Journal Logo

Empirical Investigations

Using Simulation as an Investigational Methodology to Explore the Impact of Technology on Team Communication and Patient Management

A Pilot Evaluation of the Effect of an Automated Compression Device

Gittinger, Matthew MD; Brolliar, Sarah M. MPHc; Grand, James A. PhD; Nichol, Graham MD, MS; Fernandez, Rosemarie MD

Author Information
Simulation in Healthcare: The Journal of the Society for Simulation in Healthcare: June 2017 - Volume 12 - Issue 3 - p 139-147
doi: 10.1097/SIH.0000000000000205

Abstract

Cardiac arrest management is complex, requiring the resuscitation team to perform patient care and medical decision-making under dynamic and time-pressured conditions.1 Greater depth of chest compression, greater cardiopulmonary resuscitation (CPR) fraction, briefer perishock pauses, and optimal compression rate are associated with improved outcomes after out-of-hospital cardiac arrest.2–6 However, coordinating and monitoring compressions during a cardiac arrest event to ensure the adequacy of their delivery are challenging.2,7,8 Importantly, the effort and attention required during active chest compressions encumber not only the individuals performing the task but also the team leader coordinating them. This adds significantly to the resuscitation team's workload, monitoring requirement, and resource demand, thus potentially limiting their ability to focus on other critical patient care diagnosis and treatment-related tasks.

The LUCAS Chest Compression System (Physio-Control Inc, Redmond, Wash) is designed to deliver uninterrupted compressions in a manner consistent with current emergency cardiovascular care guidelines.9 LUCAS is intended to reduce the need for continued management and oversight of chest compressions by the healthcare team after its application to a patient. This may impact teamwork in several ways. First, automating chest compressions may allow team members who would otherwise be engaged in performing manual compressions to engage in the team's planning and diagnostic activities. Second, reducing the need to monitor the depth and rate of compressions may allow the team leaders to focus on more complex diagnostic and treatment issues. Finally, the ability of LUCAS to deliver ongoing compressions during defibrillation reduces the need to coordinate team members on and off compressions, again potentially increasing the team's cognitive capacity to perform higher-level diagnostic decision-making. As such, automated compression devices may have a positive impact on teamwork during cardiac arrest events. This impact may translate into improved patient management and outcomes.

Initial studies of the LUCAS device focused on critical time-based actions (eg, no flow time, time to initiation of compressions) and more distal patient outcomes. To our knowledge, the impact of automated compression devices on team communication has not been evaluated. Simulation provides an important mechanism to systematically evaluate the impact of introducing new technology such as the LUCAS device into the healthcare environment. By providing a standardized evaluation platform to capture clinical behavior and communication, simulation supports rigorous assessment of human factors during complex clinical care. We conducted a pilot study to identify potential effects and associations between use of an automated chest compression device and team communication and patient management during an emergency department cardiac arrest. This is an important first step to further understand how technology such as the LUCAS device might affect clinical care during cardiac arrest resuscitations.

METHODS

Study Design

This was a randomized, controlled design study evaluating the impact of the LUCAS automated chest compression device as compared with manual compressions on team communication and patient management during a simulated cardiac arrest (Fig. 1). Approval for this study was obtained from the University of Washington Institutional Review Board.

F1
FIGURE 1:
Participant flow.

Study Setting and Population

Participants were board-certified or board-eligible emergency medicine physicians (n = 12), emergency nurses (n = 24), emergency department medical assistants (n = 7), and fourth-year Basic Life Support (BLS)-certified medical students (n = 5) at Harborview Medical Center. At our institution, a TeamSTEPPS Master training site,10 all emergency department staff and trainees receive 4 hours of simulation-based TeamSTEPPS-based training upon hire, with regular refresher training. All nursing, medical assistant, and physician participants received 1 hour of training on the LUCAS device, followed by a 15-minute refresher training before beginning data collection. Training addressed how to operate the LUCAS as well as how to incorporate application of the LUCAS into standard cardiac arrest management. All participants demonstrated competence applying and using the LUCAS device on a Laerdal Resusci Anne (Stavanger, Norway) mannequin before beginning the study. Training and simulations were conducted from July 2013 through September 2013 within a nonclinical space designed to replicate the resuscitation area of an emergency department. Participants received US $50 compensation for their participation.

Study Protocol

Eligible participants provided written consent and were assigned to 4-member teams (1 physician, 2 nurses, 1 medical assistant or medical student) on the basis of schedule availability (Fig. 1). A total of 12 teams were randomized using sealed envelopes containing condition assignment. Participants completed a demographic questionnaire before performing the simulation. Each team performed a simulated cardiac arrest resuscitation using either automated (intervention) or manual (control) chest compressions. All simulations were video recorded using 2 stationary cameras, 1 placed at the foot of the bed to capture all team behaviors and 1 positioned to capture the cardiac monitor. Both video feeds were synchronized and entered into Noldus Observer XT (Leesburg, Va) software for coding of patient management and communication.

Cardiac Arrest Simulation

All simulated resuscitations were performed using a Laerdal (Stavanger, Norway) SimMan human patient simulator and a scripted nurse (actor) who would provide necessary information and prompts to the team when indicated (see figure, Supplemental Digital Content 1, https://links.lww.com/SIH/A313). The scripted actor was trained and monitored during the study to ensure that there was no drift in performance. We modified an event-based cardiac arrest scenario previously used to assess resuscitation team performance.11 Briefly, the simulation involved 2 phases: phase 1 (baseline) required management of an unresponsive, unstable patient with ventricular tachycardia (VT) and phase 2 (arrest) required management of ventricular fibrillation (VF). Phase 1 did not involve cardiac compressions and therefore provided baseline assessment of team communication and patient management regardless of chest compression method. After 2 minutes in phase 1, the simulation progressed to phase 2 regardless of patient management. Teams randomized to perform automated compressions were instructed to use the LUCAS chest device after they communicated their intent to initiate chest compressions. Teams randomized to perform manual compressions did not have access to the LUCAS device even if requested. Simulations ended after 12 minutes in VF.

The need to modify our original scenario necessitated collecting new validity evidence.12 The final scenario (clinical content, behavioral triggers, timing, and expected responses) was content validated by resuscitation experts (n = 11) and piloted using code teams (n = 4) consisting of 1 emergency medicine attending physician, 2 emergency nurses, and 1 medical assistant. Pilot team members were interviewed to establish scenario representativeness, an understanding of diagnostic reasoning during the scenario, degree of psychological fidelity, and team interactions provoked by the scenario. This constituted response process evidence of validity. Finally, reliability of the scenario was assessed throughout the study to ensure that all clinical cues and triggers were presented in a standardized manner for all teams.

Outcome Measures

Communication

Team communication was coded in the following 4 areas defined hereafter: (1) teamwork focus, (2) huddle events, (3) clinical focus, and (4) profession of team member making the statement (Table 1). Statements purely containing sentiments of agreement or disagreement (eg, yup, yes, no, nope) were not coded separately. Statements made by the nurse actor in the simulation were also not coded.

T1
TABLE 1:
Communication Category Definitions
  1. Teamwork focus of each statement was assessed using a coding framework adapted from an existing teamwork taxonomy.13,14 First, statements were determined to be either planning oriented or action oriented. We chose this framework because it is supported by evidence of validity15 and has been successfully used in emergency medicine teams.11
  2. After coding for teamwork focus, team communication was specifically evaluated for the presence of team huddle events.16 Team huddle events are unscheduled, as-needed brief communications that ensure all members of the team quickly establish a shared understanding of the situation and can rapidly adapt to patient decompensation.17 Poor situation awareness in high-risk, highly dynamic teams is associated with errors and adverse events.18,19 In a cardiac arrest resuscitation, huddle events represent attempts to focus team members on diagnostic and treatment options, guide shared decision-making, and build situation awareness for all team members.20 We coded communication for these events because they represent important components of team interaction and could contain both planning and action statements.
  3. The clinical focus of each statement was also coded. We classified all statements into the following 3 categories: (a) statements focused on Advanced Cardiovascular Life Support (ACLS) management, including patient ventilation and endotracheal intubation; (b) statements focused on patient care outside of the ACLS algorithm (patient care [other focus]); and (c) statements focused on coordinating or improving the quality compressions (compression focus). These statements are further defined in Table 1. We selected these categories because we are interested in understanding the quantity of communication focused on chest compression coordination versus other aspects of cardiac arrest management.
  4. Each statement was coded according to which team member made the statement (physician, nurse, assistant).

Patient Management

We used an ACLS guideline–based patient management checklist previously shown to have good interrater reliability and validity (Table 2).11 During phase 1, dichotomous checklist items included 10 critical actions appropriate for the care of a patient. In phase 2, patient management was measured as follows:

T2
TABLE 2:
Patient Management Outcome Definitions
  1. Time to first defibrillation
  2. Time to initiation of compressions
  3. Total time to place LUCAS device
  4. Percent no flow time
  5. Length of intubation
  6. Time to first vasopressor dose
  7. Antiarrhythmic administration adherence.

Data Coding

We used a data coding strategy following evidence-based practices as previously described.21 Briefly, 2 emergency medicine physicians (R.F., M.G.) independently coded video recordings of resuscitations for patient management behaviors and communication using Noldus Observer XT software. Because of the nature of the intervention (LUCAS device), blinding of raters was not possible. Before initiating coding, both raters were trained on practice videos until rater agreement was more than 90%. All items were independently coded in duplicate, and Cohen's kappa was used to estimate initial interrater reliability before resolving any disagreements. For timed items, raters were considered in agreement if they were within 5 seconds of each other's ratings. Disagreements on checklist items and times were resolved via a meeting between the 2 raters to achieve consensus.

Statistical Analyses

Descriptive statistics (median and interquartile range [IQR]) were computed for all outcomes as is recommended for pilot study design.22

RESULTS

Forty-eight participants (12 teams) completed all aspects of the study. Teams were similar with respect to demographics and overall resuscitation experience (Table 3). Outcomes by intervention are detailed in Tables 4, 5.

T3
TABLE 3:
Participant Characteristics
T4
TABLE 4:
Communication Outcomes by Intervention and Profession
T5
TABLE 5:
Patient Management Outcomes

Communication

Interrater reliability across all communication coding was high with an average Cohen's kappa value of 0.66 (Table 4).23 During phase 1 (baseline), the total number of communication statements was similar between both groups. In phase 2, the median number of communication statements in the teams performing manual compressions tended to be greater as compared with the LUCAS intervention group. This was reflected in an increase in the number of planning and, to a lesser extent, action statements focused on ACLS management in the manual compression group. Teams randomized to the control group tended to make more statements focused on chest compressions as compared with intervention teams using the LUCAS device (22.5; IQR, 19.8–39.3 vs. 10.3; IQR, 7.5–13.8). Huddle statements tended to be performed twice as often in teams randomized to perform automated compressions (4.0; IQR, 3.1–4.3) as compared with teams randomized to the control group (2.0; IQR, 1.4–2.6). During phase 1, no huddle statements were made regardless of condition.

In phase 1, the teams in the manual compression group tended to have a greater percentage of communication attributed to nurses (50.4; IQR, 23.0–57.7) as compared with the LUCAS intervention teams (36.3; IQR, 21.0–47.8). This suggests a difference in teams with respect to nursing level of discussion, because phase 1 represents a baseline where conditions were identical for both groups. This difference did not carry over into phase 2, where both conditions had similar levels of communication by profession, suggesting a decrease in communication for nurses once manual compressions were required.

Patient Management

For patient management items, the average Cohen's kappa value of 0.9 for all point events and 0.65 for interval events (no flow time and placement of LUCAS device, Table 5).21 Point events included time to a specific event (eg, initiation of compressions) as well as antiarrhythmic administration adherence. For 2 of the events (no flow time and placement of LUCAS device), there were multiple intervals of time summed to create a total time interval. For example, a team might start to place the LUCAS device, pause, and then resume placement later in the code. Likewise, there were multiple intervals of “no flow time” that had to be summed to create a total no flow time. Reliability of these interval events, taking into account rater variability at each interval, was 0.65.23 During phase 1, patient management scores were similar between intervention (6.8; IQR, 6.4–8.1) and control teams (9.0; IQR, 6.9–9.6). During phase 2, no flow time, time to initiate compressions, and length of intubation were similar in both manual (control) and automated (intervention) compression groups. Teams randomized to the automated compression intervention tended to have increased time to first defibrillation (208.3; IQR, 153.3–222.1 seconds) as compared with control (manual compression) teams (63.2, 30.1, 397.2 seconds) in control teams. Placement of the LUCAS device took a median of 30.2 seconds (IQR, 15.2–40.6) but did not significantly increase no flow time. Eighty-three percent of teams randomized to the automated compression intervention delayed defibrillation until the LUCAS device was in place.

DISCUSSION

This study identified several potential effects of the LUCAS device critical to team communication processes during cardiac arrest resuscitation. We found an overall increase in communication statements made during cardiac arrest by teams performing manual compressions. This was somewhat unexpected and could not be explained solely by an increased need to coordinate the quality and delivery of chest compressions. Rather, the bulk of this difference seemed to be attributable to both planning and action statements containing ACLS or airway management content. We hypothesize that this may be due to challenges associated with nonverbal communication in teams performing manual chest compressions. In the manual compression teams, two of the team members were always engaged in performing chest compressions and likely less aware of the patient's overall situation. Under such conditions, it is possible that the team leader needed to be more directive with more frequent explanation, resulting in an increase in communication statements.

It is important to note that more communication is not necessarily more effective communication. Most teams using the LUCAS device demonstrated multiple huddle events where team communication is directed toward reorienting team members to what has been done, what progress has been made, and which next steps should be taken to diagnose and/or treat the patient. Such team behavior is thought to help team members establish shared understanding of the situation to facilitate problem solving by making thought processes more explicit.24–26 Huddles are thought to improve team efficiency and quality of information sharing.20 This has significant implications for error prevention and optimization of team performance during cardiac resuscitations.

The differences we noted in communication did not translate into a clear difference in patient management during VF arrest. In similar simulation studies, explicit discussion to establish mutual understanding correlated with better performance on ambiguous tasks requiring complex diagnostic decision-making.27 It is possible that a patient scenario containing a larger number of treatment and diagnostic possibilities, for example, pulseless electrical activity, would be more affected by team huddles that establish situational awareness and mutual understanding.24 It is also important to note that we used a simulated platform with a relatively small number of team members who were all present for the entire resuscitation. In actual cardiac resuscitations, team size can be large, with low levels of familiarity among individual team members who often arrive asynchronously.28 Under these difficult conditions, teams that can establish situation awareness are better able to assimilate new team members and use their expertise effectively to accomplish more complex tasks simultaneously.26,29 Extending our work to the clinical environment would allow the evaluation of team communication under more variable conditions.

The more frequent huddle statements noted in teams using the LUCAS device were not associated with a significant decrease in time to first defibrillation. This is not surprising, considering that the scenario was designed to progress to VF after just 2 minutes, making it a relatively early yet significant event. Teams were often still gathering information and attempting to set up early plans and priorities. No team made a huddle statement during phase 1, possibly because they did not appreciate that there was a significant amount of information to share. It is also possible that adding a new technology or device very early in the course of the resuscitation (ie, within 2 minutes of patient arrival) was more cognitively challenging than initiating manual chest compressions, thus delaying recognition and implementation of the true priority—minimal pause defibrillation. Regardless of the cause, identification of potential solutions such as refresher training is important to help mitigate threats to clinical care associated with new technologies such as the LUCAS device.

We found that teams using the LUCAS device delayed defibrillation until they transitioned from manual to automated compressions, potentially delaying initial defibrillation. This is consistent with another simulation-based study in which time to first defibrillation was on average 1 minute longer in resuscitations employing the LUCAS device.30 Since time to first defibrillation is a critical determinant of survival in cardiac arrest,31 recent recommendations for out-of-hospital use of the LUCAS device suggest delaying application until after the first defibrillation.32 This was not specified in the training protocol for practitioners at the study site before this study, underscoring the importance of rigorous device training before implementation.

The impact of technology on individual and team performance is complex, and evaluating this effect is critical to patient care. Often the initial assumption is that automating a task such as chest compressions decreases workload. Because we hypothesized earlier, this could translate into improved teamwork by decreasing the cognitive load associated with coordinating delivery of chest compressions and monitoring compression quality, thus allowing team members to participate in more effective communication. Cognitive load theory supports this idea33–35; however, the tendency to oversimplify and focus on the benefits of task automation can be dangerous. More often than not, task automation shifts a portion of workload to a different task or time.36 For example, although the LUCAS device does remove the need to perform hands-on compressions, the workload associated with initiating compressions is higher, because the process of placing the device is more complex than initiating manual compressions. This likely contributed to our finding that time to first defibrillation was longer in patients using the LUCAS device.

An automated compression device also alters monitoring processes associated with the delivery of chest compressions. Vigilance studies demonstrate that even highly motivated individuals cannot maintain visual attention on a process that contains minimal variance.37 As such, carefully monitoring the timing of compressions can actually become more challenging. Automated compressions also lack physical “alerts” such as fatigue and changes in recoil that can serve as feedback and early warning “alarms” during manual compressions.36 Whether task automation is beneficial is likely dependent on multiple factors, including the characteristics of the task or process being automated, the level of dynamic change in the environment, and the nature of the workload.

Simulation studies such as ours provide an important mechanism to systematically evaluate the downstream effects of technology on patients, providers, and the environment beyond what can be initially understood from clinically based studies. Our findings highlight the benefits of full-scale simulation to the introduction of new devices and the importance of considering how different patient scenarios (ie, witnessed arrest vs. unwitnessed arrest) impact implementation of new technology in the clinical setting. By using a highly standardized, event-based simulation, we were able to identify critical measurement targets and associations for further exploration. We demonstrate the feasibility of using simulation to assess the impact of an automated chest compression device on team communication. In addition, our coding process yielded strong interrater reliability for both communication and patient management measures.

This pilot study provides important insight into how technology may influence team processes and communication during critical events. Although 2 large randomized controlled trials of out-of-hospital cardiac arrest have not demonstrated a benefit for automated CPR over manual CPR with respect to survival,38,39 neither study evaluated teamwork, team communication processes, or medical errors. It will also be important to study automating chest compressions effects in-hospital cardiac arrest. In the latter setting, team structure is much more complex, with increased accessibility of advanced treatment and diagnostic options. It is possible that under these conditions, automation of chest compressions would have a more significant impact on resuscitation team performance. Simulation can directly addresses key knowledge gaps identified in the 2015 Institute of Medicine report on cardiac arrest survival by evaluating the impact of new technologies on team performance during resuscitation.40

Limitations

It is important to note that simulated environments do present limitations to internal and external validity. One of the more surprising findings from this study is related to the wide variability in time to defibrillation, especially within the control conditions. Although this may translate to actual clinical environments, quality assurance–related review of in-emergency department cardiac arrest events at the study institution does not suggest this to be true (unpublished data). It is possible that the team members' knowledge that the situation was simulated decreased the level of urgency with which they approached the situation. In addition, we did not conduct the simulations in situ. Although this had the advantage of allowing multiple camera angles and use of microphones to optimize data capture, it may have resulted in an artificial approach to the resuscitation by the team. The investigators performed an orientation with all subjects before the simulation to remind them to treat the simulation as if it was an actual clinical event, but this may not have been enough to counteract the simulated setting.

This study has some other limitations. First, this is a pilot study and was not powered to detect statistical significance.22 Second, the nature of the LUCAS device is such that we were unable to blind raters to condition during behavioral coding. In addition, raters were familiar with both participants and study objectives. Future research could mitigate threats to internal validity related to rater bias by employing raters without knowledge of study participants or study hypotheses. Finally, we acknowledge that although the use of simulation provided task standardization, we did not replicate the high variability present in actual cardiac arrest patients, nor did we replicate larger interdisciplinary teams that are often involved in resuscitations. As such, our simulations likely underrepresented the complexity of communication and teamwork present when individuals from multiple professions and disciplines are caring for a critical patient.

Future Research

We propose several next steps to expand the proposed study and address key limitations.

  1. Further investigate the possible impact of the LUCAS device on key patient management steps.
  2. This could involve (1) a more detailed assessment of cardiac arrest management that considers the criticality of each step,41 (2) including an ordering of events/clinical actions to better understand what teams are prioritizing in the early part of treatment, and (3) expansion of clinical scenarios to include multiple different clinical events and clinical environments.
  3. Perform a failure mode effects analysis to evaluate potential process failures and identify solutions to mitigate risk. Such analyses would help proactively explore the impact of LUCAS on patient safety.
  4. Formally test key relationships between variables hypothesized to be impacted by automating chest compressions to begin to establish causality. For example, assessing team process beyond communication could help identify the specific team-based behaviors that may influence or be influenced when chest compressions are automated. Use of a well-studied team process assessment tool (eg, Team Emergency Assessment Measure, Mayo High Performance Teamwork Scale)42,43 or team process metrics tied specifically to the scenario11 could help identify effects on coordination, monitoring, planning, and other team behaviors.
  5. Evaluate clinical factors that might mediate or moderate observed effects of the LUCAS technology on team process and patient management.
  6. For example, if the LUCAS device impacts team behaviors by decreasing workload, it would be important to measure workload and formally test this hypothesis. Likewise, if the workload of a clinical event moderates the effect of the LUCAS device on communication, a measure of workload such as the NASA Task Load Index would be helpful.

CONCLUSIONS

This pilot study demonstrates the feasibility of using simulation to study the impact of an automated chest compression device on team communication during cardiac arrest and identifies some dimensions of team communication and patient management that may be impacted by such technology. It will be important to extend this line of research to explore how task automation affects team dynamics and clinical care across a range of cardiac arrest events.

ACKNOWLEDGMENTS

The authors thank the Harborview Medical Center Emergency Department nurses, physicians, students, and staff for their time and expertise as participants.

REFERENCES

1. Falcone RA Jr, Daugherty M, Schweer L, Patterson M, Brown RL, Garcia VF. Multidisciplinary pediatric trauma team training using high-fidelity trauma simulation. J Pediatr Surg 2008;43(6):1065–1071.
2. Stiell IG, Brown SP, Christenson J, et al. What is the role of chest compression depth during out-of-hospital cardiac arrest resuscitation? Crit Care Med 2012;40(4):1192–1198.
3. Christenson J, Andrusiek D, Everson-Stewart S, et al. Chest compression fraction determines survival in patients with out-of-hospital ventricular fibrillation. Circulation 2009;120(13):1241–1247.
4. Vaillancourt C, Everson-Stewart S, Christenson J, et al. The impact of increased chest compression fraction on return of spontaneous circulation for out-of-hospital cardiac arrest patients not in ventricular fibrillation. Resuscitation 2011;82(12):1501–1507.
5. Cheskes S, Schmicker RH, Christenson J, et al. Perishock pause: an independent predictor of survival from out-of-hospital shockable cardiac arrest. Circulation 2011;124(1):58–66.
6. Idris AH, Guffey D, Aufderheide TP, et al. Relationship between chest compression rates and outcomes from cardiac arrest. Circulation 2012;125(24):3004–3012.
7. Weidman EK, Bell G, Walsh D, Small S, Edelson DP. Assessing the impact of immersive simulation on clinical performance during actual in-hospital cardiac arrest with CPR-sensing technology: a randomized feasibility study. Resuscitation 2010;81(11):1556–1561.
8. Fernandez Castelao E, Russo SG, Riethmüller M, Boos M. Effects of team coordination during cardiopulmonary resuscitation: a systematic review of the literature. J Crit Care 2013;28(4):504–521.
9. Berg RA, Hemphill R, Abella BS, et al. Part 5: adult basic life support: 2010 American Heart Association Guidelines for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Circulation 2010;122(18 Suppl 3):S685–S705.
10. TeamSTEPPS National Implementation. Available at: http://www.teamsteppsportal.org/teamstepps-master-training-course/uw. Accessed July 15, 2016.
11. Fernandez R, Pearce M, Grand JA, et al. Evaluation of a computer-based educational intervention to improve medical teamwork and performance during simulated patient resuscitations. Crit Care Med 2013;41(11):2551–2562.
12. Messick S. Validity. In: Linn RL, ed. Educational Measurement. vol. 3. New York: American Council on Education and Macmillan; 1989:13–103.
13. Marks MA, Mathieu JE, Zaccaro SJ. A temporally based framework and taxonomy of team processes. Acad Manage Rev 2001;26(3):356–376.
14. Fernandez R, Kozlowski SW, Shapiro MJ, Salas E. Toward a definition of teamwork in emergency medicine. Acad Emerg Med 2008;15(11):1104–1112.
15. LePine JA, Piccolo RF, Jackson CL, Mathieu JE, Saul JR. A meta-analysis of teamwork processes: tests of a multidimensional model and relationships with team effectiveness criteria. Pers Psychol 2008;61:273–307.
16. Weaver SJ, Rosen MA, DiazGranados D, et al. Does teamwork improve performance in the operating room? A multilevel evaluation. Jt Comm J Qual Patient Saf 2010;36(3):133–142.
17. Webster JS, King HB, Toomey LM, et al. Understanding quality and safety problems in the ambulatory environment: seeking improvement with promising teamwork tools and strategies. In: Henriksen K, Battles J, Keyes M, eds. Advances in Patient Safety: New Directions and Alternative Approaches. vol. 3. Rockville, MD: Agency for Healthcare Research and Quality; 2008.
18. Endsley MR. Toward a theory of situation awareness in dynamic systems. Hum Factors 1995;37(1):32–64.
19. Endsley MR. Measurement of situation awareness in dynamic systems. Hum Factors 1995;37(1):65–84.
20. Goldenhar LM, Brady PW, Sutcliffe KM, Muething SE. Huddling for high reliability and situation awareness. BMJ Qual Saf 2013;22(11):899–906.
21. Feldman M, Lazzara EH, Vanderbilt AA, DiazGranados D. Rater training to support high-stakes simulation-based assessments. J Contin Educ Health Prof 2012;32(4):279–286.
22. Thabane L, Ma J, Chu R, et al. A tutorial on pilot studies: the what, why and how. BMC Med Res Methodol 2010;10:1.
23. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977;33(1):159–174.
24. Waller MJ, Uitdewilligen S. Talking to the room: collective sensemaking during crisis situations. In: Roe R, Waller MJ, Clegg S, eds. Time in Organizational Research. New York, NY: Routledge; 2009:186–203.
25. Kolbe M. Five simple processes that improve high-risk team effectiveness. In: Salas E, Tannenbaum S, Cohen D, Latham G, eds. Developing and Enhancing Teamwork in Organizations: Evidence-Based Practices and Guidelines. San Francisco, CA: Jossey-Bass; 2013.
26. Kolbe M, Grote G, Waller MJ, et al. Monitoring and talking to the room: autochthonous coordination patterns in team interaction and performance. J Appl Psychol 2014;99(6):1254–1267.
27. Tschan F, Semmer NK, Hunziker S, Marsch S. Decisive action vs. joint deliberation: different medical tasks imply different coordination requirements. In: Duffy VG, ed. Advances in Human Factors and Ergonomics in Healthcare. Boca Raton, FL: CRC Press; 2011:191–200.
28. Klein KJ, Ziegert JC, Knight AR, Xiao Y. Dynamic delegation: hierarchical, shared and deindividualized leadership in extreme action teams. Adm Sci Q 2006;51(4):590–621.
29. Waller MJ, Gupta N, Giambatista RC. Effects of adaptive behaviors and shared mental models on control crew performance. Manage Sci 2004;50(11):1534–1544.
30. Putzer G, Braun P, Zimmermann A, et al. LUCAS compared to manual cardiopulmonary resuscitation is more effective during helicopter rescue-a prospective, randomized, cross-over manikin study. Am J Emerg Med 2013;31(2):384–389.
31. Guidelines 2000 for Cardiopulmonary Resuscitation and Emergency Cardiovascular Care. Part 4: the automated external defibrillator: key link in the chain of survival. The American Heart Association in Collaboration with the International Liaison Committee on Resuscitation. Circulation 2000;102(Suppl 8):I60–I76.
32. Rubertsson S, Silfverstolpe J, Rehn L, et al. The Study Protocol for the LINC (LUCAS in Cardiac Arrest) Study: a study comparing conventional adult out-of-hospital cardiopulmonary resuscitation with a concept with mechanical chest compressions and simultaneous defibrillation. Scand J Trauma Resusc Emerg Med 2013;21:5–5.
33. Sweller J. Cognitive load during problem-solving: effects on learning. Cogn Sci 1988;12(2):257–285.
34. Riem N, Boet S, Bould MD, Tavares W, Naik VN. Do technical skills correlate with non-technical skills in crisis resource management: a simulation study. Br J Anaesth 2012;109(5):723–728.
35. Sawyer T, Leonard D, Sierocka-Castaneda A, Chan D, Thompson M. Correlations between technical skills and behavioral skills in simulated neonatal resuscitations. J Perinatol 2014;34(10):781–786.
36. Bainbridge L. Ironies of automation. Automatica 1983;19(6):775–779.
37. Mackworth NH. Researches on the measurement of human performance. Med Res Council Spec Rep Ser 1950;268:156–158.
38. Perkins GD, Lall R, Quinn T, et al. Mechanical versus manual chest compression for out-of-hospital cardiac arrest (PARAMEDIC): a pragmatic, cluster randomised controlled trial. Lancet 2015;385(9972):947–955.
39. Rubertsson S, Lindgren E, Smekal D, et al. Mechanical chest compressions and simultaneous defibrillation vs conventional cardiopulmonary resuscitation in out-of-hospital cardiac arrest: the LINC randomized trial. JAMA 2014;311(1):53–61.
40. Graham R, McCoy M, Schultz A. Institute of Medicine. Strategies to Improve Cardiac Arrest Survival: A Time to Act. Washington, DC: National Academies Press; 2015.
41. Gaba DM, Howard SK, Flanagan B, Smith BE, Fish KJ, Botney R. Assessment of clinical performance during simulated crises using both technical and behavioral ratings. Anesthesiology 1998;89(1):8–18.
42. Cooper S, Cant R, Porter J, et al. Rating medical emergency teamwork performance: development of the Team Emergency Assessment Measure (TEAM). Resuscitation 2010;81(4):446–452.
43. Malec JF, Torsher LC, Dunn WF, et al. The Mayo high performance teamwork scale: reliability and validity for evaluating key crew resource management skills. Simul Healthc 2007;2(1):4–10.
Keywords:

Human factors; Simulation; Cardiac arrest; Teamwork

Copyright © 2017 Society for Simulation in Healthcare