As a maturing field, healthcare simulation faces a number of pedagogical dilemmas that require rigorous exploration to determine what is educationally and emotionally best for our learners.1–5 The use of mannequin death is one such area.4,6 One influential taxonomy divides mannequin death into the following 3 categories: “expected death,” where both facilitators and learners are aware that death is part of the subject matter; “unexpected death,” where mannequin death is planned by facilitators but learners are not alerted; and “death due to action or inaction,” where mannequin death is both unplanned by facilitators and unexpected by learners.7 In this last category, mannequin death is dynamically introduced into a case because of perceived learner error or inappropriate decision making.7 Although little controversy exists regarding the use of simulation to teach learners about managing a patient’s death, mannequin death due to action or inaction has generated a great deal of more disagreement.6,8
Should mannequin death due to learner action or inaction be allowed to occur, or should the learner be protected from death via facilitator interventions that allow the mannequin to survive regardless of their actions? The answer to this question depends on the balance between the psychological stress generated by the event and its subsequent effect on learner education. If the psychological stress generated negatively impacts the educational value of the session, then mannequin death due to action or inaction is best avoided, but if that psychological stress improves fidelity and/or enhances retention of learning points, then the technique may be beneficial.7,9 Studies addressing this issue have shown mixed results. In 1 study, medical students exposed to unexpected mannequin death experienced increased cognitive load and diminished learning.10 In contrast, a study of stressors in simulation among students uncovered a significant positive correlation between imposed emotional stress and long-term (6-month) competency.8 These observations are further supported by a survey of 63 physicians and 175 air medical team members, who opined that mannequin death should occur if care provided during the scenario was inadequate or inappropriate and that the possibility of mannequin death was not a distraction to learners.11 Participants in this study strongly felt that students should not be exempted from the experience of simulated death.11
Each of these studies addresses a different “part of the puzzle,” but little guidance exists on how these separate data points can be integrated into a larger picture. Given the potential power of this technique to generate emotional stress and impact student learning, a global model of mannequin death due to action or inaction (referred to simply as mannequin death in the rest of the manuscript) would be a significant advance in knowledge and could serve to guide practice and shape future research. Our team attempted to address this by developing a pilot theoretical framework addressing the effect of simulator death due to action or inaction on the educational and psychological state of learners. Our simulation program permits facilitators to dynamically employ mannequin death if warranted by learner actions and if the learners were deemed to have sufficient clinical experience and maturity, thus providing a natural laboratory. By developing a pilot model that can be built upon, modified, and tested, our hope was to take a step toward unifying our knowledge about this issue and create a common base for future investigators.
This retrospective study was declared exempt by the University of Louisville Institutional Review Board.
We conducted the study in 2 phases. The first phase focused on the development of an initial theoretical framework via analysis of focus group transcripts using a qualitative, grounded theory approach. Because our goal was to base the model on the perception of learners who had experienced mannequin death, we chose grounded theory because of its ability to create novel models using subject interview data.12 The second phase sought to further refine the framework using a survey-based, mixed methods technique.
Subjects for phase 1 consisted of learners who had experienced mannequin as a part of routine educational activities. A recruitment e-mail was sent to all pediatrics residents, medicine-pediatrics residents, and pediatric fellows at the University of Louisville and all critical care nurses and respiratory therapists in the pediatric intensive care unit at Kosair Children’s Hospital. A total of 11 clinicians responded, who were organized into 3 focus groups (2 composed of nurses and 1 composed of residents and fellows). Participant characteristics are provided in the results section. Focus group questions were prepared by the authors (a group consisting of simulation experts, clinicians, and educators with experience in qualitative analysis) using an iterative development process. Question development focused on minimizing leading language and phrasing questions in a way that would promote open discussion. Topics addressed included how the learner approached the simulation, the death of the mannequin, the postsimulation debriefing, and the impact of the experience on clinical practice.
Focus groups were conducted during a 2-month period by a moderator unaffiliated with the simulation program (to avoid bias) with special training in communication and qualitative methodology. Conversations were audio recorded and transcribed verbatim. Transcripts were then independently analyzed by 3 coders to generate a list of relevant themes.13–16 Coders were selected on the basis of a lack of personal opinion regarding the utility of mannequin death to avoid bias. After analysis, all coders and authors met to triangulate and synthesize the themes. This process involved iterative re-examination of the transcripts and synthesis of concepts. Disagreement was resolved by open discussion with the goal of exploring perceptual differences in transcript themes and using these differences to gain deeper understanding. Given the small sample size, prevalence of initial themes in the 3 transcripts was examined to assess thematic saturation, defined as the point at which no new themes emerge. Once a final theme list was generated, the group discussed how best to link these themes together. A number of organizational schemes were discussed, but ultimately, it was decided to use the temporal flow of a simulation session as a logical framework because we felt that this most closely represented how the learners themselves would perceive relationships between the themes.
Phase 2 began with survey generation using the themes and theoretical framework as a guide for content and structure. An iterative process was again used to create and refine questions, resulting in a final 30-question online instrument containing 5-point Likert scale, multiple choice, and free-response questions. Questions were framed with reference to the most recent simulation session including mannequin death experienced by participants. This survey was subsequently distributed to all learners in our institution via a recruitment e-mail. Explicit statements in the e-mail, questions within the survey itself, and programmable survey logic were used to exclude those learners who had not experienced mannequin death. Those who participated in the focus groups were explicitly included in this phase, because the final survey was substantially more detailed than the initial questions and we did not wish to exclude any potential insights. The final survey is included as Survey Regarding the Experience of Mannequin Death Due to Learner Action or Inaction, Appendix A, http://links.lww.com/SIH/A264, Supplemental Digital Content 1.
Survey Data Analysis
Quantitative data generated by the survey were initially analyzed descriptively. Items that seemed to be correlated were formally compared using Spearman rank-order correlation or χ2 analysis as appropriate. Free-response questions were textually analyzed for common ideas and phrases. Statistical correlations and textual data were used to refine the framework.
A total of 11 learners participated in 3 focus groups. These consisted of 2 pediatrics residents in their second and third years of training, 1 first-year pediatric pulmonology fellow, 1 third-year pediatric critical care fellow, and 7 pediatric intensive care staff nurses with clinical experience ranging 1 to 20 years. Focus group participants experienced 1 to 14 simulations as part of our program. Each learner had experienced at least 1 mannequin death.
Assessment of Saturation
Table 1 shows the prevalence of initial, independently coded themes as distributed among the transcripts. The presence of most initial themes in all 3 transcripts and all themes in at least 2 transcripts suggests that new themes were not emerging with each group and hence that adequate saturation was reached.
Although initial emotional reactions to the mannequin’s death were negative, all participants noted a significant difference in the content of those emotions when reacting to a mannequin death as compared with the death of a real patient. Learners seemed to process the mannequin death primarily as an emotionally difficult learning experience, perhaps similar to failing a test.
Based on transcript content, it was hypothesized that this initial negative reaction stemmed in part from the way individual learners felt they were perceived by the rest of the code team (ie, placing of blame). After the debriefing, however, learners recalled a positive change in both their attitude and emotional response. Focus group participants perceived the experience of mannequin death to have an overall positive impact on long-term retention of learning points and on practice.
Triangulation and analysis resulted in the following final list of themes: (1) learner preparation and suspension of disbelief; (2) differences in emotional response between real and simulated death; (3) effect of simulated death on future emotional engagement with learning process; (4) learner self-perception and perception by team; (5) impact of debriefing on emotive response; (6) impact of debriefing on knowledge and skills learned; and (7) effect of mannequin death on knowledge retention and practice change. These themes are explored in Table 2, supporting quotes are presented in Table 3. The initial model was constructed from these themes.
The survey was completed by 41 participants (19 nurses, 19 residents and fellows, and 3 respiratory therapists). Median clinician experience was 3 years (range, 1–42 years) and each had experienced a median of 1 mannequin death (range, 1–10 with 1 outlier stating they had experienced 20 deaths). Survey data are presented later by theme.
Learner Preparation and Suspension of Disbelief
Sixty-nine percent of learners agreed or strongly agreed that they felt prepared to engage in the session before participation and 66% of learners agreed or strongly agreed that they were able to suspend disbelief. Comments indicated that mannequin appearance, case structure, environmental fidelity, personal perception of the stressfulness of the situation, and engagement of other learners facilitated each learner’s ability to maintain this suspension.
Differences in Emotional Response Between Real and Simulated Death
Although learners were divided on the question of whether mannequin death generates an emotional response (Fig. 1), 80% of respondents reported that mannequin death did not evoke the same emotional response as real patient death. Although learners reported feelings of frustration, disappointment, or embarrassment with mannequin death, they linked this more closely with perceived knowledge or skill deficits than with an emotional identification with the hypothetical patient. One respondent stated “A real death is on a completely different emotional level than the death of a mannequin. When a mannequin dies, I just have a feeling of failure in my clinical skills.” Another respondent stated “A real patient hurts worse. A mannequin elicits similar emotions on a smaller scale. I have difficulty believing anyone could feel the same for a mannequin as they do an actual patient.”
Effects of Simulated Death on Future Emotional Engagement in Learning Activities
Learners were also divided on whether knowledge that mannequin death may occur increases their anxiety about participation before simulation sessions (Fig. 1). This distribution was modestly correlated with the emotional reaction to simulation death (rs = 0.46; P < 0.001; Fig. 2), supporting the possibility of a more emotionally sensitive subpopulation. Eighty-nine percent of participants disagreed or were neutral, however, with the statement that mannequin death made them reluctant to participate in future simulations. However, one learner stated “I felt that I will be more hesitant to take a leadership role in this kind of scenario because I felt a loss of confidence,” suggesting that some carryover of negative emotion to future simulations does occur.
Learner Self-Perception and Perception by Team
Eighty-eight percent of learners had a neutral or positive response toward the team during the simulation, regardless of mannequin death. Similarly, 92% of learners had either neutral or positive perceptions of the team’s response toward them during the simulation.
Impact of Debriefing on Emotive Response to the Simulation
Learners perceived that debriefing significantly altered postsession emotions, with a reduction in frustration (49%–22%, P = 0.02), disappointment (68%–44%, P = 0.045), and embarrassment (49%–19%, P = 0.02). In addition, debriefing was associated with a perceived increase in satisfaction (5%–31%, P = 0.003; Fig. 3). One learner stated “The debriefing is where the mystery gets unraveled. I get to find out where we went wrong, what could have been done differently, and how it was supposed to happen. It’s what makes the experience worthwhile. I think I learn the most during the debriefings after a mannequin death.” Another opined that after the debriefing, “The disappointment was still there, but the experience felt like a learning experience instead of a failure.”
Impact of Debriefing on Knowledge and Skills Learned
Ninety-seven percent of learners agreed or strongly agreed that the debriefing positively impacted their knowledge base. Sixty-six percent of learners agreed or strongly agreed that the death of the mannequin enhanced the learning occurring during debriefing, whereas 25% were neutral regarding this.
Perceived Effect of Mannequin Death in Knowledge Retention and Practice Changes
Ninety-two percent of learners reported that they would use the knowledge gained during these simulations. Similarly, 69% of respondents reported that mannequin death enhanced or significantly enhanced their learning retention, whereas 31% reported that the effect of mannequin death was neutral in terms of their learning retention. Several learners made this link explicit in comments such as “The mannequin dying certainly made me remember hyperkalemia and managing arrhythmias” and “If they can’t die, it’s akin to taking a test you can’t fail. When death is a possibility, it enhances the fidelity of the whole experience.”
The model was further refined on the basis of these results. The final model is presented in Figure 4.
The goal of this study was the development of a pilot model describing the educational and psychological aspects of mannequin death due to learner action or inaction. In this discussion, we explore the findings contributing most to the final model, offer a description of the model itself, explore linkages to the current literature, and address the intersection of our model with issues of psychological safety and adult learning.
The focus groups and subsequent survey uncovered a number of interesting findings that were critical in the development of the final model. First, it seemed that most learners make a sincere attempt to suspend disbelief but nevertheless do not perceive that mannequin death possesses the same emotional valence as that of a real patient. Although learners may engage with us in the “fiction contract,” it seems that the distinction between reality and the “as-if” of the simulation remains clearly in view.17,18 Second, the bimodal distributions with respect to emotions invoked by mannequin death and the effect of possible mannequin death on anticipatory learner anxiety are moderately correlated, suggesting that the same learners potentially experience both reactions. This may represent a particularly sensitive subpopulation, although the nature of that sensitivity is at present unclear and will require further exploration.
Learner perception of the team seemed generally positive despite the death of the mannequin. Indeed, both focus group and survey data suggest that learners are more likely to judge themselves than others. This is reassuring given our initial concerns about the possible negative effects of distress and ill will within the learner group.5 Given the limitations of our sample size, however, we cannot necessarily conclude that such effects never occur.
Perhaps, the most significant finding was the perceived positive effect of debriefing on emotional state. The ability of debriefing to mitigate learner emotion and create psychological safety has long been accepted by the psychology and simulation communities, and these results support this.18–24 These debriefings, however, were conducted by experienced faculty, which is likely reflected in these results. Most also felt that mannequin death positively or neutrally impacted knowledge acquisition during the debriefing process. Although this is a largely reassuring response, it was not completely uniform, because 9% felt that mannequin death detracted from learning. Although small, we do not wish to ignore the significance of this group, because its existence again raised the possibility of a “more sensitive” subpopulation with characteristics requiring further study. Finally, the vast majority of respondents perceived mannequin death to have a positive impact on long-term practice and knowledge. Respondents explicitly stated that the opportunity to experience the consequences of their actions was highly valued and emphasized the learning points of the case.
Explanation of the Model
When developing the model, we strove to maintain a clear focus on learner perceptions, and thus, after much discussion, we chose to use the temporal organization of a typical simulation session as a framing device. By doing this, we hoped to arrange the themes in the sequence that would make the most intuitive sense to learners. The model begins with the recognition of a number of “emotional modulators” that potentially impact the response of learners to an instance of mannequin death. Relevant themes here include theme 1, which addresses the preparatory approaches used by learners to ready themselves for the simulated environment; theme 2, which encapsulates the individual differences that exist in emotional responsiveness; and theme 3, which addresses the possibility of anticipatory anxiety. Arrows delineate the ways in which these themes interact. A linkage of particular note exists between themes 2 and 3, where bidirectional arrows represent the possible subpopulation of emotionally sensitive learners (which bridge these 2 themes) noted in our analysis.
Once the event of mannequin death is experienced, the flow of the model proceeds to the effect of that experience on self and team perception (theme 4). This is linked to the debriefing process, because unaddressed tension within the team could potentially have a significant effect on the debriefing, requiring a degree of facilitator skill to mitigate.1,2,20,25 Because our survey results suggested that this was not a significant issue among our population, we depicted the links between theme 4 and those related to the debriefing using dashed lines to indicate its questionable strength. Study of other populations and institutions will be needed to clarify the strength of this theme. The event of mannequin death also impinges directly upon the emotional and educational aspects of the debriefing, represented in the 5th and 6th themes. These, in turn, influence how the experience affects long-term knowledge and practice (theme 7).
Finally, the entire model is itself situated within the context of previous and subsequent simulation-based educational experiences as represented by the clear boxes framing the model. The influence of previous experiences is envisioned as occurring primarily via interactions with the differing emotional responses and anxiety levels represented by themes 2 and 3. As the vast majority of our survey respondents indicated that the experience of mannequin death did not adversely affect their desire to participate in future educational activities, the link to theme 3 link was also depicted as a dashed line. The influence of these themes on future simulations is represented by the arrows continuing to the right of the diagram.
As a test of the final model, we re-examined existing studies of mannequin death, as we would expect the subjects of these articles to correlate with aspects of the model if it is complete. This comparison is depicted in Table 4. Of the 6 current studies that report empirical data, 5 attempted in some way to assess theme 7.10,11,27,28 Each of these 5 articles also focused on 1 other theme.10,11,26–28 The sixth article addressed theme 2 only.26 Themes 1 (learner preparation and suspension of disbelief) and 4 (learner self-perception and perception by team) were unaddressed. This comparison both supports the applicability of the model and demonstrates how its use could suggest aspects of mannequin death requiring further study.
Exploring External Connections
Although the model was created on the basis solely of learner data, it is helpful to explore parallels between it and other published models of adult education and psychological safety. Indeed, significant parallels exist to the work by David Kolb.29 Kolb’s Cycle depicts adult education as a 4-step process beginning with a concrete experience that then proceeds through active reflection and observation and abstract conceptualization as learners attempt to alter their understanding of that experience. This is completed by a time of active experimentation in which they enact what they have learned, which in turn generates new experiences that begin the cycle anew. With regard to our model, the preparation for (theme 1) and the experience of mannequin death forms the fundamental experience, with reflection on both self and team occurring as part of theme 4. This naturally leads to the conceptual modification that occurs in the debriefing (themes 5 and 6). Active experimentation can be envisioned as occurring during theme 7. The iterative nature of Kolb’s Cycle is captured by themes 2 and 3, which serve to modify the nature of the next experience, as well as the explicit links to past and future sessions.
Aspects of our model also coincide with other recently published frameworks and discussions addressing psychologically difficult simulations.1,4,5,30 One recent model suggests an approach to the ethically difficult simulation in which educators actively consider aspects of the session facilitator, learner, case, educational goal, and institution that might impinge on the psychological safety of learners.30 Although our model has a somewhat different purpose, these considerations are somewhat reflected in themes 2 and 3, where learner-specific differences and their relationship to previous experiences are invoked, and in themes 5 and 6, in which facilitator experience with debriefing plays a crucial role.30
Perhaps the most significant limitation of this study is the initial focus group sample size and its possible effect on saturation. The largely homogenous distribution of initial, independently coded themes among the transcripts, however, suggests that despite the small size, adequate saturation was reached for the purposes of our pilot.
A second cluster of limitations centers on the potential for bias. Given the retrospective nature of the study, a degree of recall bias is possible in the identification of subjects and collection of data. Potential bias also exists in the study population, which consisted of experienced physicians and nurses in pediatric acute care environments, many of whom have experienced real patient death, and among investigators, because our program is one that supports the use of mannequin death. In addition, our faculty primarily uses an “advocacy-inquiry” approach, which may limit the generalizability of our findings if other debriefing models are used (though this seems unlikely).21,22 Although care was taken to minimize these potential biases, the nature of our investigator and study populations render our model a pilot at best, and further study in less experienced populations, at sites where mannequin death is not allowed, and in nonpediatric settings may result in significant modifications or new themes.14,15
In addition, the survey referred to learner emotions in an intentionally ambiguous way. This approach was deliberate, because we did not wish to drive subjects toward specific emotional responses. No preliminary cognitive testing was performed on the survey, however, and thus, we cannot be sure if this achieved the desired effect. Finally, limitations in our program’s demographic databases (which does not record the dates at which learners experience mannequin death) prevent us from calculating true survey response rates or commenting on the effect of the time interval between experiencing mannequin death and study participation.
We report on the creation and exploration of a pilot model addressing the effect of mannequin death due to learner action or inaction. By exploring this issue among learners who have experienced it, we were able to generate themes that encapsulated the psychological and educational effects of this technique. Although the model stems from a limited population, it is our hope that it can serve as scaffolding for future qualitative work with a more diverse population, as well as a means of hypothesis generation for quantitative studies.
The authors thank Michele Morison for her invaluable work in transcribing the focus group recordings.
1. Calhoun AW, Boone MC, Miller KH, et al. Case and commentary: using simulation
to address hierarchy issues during medical crises. Simul Healthc 2013;8(1):13–19.
2. Calhoun AW, Boone MC, Porter MB, et al. Using simulation
to address hierarchy-related errors in medical practice. Perm J 2014;18(2):14–20.
3. Joels M, Pu Z, Wiegert O, et al. Learning
: how does it work? Trends Cogn Sci 2006;10(4):152–158.
4. Gaba DM. Simulations that are challenging to the psyche of participants: how much should we worry and about what? Simul Healthc 8(1):4–7.
5. Truog RD, Meyer EC. Deception and death in medical simulation
. Simul Healthc 8(1):1–3.
6. Corvetto MA, Taekman JM. To die or not to die? A review of simulated death. Simul Healthc 2013;8(1):8–12.
7. Leighton K. Death of a simulator. Clin Simul Nurs 2009;5(2):e59–e62.
8. Demaria S Jr, Bryson EO, Mooney TJ, et al. Adding emotional stressors to training in simulated cardiopulmonary arrest enhances participant performance. Med Educ 44(10):1006–1015.
9. Bryson EO, Levine AI. The simulation
theater: a theoretical discussion of concepts and constructs that enhance learning
. J Crit Care 2008;23(2):185–187.
10. Fraser K, Huffman J, Ma I, et al. The emotional and cognitive impact of unexpected simulated patient death: a randomized controlled trial. Chest 2014;145(5):958–963.
11. Phrampus P, Dorfsman M, Cole J. Death during simulation
training: feedback from trainees [abstract]. In: International Meeting on Medical Simulation
. Miami, FL: Society for Technology in Anesthesia; 2005:95.
12. Giles T, King L, de Lacey S. The timing of the literature review in grounded theory
research: an open mind versus an empty head. ANS Adv Nurs Sci 2013;36(2):E29–E40.
13. Lingard L, Albert M, Levinson W. Grounded theory
, mixed methods, and action research. BMJ 2008;337:a567.
14. Reeves S, Albert M, Kuper A, et al. Why use theories in qualitative
research? BMJ 2008;337:a949.
15. Sargeant J. Qualitative
research part II: participants, analysis, and quality assurance. J Grad Med Educ 2012;4(1):1–3.
16. Sullivan GM, Sargeant J. Qualities of qualitative
research: part I. J Grad Med Educ 2011;3(4):449–452.
17. Dieckmann P, Gaba D, Rall M. Deepening the theoretical foundations of patient simulation
as social practice. Simul Healthc 2007;2(3):183–193.
18. Rudolph JW, Raemer DB, Simon R. Establishing a safe container for learning
: the role of the presimulation briefing. Simul Healthc 2014;9(6):339–349.
19. Arafeh JM, Hansen SS, Nichols A. Debriefing
in simulated-based learning
: facilitating a reflective discussion. J Perinat Neonatal Nurs 24(4):302–309.
20. Lederman LC. Debriefing
: toward a systematic assessment of theory
and practice. Simul Gaming 1992;23(2):145–160.
21. Rudolph JW, Simon R, Dufresne RL, et al. There’s no such thing as “nonjudgmental” debriefing
: a theory
and method for debriefing
with good judgment. Simul Healthc 2006;1(1):49–55.
22. Rudolph JW, Simon R, Rivard P, et al. Debriefing
with good judgment: combining rigorous feedback with genuine inquiry. Anesthesiol Clin 2007;25(2):361–376.
23. Smith SS, Richardson D. Amelioration of deception and harm in psychological research: The important role of debriefing
. J Pers Soc Psychol 1983;44(5):1075–1082.
24. Stocker M, Burmester M, Allen M. Optimisation of simulated team training through the application of learning
theories: a debate for a conceptual framework. BMC Med Educ 2014;14:69.
25. Arafeh JM, Hansen SS, Nichols A. Debriefing
in simulated-based learning
: facilitating a reflective discussion. J Perinat Neonatal Nurs 2010;24(4):302–309.
26. Nickerson M, Pollard M. Simulation
philosophy and practice: simulator patient death versus survival. Clin Simul Nurs 2009;5:e147.
27. Phrampus PE, Cole JS, Phrampus PE. Perceptions of experiencing simulated death. Simul Healthc 2006;1(2):117.
28. Fraser K, Huffman J, Ma I, et al. Death of a manikin: adverse effect on learning
and mechanisms. Chest 2011.
29. Kolb DA, Fry R. Toward an applied theory
of experiential learning
. In: Cooper C, ed. Theories of Group Process. London: John Wiley; 1975.
30. Calhoun AW, Pian-Smith MC, Truog RD, et al. Deception and simulation
education: issues, concepts, and commentary. Simul Healthc 2015;10(3):163–169.