tasks, and clinical settings.
In recent years, the teamwork literature as a whole has evolved to consider teamwork as more dynamic, explicitly stating that teamwork is not stable (Kozlowski, 2015; Waller, Okhuysen, & Saghafian, 2016). Explorations of the patterns of coordination reveals that to be most effective, teams engage in a combination of explicit and implicit coordination, and that there may be certain patterns that are more effective than others (Grote, Kolbe, Zala-Mezö, Bienefeld-Seall, & Künzle, 2010; Kolbe et al., 2014; Rico, Sanchez-Manzanares, Gil, & Gibson, 2008). Few studies have investigated adaptive coordination processes in health care teams (e.g., Bogdanovic, Perry, Guggenheim, & Manser, 2015; Burtscher et al., 2011; Manser, Harrison, Gaba, & Howard, 2009). There are significant research and practical implications for building on this work to better understand optimal team performance in the high-risk domain of health care.
The goal of this article is to identify common coordination characteristics that enable health care action teams to ensure effective patient care and to discuss specific examples of adaptive coordination across three different high-risk health care domains. Action teams are groups of highly skilled, multidisciplinary specialists who conduct interlinked tasks during complex, time-limited performance events in challenging environments (Sundstrom, McIntyre, Halfhill, & Richards, 2000).
The contribution of this work is (a) a clearer definition of common coordination characteristics and (b) specific examples of skills that can be utilized to create opportunities for adaptive coordination.
Adaptation
The ability of a team to adapt is a main characteristic of high-performing teams (Burtscher, Wacker, Grote, & Manser, 2010; Lei, Waller, Hagen, & Kaplan, 2015). Team adaptation—the process of change within a team—is a dynamic process, that is, highly dependent on task and team characteristics (Burke, Stagl, Salas, Pierce, & Kendall, 2006; Maynard, Kennedy, & Sommer, 2015) and most frequently occurs after a disruption in team process (Maynard et al., 2015).
There is most often a cue preceding adaptation (Burke et al., 2006). A cue is an indicator for action, which exists in the work environment. A trigger for adaptation is an external signal that indicates when adaptation should occur (Maynard et al., 2015). A cue can be a trigger once it has been recognized by the team. Not all cues for adaptation are created “equal” (Baard, Rench, & Kozlowski, 2014; Burke et al., 2006), meaning that adaptation is shaped by the cue and the trigger, whether they are team-based or task-based, and at what temporal stage the trigger occurs (Okhuysen, 2001; Okhuysen & Waller, 2002). Therefore, coordination processes, which occur postcue/trigger, must be adapted for the team to continue improving in the face of cues for adaptation.
Coordination in Action Teams
Coordination is a critical aspect of team performance, especially in interdependent teams executing complex tasks (Fiore, Salas, & Cannon-Bowers, 2001; Okhuysen & Bechky, 2009). Coordination has been defined as “the process by which team resources, activities, and responses are organized to ensure that tasks are integrated, synchronized, and completed within established temporal constraints” (Cannon-Bowers, Tannenbaum, Salas, & Volpe, 1995, p. 345). Coordination requirements vary as a function of task characteristics (e.g., complexity), team characteristics (e.g., familiarity; Foushee, Lauber, Baetge, & Acomb, 1986; Kanki & Foushee, 1989), and the dynamically evolving situation (e.g., time pressure; Kontogiannis & Kossiavelou, 1999).
Previous research has articulated behaviors that are indicative of coordination (Rosen et al., 2011). To be effective, the presence of coordination behaviors alone is insufficient. Teams must have an accurate and overlapping team situation model (TSM). A TSM is a dynamic, context-driven mental representation of the areas of the team’s work and is influenced by multiple characteristics including team composition, team attributes, and team performance environment (Rico et al., 2008). Coordination behaviors can contribute to establishing a TSM (Kolbe, Burtscher, Manser, Kunzle & Grote, 2011; Tschan et al., 2011).
Achieving optimal coordination is particularly challenging in action teams. They must simultaneously execute several tasks, avoid conflict with other team members, and seamlessly handover tasks to be performed during a critical event (Kontogiannis & Kossiavelou, 1999; Watts & Monk, 1998). Because of changing task goals, time pressure, and shifting team membership (Klein, Ziegert, Knight, & Xiao, 2006) coordination requirements may evolve throughout the task, making coordination in action teams very challenging, especially in the case of critical events.
Adaptive Coordination in Action Teams
Taking account of the literature from adaptation, action teams, and coordination studies, we are interested in further understanding adaptive coordination in action teams, particularly those teams which work within dynamic health care settings.
Burke et al. (2006) have formulated a comprehensive model of team adaptation, based on evidence from empirical studies conducted in various domains, to describe challenges facing teams as they continuously adapt to changing task and situational demands. Their model distinguishes between adaptations concerning (a) inputs into the teamwork process such as a mobilization of additional resources or a structural reconfiguration of the team as well as (b) team processes, that is, changes in coordination mechanism, decision making, and communication patterns in response to unexpected events (Brehmer, 1996; Entin & Serfaty, 1999; Serfaty, Entin, & Deckert, 1993). Team members draw from their individual and shared resources to detect, frame, and act on triggers or sets of triggers that signal the need for team-level adaptation (Burke et al., 2006). The ability to recognize these triggers is an essential requirement for team adaptation.
Although the framework by Burke and colleagues (2006) describes adaptive performance in a generic way, other research has focused on adaptation of specific team processes such as coordination (Burtscher et al., 2010; Grote et al., 2010). Adaptive coordination has been defined as a team’s ability to change its coordination activities in response to unexpected external events or changes in task or team characteristics (Burtscher et al., 2010; Entin & Serfaty, 1999). Thus, team members need to be aware of their environment and search for cues that could affect the success of the team’s mission.
Because of the complexity of tasks undertaken by action teams, a single point or summary assessment of coordination would be of limited utility. Coordination requirements change throughout a task when new information is discovered and integrated, or feedback is given (Rosen et al., 2011). Although current team-training models often include “adaptability” or “updating” as an important aspect of teamwork, it is necessary to help learners better understand exactly what adaptation means, and how they can identify situations that may require adaptation (Gorman & Cooke, 2010; Pulakos, Arad, Donovan, & Plamondon, 2000). Within health care, situational cues exist within a dynamic clinical context and can serve as triggers for adaptation; therefore, cue recognition is a necessary skill (Friedman et al., 1991; Moulton, Regehr, Mylopoulos, & MacRae, 2007; Salas, Cannon-Bowers, Fiore, & Stout, 2001).
Adaptive Coordination in Health Care Action Teams
Health care teams, especially in acute care environments, are assembled ad hoc, have dynamic membership, often work together only for a short period of time and involve specialists or specialist subgroups (Kolbe et al., 2011; Schmutz, Hoffmann, Heimberg, & Manser, 2015). Research on health care action teams has shown that tasks frequently evolve outside of protocolized standards due to changes in situational requirements (Fernandez Castelao, Russo, Riethmüller, & Boos, 2013; Klein et al., 2006). For example, the standard protocol for performance in a resuscitation would be to give fluid if the patient is hypotensive but because of situational changes (e.g., the patient’s blood pressure drops during transfer from the emergency department to the ward), the team must decide whether to deviate from protocol based on the new task requirements (e.g., stabilize the patient en route).
Evidence from interview and observational studies in various care settings underlines the relevance of adaptive coordination in health care action teams (Bogdanovic et al., 2015; Burtscher et al., 2010; Manser et al., 2009). For example, changes in task load were found to influence the explicitness of coordination behaviors of anesthesia teams during induction (Zala-Mezo, Wacker, Kunzle, Bruesch, & Grote, 2009). Anesthesia teams were also found to exhibit better treatment performance when adapting their coordination patterns in response to unexpected events (Burtscher et al., 2011). Adaptation can be improved through targeted team training, indicating that it is malleable and can be taught (Riethmüller, Castelao, Eberhardt, Timmermann, & Boos, 2012).
Due to the fact that coordination as well as adaptive coordination is highly dependent on team and task characteristics, current frameworks are not able to provide specific sets of behavioral responses for frontline providers and thus for team training. Situational cues for adaptation as well as the appropriate behavioral responses need first to be defined using a systematic approach. By applying task analysis in three different clinical settings, we attempt to address this gap in the current understanding of adaptive coordination in action teams. While specific behavioral responses can hardly be generalized for different task contexts, common characteristics of adaptive coordination are identified that can inform training. By leveraging data from three distinct health care settings, we attempt to identify coordination and adaptive coordination mechanisms that are applicable across these settings. By identifying these commonalities, our findings may prove valuable to managers to assist their teams to become more adaptive and also shape the content of training interventions seeking to improve team adaptation.
Method
Study Setting
Three unique task environments of health care action teams are included in this article in which adaptive coordination is essential. This is a secondary analysis of data collected during earlier studies (Henrickson Parker, 2015; Manser, Howard, & Gaba, 2006; Schmutz et al., 2015). Although the focus of previous analyses of these data had been the identification of coordination behaviors to be included in taxonomies for observational research and the investigation of relationships between specific coordination behaviors and team performance, the contribution of this work is to identify the common coordination requirements across three different clinical settings. This study is unique from the previous three in that it explores different clinical settings in an attempt to define common coordination and adaptive coordination mechanisms across environments, thus resulting in more general principles.
Cardiac anesthesia
The first task environment of interest is cardiac anesthesia. The goal of this study was to understand coordination requirements within this unique team environment, focusing specifically on the coordination of the anesthesia crew, including anesthesia resident, attending and certified registered nurse anesthetist. In this environment, there are many complicated, but routine tasks and the anesthesiologist must work with both the cardiac surgeon and perfusionist closely to monitor the patient’s status. Although this work is rather protocolized over long periods, should something go wrong, the team has to respond quickly to maintain stability of the patient’s safety.
Pediatric resuscitation
The second task environment of interest is a simulated emergency for pediatric septic shock. The goal of this study was to better understand the performance relevant effects of three specific coordination tasks (task distribution, provide information without request, and closed loop communication) in an emergency simulation. This type of emergency can occur in multiple clinical environments (ICU, pediatric ward, etc.), and it involves a patient unexpectedly decompensating, often very quickly, due to the rapid onset of shock. Rapid recognition is critical, and once recognized, goal-directed therapy is required.
Trauma resuscitation
The final task environment of interest is acute trauma resuscitation at a Level 1 trauma center. The goal of this study was to better understand coordination across team members, throughout a dynamic and adaptive task. In this environment, the team may have little knowledge of the patient’s needs prior to the patient arriving, and the patient is often acutely sick. The team may be ad hoc and determines their goals based on the information at hand, frequently updating upon patient arrival and throughout caring for the patient.
Procedure for Initial Task Analyses Focused on Identifying Coordination Requirements
For each of the three studies, the goal was to identify coordination requirements, and to better understand how these requirements need to be adapted to accomplish specific clinical goals. A similar approach was taken across the three studies. First, familiarizing observations were conducted to allow the researcher to deconstruct the task and to see beyond the extraordinarily technical aspects of clinical care (e.g., open heart surgery) and focus on the team coordination behaviors (Thomas, Sexton, & Helmreich, 2004). Once familiar with the task environment, the main task (cardiac anesthesia, pediatric septic shock care, trauma resuscitation) was deconstructed using standard task analysis techniques for teams (Annett, 2005; Kirwan & Ainsworth, 1992; Stanton, Salmon, Guy, Baber, & Jenkins, 2005). Task analysis allows researchers to identify both operational requirements (task work) and team requirements (teamwork) among the different members of a team (Marsch et al., 2004; Tschan et al., 2011).
For each procedural step, groups of steps, or task sequences in the task analyses, coordination requirements were gathered from expert interviews, using an augmented hierarchical task analysis approach (e.g., Annett, 2005; Kirwan & Ainsworth, 1992). To determine the coordination requirements, specific probing questions were asked of the participants on how they gather information for each task step, where they find information, how they share information, and how they monitor progress toward task goals.
Comparing Coordination Requirements Across Health Care Action Team Settings
The results of the three previous task analyses were distributed among the research team. Research team members then familiarized themselves with each analysis and held multiple detailed discussions on the task analysis and coordination requirements necessary to accomplish the clinical goals for patients in different clinical environments. Because the focus of this work is identification of core coordination requirements for action teams in health care, we chose to utilize an abductive approach (Timmermans & Tavory, 2012). Abductive analysis allows for theory generation while placing findings against a background of existing theories. In this case, we utilized the extensive literature on coordination and adaptive coordination, and attempted to apply it to the dynamic work of health care action teams.
The research team (the authors) first examined each task analysis in detail, then identified common coordination requirements across all three clinical environments. The authors are all PhD-level industrial/organizational psychologists with research history in team coordination within health care. The team then discussed each task step and associated common coordination requirement, and attempted to identify common themes. Discussions took place over a 6-week period, were held in person, and were each approximately 1 hr long. Because we used pooled data that were not collected for the purposes of comparative analysis, we did not have a priori categories that we used for analysis. Rather, we drew on the identified similarities among our findings regarding specific coordination behaviors across settings (Bechky & Okhuysen, 2011). We were specifically interested in situational triggers for adaptation. For example, between the three different environments, in instances where tasks were tightly coupled, how did coordination requirements change? Are there phases or stages to the task that serve as coordination anchors for determining next steps? What specific coordination behaviors overlap through a case?
Results
Characteristics of Coordination Requirements Across Health Care Action Team Settings
Three overall themes emerged from the comparison of the task analyses pertaining to coordination requirements in health care action teams. General adaptive coordination requirements include (a) continually appraising the dynamic environment, (b) identifying and defining anchoring points of coordination between team members, and (c) responding to the demands of nonroutine events through making coordination highly explicit. All of the tasks included in the task analyses were based upon protocols and procedures specific to each clinical care domain, and explicitly stated within normal care procedures (e.g., Advanced Trauma Life Support [ATLS] guidelines for trauma resuscitation). The tasks could be relatively stable (cardiac anesthesia) to requiring some adaptation, which would then result in stabilization (sepsis response) or finally, those necessitating improvisation and continuous adaptation (trauma resuscitation).
Continually Appraise the Dynamic Environment
With all tasks, teams have to continually evaluate the patient’s status, the team’s status, and ongoing changes, however, certain tasks have specific points at which re-evaluations may make more sense. Regardless of the initial stability of a clinical setting, the stability of the previous minute does not indicate the requisite level of coordination required for the next. We have found across health care action teams, that often define these cues themselves (e.g., “if the BP drops below xyz, call me so that we can discuss changing the medication”) to make sure other team members interpret them as a trigger for recoordination. Constant reassessment of coordination requirements is necessary for smooth and effective task performance.
In our data sets, this shift often took the form of building and updating a template for performance. A template is defined as a preset format that serves as an exemplar pattern for team tasks that does not need to be recreated each time it is used. For example, though the task of cardiac anesthesia is largely template-driven, at any given point, the patient could become unstable or the surgical environment could change. At this point, the situation would require reassessment, potentially switching to a different template and an updating of coordination requirements.
In pediatric septic shock, the situation is initially dynamic requiring intense coordination to integrate available diagnostic information, but once a diagnosis is made, the task requirements are well defined. Thus, the focus of coordination requirements shifts from information gathering to a more templated structure requiring task management. Similar to cardiac anesthesia, a patient may become acute at any time, and therefore, coordination requirements may need to be re-evaluated and updated.
In trauma resuscitation, the team has a very rough template to work from when the patient arrives but learns more about their task as they are engaged in it. For example, in one live observation, a patient entered the trauma bay labeled as a “pedestrian hit by a car” but upon visualization of an emergency CT scan, it could be seen that the patient had a gunshot wound to the head, received while at a traffic stop, thus completely changing the current working template for the team. The initial template of “blunt trauma, possible internal injury” must be updated to “penetrating injury to the brain,” which radically changes the clinical and coordination task. The main coordination challenge related to this situation is that this change has to be communicated across the team and will also alter the associated coordination processes (setting of priorities, reassigning team member responsibilities, workload distribution, etc.).
Even if templates are available for each of the relevant treatment episodes, once a diagnosis has been established and the templates are familiar to all team members, the tasks and associated coordination processes are valid for longer or shorter periods of time depending on the dynamicity of the task. Teams in all three task environments have to adapt their coordination requiring effective management of task-related information (e.g., initial assessment and diagnosis), task execution (e.g., giving fluids), and current team status (e.g., do we have the skills we need as a team?). The critical coordination competency highlighted by our analysis is that the team has to continually reassess if the template they are working from is still valid and if not, integrate the available information and communicate the necessary shift.
Identify and Define Anchoring Points of Coordination Between Team Members
Across all three tasks, “anchors” of coordination emerged. Anchors are brief pauses to reassess and redistribute task and resource needs. At these anchors, there were particular task characteristics that were indicative of the need for a different type of coordination. For example, there are defined procedural steps in cardiac anesthesia, such as going on and coming off bypass where the surgical and anesthetic teams are highly interdependent, thus intensifying the coordination requirements. In addition to these anchors inherent in care episodes, team members may define additional anchors based on changes in clinical conditions or their anticipation of increased need for re-evaluation and decision making for a specific patient.
In pediatric septic shock and trauma resuscitation, anchors were also present but less obvious at the beginning of the task. Similar to anesthesia, anchors were significant clinical moments (such as transitions between subgoals) or points at which the patient parameters would indicate that a decision had to be made or a different level of intervention was necessary (e.g., the patient losing consciousness or the patient not sustaining a patient airway). Anchors were closely associated with significant clinical decisions arising from situational changes rather than transitions between predefined procedural steps.
Across all three clinical environments, our task analyses showed that if a template is no longer felt to be correct, any member of the team could stop the team and define an “emergent anchor.” Emergent anchors are opportunities for the team to reassess their coordination and update the team needs based on the situation. Recognizing the need for an emergent anchor and detecting those anchors are both critical competencies for coordination in health care action teams. However, this skill may be identified as part of the clinical task, rather than as part of the necessary coordination to achieve the team’s goals.
In routine situations, coordination anchors remain stable, and are often effectively supported by implicit coordination. In all three task environments, nonroutine events necessitated a shift to more explicit coordination, even if only to indicate moving from one template to another. For example, in cardiac anesthesia, all team members possess an in-depth understanding of interrelations between the tasks of the various subteams and coordinate implicitly over extended periods of a case. However, if an unexpected event occurs, such as unexpected bleeding, this clinical event would amount to a new emergent anchor indicating the need to reallocate tasks, resources, and to adjust the plan. The septic shock scenario starts with routine tasks and coordination is both implicit and explicit because the team is aware of the template for the initial assessment, yet the protocol dictates some explicit coordination (e.g., communication about the state of the airway, etc.). The series of events when a patient is trending toward instability and possible decompensation indicates the beginning of a nonroutine event that affects coordination requirements. Because of the patients’ critical state, the team has to act quickly and communication is mostly explicit and directive (e.g., stating decisions, giving orders, distributing tasks) shifting to a new type of coordination. As soon as the patient is stable, there is a shift again to more implicit coordination within a template for treating the septic shock (e.g., nurses preparing the next fluid bolus without explicit instruction). Each of these shifts could require an anchor, ensuring that the team is appropriately recoordinating their actions.
Trauma resuscitation provides a unique environment to examine templates and emerging anchors. Because the task involves multiple templates that are constantly updated and reassessed, it may be that nonroutine events are actually considered “routine” by clinicians working in this task environment. For example, the task analysis showed that there is often initial information gathering and triage, which serve as a coordination anchor. Because of the nature of the injury, however, this coordination anchor may recur after initial assessment, and the output of the anchor in terms of templates and coordination requirements is highly variable. Overall, coordination in this task environment is much more explicit, whereas coordination within subteams (such as the nursing team) may be implicit within a specific procedural step. There is a recognized need for constant explicit updating to ensure effective overall coordination within and between subteams. Therefore, within this environment, emergent anchors are frequent and often dynamically updated. The complexity of the coordination requirements in this setting has resulted in many trauma centers requiring the role of a trauma team leader who, rather than engaging in hands-on care of the patient, is present to independently manage and convey the excessive amount of explicit and emergent coordination anchors that occur.
Respond to the Demands of Nonroutine Events Through Making Coordination Highly Explicit
Because of the existence of emergent anchors, health care action teams seem to engage in microloops of coordination. In line with the conceptualization of episodic work by Marks, Mathieu and Zaccaro (2001), we define microloops as short periods of task execution after which there is a reassessment of coordination needs occurring in a rapid cycle. During each loop, the team may operate based on a template but at transition points there may be a need to switch between templates. In our task analyses, particularly for the more dynamic team tasks such as septic shock and trauma resuscitation, we identified many indications of fast-paced reassessment loops. Some of these loops are built into treatment protocols (e.g., the ATLS protocol for trauma). For example, in a septic shock scenario, the main goal is the administration of fluid. According to guidelines, three fluid boluses should be administered. Every administration of fluid is followed by a reassessment phase (i.e., assess breathing and circulation). We found that for each of these microloops in the clinical task, there was a mirroring loop in coordination. That is, a decision of the course of action needs to be communicated to the team both before and after each reassessment to maintain a coordinated treatment of the patient (Schmutz, Eppich, Hoffman, Heimberg, & Manser, 2014).
We also found that within large complex teams, subteams may have their own subtemplates, anchors and microloops. For example, during trauma resuscitation, the nursing staff working on a defined subgoal has different coordination requirements that are nested within the coordination requirements of the overall team. Actively engaging in microloops of coordination, and recognizing the need for coordination updating, based on situational demands are both necessary skill competencies to avoid coordination breakdowns in these teams. Yet, they differ among team members, despite the overarching goal being the same. These differences could necessitate further explicit communication, increased need for microloops of communication and explicit anchors to ensure the team’s shared awareness.
Discussion
The goal of this article is to identify common coordination characteristics that enable health care action teams to adaptively ensure effective patient care and to discuss specific examples of adaptive coordination within the health care setting. Findings across three diverse clinical settings showed that expert action teams (a) continually appraise their dynamic environment, (b) identify and define points of coordination between team members, and (c) respond to the demands of nonroutine events through making coordination highly explicit. They do these activities through using and updating templates of performance, using predefined and emergent coordination anchors, and engaging in microloops of coordination. Our results show that common behavioral responses for adaptive coordination can be found across multiple health care environments. We believe that by using task analysis, we are able to link triggers for actions to specific tasks, connecting the general notion of adaptive behavior to the specifics of the task. In addition, because these responses were behavioral in nature, there are implications for these findings to health professions education and teamwork training in general.
We believe this article provides additional knowledge on how adaptive coordination functions in health care action teams. Within current teamwork behavior taxonomies that dominate health care (e.g., TeamSTEPPS in the United States), behaviors such as mutual performance monitoring and backup behavior and communication, and role organization are considered present or absent, and the presence of these skills, when combined, will yield a highly flexible and adaptive team (King et al., 2008). Our findings further this conclusion, showing that the situation dictates additional information to the team about their performance requirements. It may be inadequate to consider these skills as present or absent. Rather, it is the manifestation of skills within a situation with continual updating that ensures a team is acting appropriately within a situation.
Because context drives much of the adaptive coordination required within action teams, a more detailed approach that provides guidance on what works (specific behaviors), for whom (experience levels), and in what context (work and task characteristics) is warranted (Maloney, Bresman, Zellmer-Bruhn, & Beaver, 2016). With this article, we provide an initial basis for identifying relevant situational cues for adaptation and the corresponding adaptive processes (i.e., templates, anchors, and microloops of coordination).
The results of our analysis show that to effectively train coordination and importantly, adaptive coordination in action teams, focusing on coordination templates, anchors, and microloops is critical. If teams do not acknowledge coordination anchors (i.e., miss triggers for adaptation), and thus do not update their templates, coordination requirements may shift, while coordination behaviors remain the same. Therefore, the team is not responding appropriately to the clinical situation, potentially increasing the likelihood for a mistake or at a minimum, causing delays in care. For example, in a trauma, if the team does not alter their coordination after a critical shift in patient status (e.g., the oxygen saturation changes after the patient is thought to be stabilized) and update their template (e.g., the patient is stable vs. the patient is no longer receiving appropriate oxygen), they may continue coordinating between each other using an old template. Although it is expected that any team member could fulfill the requirement to redefine an “emergent anchor” according to standing protocols and procedures, team members with a lower hierarchical ranking (e.g., nurse) may find it difficult to engage in appropriate “speaking-up” behaviors, given the presence of a strong hierarchy. Therefore, this skill may require specific attention in team training (Okuyama, Wagner, & Bijnen, 2014).
A microloop for coordination would facilitate adaptation in this situation. Due to the multiple parallel tasks performed in this setting, it may be difficult for team members, especially when inexperienced, to identify relevant changes as triggers for coordination. Thus, explicitly defining cues, which may trigger adaptive team processes, crucially need to be monitored.
As illustrated in this example, individuals in complex health care teams are not always able to identify when the task or coordination requirements are shifting. Cue recognition is crucial, lest the team “plow through” (Moulton et al., 2007), and miss an important emergent anchor. There can be performance “drift” in which there is a failure to transition from automatic mode (Moulton, Regehr, Lingard, Merritt, & MacRae, 2010). These findings show the necessity of recognizing subtle cues leading to necessary shifts in coordination. To identify shifts in situational requirements, teams must be able to recognize, select, interpret, and respond to important environmental cues to ensure optimal adaptation of coordination during complex tasks (Burke, Salas, Wilson-Donnelly, & Priest, 2004). Thus bursts of during-action reflection such as a reassessment of coordination needs prevent a team from blindly working toward the wrong goal (Schmutz & Eppich, 2017). In the trauma example above, a reassessment of the patient and the team’s coordination status might prevent premature closure of the team’s activities. This need increases even more during periods of high-stress, complex work. In such situations, a time-out where a team stops and reviews the process, creating a forced coordination anchor, might help identify relevant situational cues (Rall, Glavin, & Flin, 2008).
Standardized steps or templates, anchors and microloops of coordination may help facilitate cue recognition, particularly for learners. Defining the template for performance may be conducted through a skills-based training such as ATLS or Advanced Cardiac Life Support (ACLS). During these training courses, clinicians are taught specific procedural/clinical steps and often practice those steps in high and low fidelity simulations (American College of Surgeons Committee on Trauma, 2013). However, there seems to be limited training on the coordination anchors required to ensure the team is able to perform optimally (i.e., When do I need to coordinate what and with whom?). Anchors have been built in to current clinical processes, such as briefings, huddles, or structured closed loop communication (Glymph et al., 2015; Härgestam, Lindkvist, Brulin, Jacobsson, & Hultin, 2013; Marks, Zaccaro, & Mathieu, 2000). Making a coordination anchor explicit and providing both language and structure around these coordination anchors (e.g., before commencing a surgical procedure, conduct a team time-out, Henrickson Parker, Wadhera, Elbardissi, Wiegmann, & Sundt, 2009) and emergent anchors (e.g., defining team leader before commencing resuscitation) could help team members to engage in critical team tasks.
Explicit coordination anchors can facilitate this type of sharing and provide an assessment of the current state of the patient and the anticipated next steps. Research examining expert performance in other high-risk settings has shown that experts make extremely rapid situation assessments, followed by task-related decisions that have a significant impact on the course of action and the associated coordination requirements for the team (Klein, 2008). Our results highlight that expert team performance requires depth of understanding to choose templates for performance, updating to make sure the team is able to adapt to evolving situation requirements, and anticipation of changes to define coordination anchors (Klein, 1989).
Research and Practical Contributions
Research contributions
This study contributes to the literature on team adaptation to situational dynamics by providing evidence from health care for three common coordination characteristics across clinical settings. Although the three characteristics of adaptive coordination that we identified in this study must be further explored in future research, our work provides a starting point for developing a framework that has the potential to bridge the team literature on generic versus task-specific coordination requirements. The notion that action teams are highly effective through (a) continually appraising their dynamic environment, (b) identifying and defining points of coordination (i.e., cues and anchors), and (c) responding to the demands of nonroutine events through highly explicit coordination should be studied in more detail to further our understanding of how the dynamics of the task affect on the actual coordination behaviors needed to achieve optimal team performance. In particular, the recognition of situational cues, their interpretation as triggers for coordination and where necessary, the explicit definition of coordination anchors, requires the use of complementary methodological approaches such as focused behavioral observation, stimulated recall, and experimental approaches. To move this line of research forward, we propose utilizing simulation, in concert with live settings, and deliberately manipulating various aspects of the task environment to determine their specific impact on team coordination and performance.
Practical contributions
Coordination in health care has been studied to a degree, but the focus has not been on training teams how to coordinate effectively. To affect frontline training on adaptive coordination, it is necessary to have a better understanding of how teams adapt effectively in different situations. Our work explicitly contributes to this goal by articulating common adaptive coordination processes across teams. Team training in health care is largely based on static models of team performance, yet the work is highly dynamic and variable. Teaching teams to recognize changes in the situation and how to appropriately adapt coordination behaviors is critical. Changes in a patient’s clinical trajectory, which can be identified through procedural steps, care episodes or by time markers can serve as anchors for indicators of changing coordination requirements. Identifying when these anchors exist gives clinicians the opportunity to explicitly understand that requirements for coordination are changing, and thus they must adapt their behaviors. Training could be shaped around experiential moments, either through video debriefings of live events or through simulation. The medical education community has become more invested in developing behaviorally measurable competencies for teamwork, explicitly stating that the capability to work effectively as part of an interprofessional team is a core indicator for readiness for residency (AAMC-Entrustable professional activities 9; ten Cate, 2013). Research such as this offers further support for behavior-based interventions.
To our knowledge, this is the first study that analyzed and compared coordination requirements across routine and emergency situations, and to make recommendations for competencies based on adaptive coordination requirements. Our study highlights important gaps in current training, and will allow for targeted competency-based developments to address these gaps.
Limitations and Future Research
As with all studies eliciting expert knowledge of a complex task, our study is based on the experience of experts and does not provide any empirical data linked with team performance. The quality of the data was thus dependent on experts’ ability to verbalize their understanding of the issue under discussion. It was difficult for experts to use “coordination language” but not difficult to understand the concept of adaptive coordination, or to describe its importance to team performance.
Use of an established methodological approach—familiarization with environment, constructing a task analysis, and expert interviews—to understand coordination behaviors showed that coordination requirements are consistent across the three domains, but there were slight variations in the details of each task analysis. Task analysis is limited in its ability to capture the complexity of environmentally driven changes, while remaining usable. An exhaustive list of clinical permutations was too varied to be included in each representation. The three study settings differed in many ways that we discuss in our article concerning their impact on coordination requirements. Future studies will have to validate these findings by including more action team settings within and beyond health care, with clusters of team tasks that are similar according to the characteristics of the task environment.
In addition, future research must integrate the temporal nature of adaptation. Effective coordination is largely task dependent (Bedwell, Ramsay, & Salas, 2012) making it difficult to determine the team behaviors that might contribute to effective performance across settings. Particularly within the health care setting, time has a significant impact on the task, as patient status changes dynamically. In addition, the familiarity of the team and team members’ expertise likely affects the mode of adaptive coordination (e.g., implicit or explicit; Rico et al., 2008) and on the speed of cue recognition and adaptive response (Moulton et al., 2010).
Conclusion
Our integrative analysis of three task environments revealed that coordination in health care action teams is varied, but characterized by three common features: (a) continual appraisal of the environment, (b) identification of points of coordination between team members, and (c) making coordination highly explicit in nonroutine situations. Team members utilize templates, coordination anchors, and microloops of coordination to aid in updating their understanding of the situation and the coordination requirements. Health care teams should explicitly engage in and also train these forms of coordination to ensure excellent patient care.
Footnotes
Ethical Approval
Parker study was approved by MedStar Health’s IRB 2012-080 titled “Understanding and Optimizing Coordination in High-Risk Healthcare Teams”; Manser study was approved by Stanford University’s IRB 79750 titled “Coordination demands and processes within anesthesia teams”; Schmutz/Manser study was approved by the University of Fribourg Ethics Review committee; Department of Psychology; on 3/31/2014 titled “Development of checklists for evaluating clinical performance during medical emergencies in pediatrics.”
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 work was funded by the Agency for Healthcare Research and Quality, Ruth L. Kirschstein National Research Service Award (F32HS021 046-01, Parker PI), the University of Fribourg Postdoctoral overseas scholarship (Parker, PI; Manser, Sponsor), and by the Swiss National Science Foundation research professorship grant (PP00P1_128616, Manser, PI) and used data originally obtained during a Swiss National Science Foundation scholarship (PBZH1-100994, Manser, PI).
Guest Editors: Martina Buljac-Samardzic, Connie Dekker-van Doorn, and M. Travis Maynard