Get 20M+ Full-Text Papers For Less Than $1.50/day. Subscribe now for You or Your Team.

Learn More →

How does moulage contribute to medical students’ perceived engagement in simulation? A mixed-methods pilot study

How does moulage contribute to medical students’ perceived engagement in simulation? A... Introduction: Moulage is used frequently in simulation, with emerging evidence for its use in fields such as paramedicine, radiography and dermatology. It is argued that moulage adds to realism in simulation, although recent work highlighted the ambiguity of moulage practice in simulation. In the absence of knowledge, this study sought to explore the impact of highly authentic moulage on engagement in simulation. Methods: We conducted a randomised mixed-methods study exploring undergraduate medical students’ perception of engagement in relation to the authenticity moulage. Participants were randomised to one of three groups: control (no moulage, narrative only), low authenticity (LowAuth) or high authenticity (HighAuth). Measures included self-report of engagement, the Immersion Scale Reporting Instrument (ISRI), omission of treatment actions, time-to-treat and self-report of authenticity. In combination with these objective measures, we utilised the Stimulated Recall (SR) technique to conduct interviews immediately following the simulation. Results: A total of 33 medical students participated in the study. There was no statistically significant difference between groups on the overall ISRI score. There were statistically significant results between groups on the self- reported engagement measure, and on the treatment actions, time-to-treat measures and the rating of authenticity. Four primary themes ((1) the rules of simulation, (2) believability, (3) consistency of presentation, (4) personal knowledge ) were extracted from the interview analysis, with a further 9 subthemes identified ((1) awareness of simulating, (2) making sense of the context (3) hidden agendas, (4) between two places, (5) dismissing, (6) person centredness, (7) missing information (8) level of training (9) previous experiences). Conclusions: Students rate moulage authenticity highly in simulations. The use of high-authenticity moulage impacts on their prioritisation and task completion. Although the slower performance in the HighAuth group did not have impact on simulated treatment outcomes, highly authentic moulage may be a stronger predictor of performance. Highly authentic moulage is preferable on the basis of optimising learning conditions. Keywords: Moulage, Engagement, Instructional design, Medical education, Realism * Correspondence: Jessica.stokesparish@gmail.com School of Medicine & Public Health, University of Newcastle, Callaghan, New South Wales, Australia Full list of author information is available at the end of the article © The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Stokes-Parish et al. Advances in Simulation (2020) 5:23 Page 2 of 12 Introduction in a significant difference between control and experi- Engagement in simulation is described as a key to suc- mental groups where no moulage versus moulage was cess; if a participant is engaged, the learning/simulation tested in a study on paramedicine students [10]. In this must have “worked”. Grounded in the notion of active study, participants were randomised to two groups (no learning theories such as experiential learning and con- moulage or moulage) and researchers measured task structivism, engaged learners “construct knowledge from immersion, eye-tracking and interviews. Moulage is experience, meaning interpretation and having interac- gaining attention in other fields, such as radiology [11], tions with peers” (Hung et al. 2006). But what is engage- where it has not been explored before, whilst areas like ment? In gaming, engagement is described as being dermatology continue to research the use of moulage as associated with qualities that pull people in [1]. Hung a teaching method for melanoma identification [12, 13]. et al. (2006) describe engaged learning as “authentic”, In other fields of simulation, such as military or defence whereby learners are able to problem-solve, make training, highly authentic moulage is often a de facto in- choices and interact with peers and instructors [2]. clusion that is regarded highly important [14]. Simulation incorporates this in the very nature of its de- We have identified elsewhere the need to explore how livery—participants are given a case they must work moulage contributes to simulation, as opposed to a sort through, often in a group. In simulation, the word of de facto inclusion in simulation instructional design. engagement is often interchanged with the word We propose that moulage fits in the domains of realism “immersion”. Immersion is the “subjective impression suggested by Dieckmann et al. [5]. That is, moulage is that one is participating in a comprehensive, realistic ex- physical (the moulage appears real), semantic (moulage perience” [3]. This highlights the individual part of being is conceptually believable—if A occurs, B will happen, so able to suspend disbelief to participate actively in the therefore I engage) and phenomenal (I emotionally en- simulation. This concept of engagement is echoed by gage with the case because moulage enhances first im- many authors [4–6], yet there has been little discussion pressions). However, we do not understand precisely on what engagement means in the context of simulation. how moulage fits within this framework. A moulage Indeed, Padgett et al. raise this in a critical narrative re- should be believable, make sense to the viewer and not view of the definition of engagement in simulation, in a contradictory manner. We hypothesise that if a agreeing that the term engagement is used loosely and moulaged wound does not match the narrative or if it without clear definition [7]. In their terms: “Learner en- was portrayed inaccurately, this could disrupt the partic- gagement is a context-dependent state of dedicated ipants’ engagement, potentially influencing engagement focus towards a task wherein the learner is involved cog- in learning activity. This hypothesis is supported by lit- nitively, behaviourally, and emotionally” [7]. However, erature where episodes of disengagement occurred in Padgett et al. do not explore gaming literature, the con- simulations where the narrative or setting were not cept of suspending disbelief or the likeness between plausible or factual [15]. immersion and engagement [7]. For the purpose of this The aims of this study were to answer the following study, we have defined engagement as questions: the state in which the participant is observed to be 1. How does the use of moulage authenticity impact actively interacting with the simulation as if it were on engagement of participants in a healthcare real. simulation? 2. What are stakeholders perceptions of the value of With the opposite being true of disengagement, the high and low-authenticity moulage compared to participant is unable to interact as if it were real. none in the educational process? Experts posit strategies to increase engagement through realism. Moulage is increasingly described as a To answer these questions, we had the following way to increase realism in simulation. Defined as “the hypotheses: use of special effects makeup techniques to simulate ill- nesses, bruises, bleeding, wounds or other effects to a Hypothesis 1: Highly authentic moulage causes greater manikin or simulated patient, acting as visual and tactile engagement in simulation participants cues for the learner” [8], moulage is used at varying Hypothesis 2: Poorly authentic moulage causes levels in simulation scenarios. Since the publication of disengagement in simulation participants our commentary, [9], a number of studies have been published to explore its use and benefit in simulation. In the following sections, we describe the methods and One such study by Mills et al. (2018) explored how study design for this work, before moving on to the immersion is influenced by the use of moulage, resulting results. Stokes-Parish et al. Advances in Simulation (2020) 5:23 Page 3 of 12 Methods Orientation to simulation and study Participants Participants were given a standard simulation orienta- We recruited participants from the final 2 years of the tion to the location, including covering the fiction con- undergraduate medical degree (5 years) at the University tract (the process in which the participant agrees to of Newcastle in Australia. Students were eligible to par- interact in the simulation within the set rules of simula- ticipate in the study if they had participated in simula- tion), confidentiality agreement and ground rules for tions previously as a part of their degree. Students were participation in simulation, as per the International not eligible to participate if they had no previous experi- Nursing Association for Clinical Simulation and Learn- ence participating in highly immersive simulations or if ing (INACSL) Standards for Simulation [18] as well as they wore glasses (due to the eye-tracking component of an outline of how the study would flow. At this point, the study, contacts were allowed). the participants signed consent to participate. Following Based on power calculations from previous studies this, eye-tracking equipment was applied and calibrated [16] and the size of a useful or meaningful difference, we (the results of the eye-tracking study will be reported in identified that a sample of 21 participants in the control a later paper). The participant was then familiarised with group and 18 each in the experimental groups would be the manikin, props and surrounding equipment. This in- needed to detect an effect size of 0.8 with a power of cluded talking to the manikin, conducting a physical 90% between control and experimental wings. A slightly examination and meeting the confederate. The partici- larger sample size (n = 23) was required to detect differ- pant was then invited to sit outside the simulation room ences of the same magnitude (0.8) between the two ex- and read the scenario brief. This entire process was perimental conditions. Meta-analysis of over 1500 completed by the Research Assistant. educational interventions suggest that the average effect size for any intervention is 0.4, so effect sizes greater Materials than 0.4 were identified as worth pursuing and reliably Scenario detecting [17]. The scenario chosen was previously assessed for content Recruitment took place via lectures and online post- validity and a peer-reviewed trauma scenario [19]. We ings on the course website. Flyers were placed in the stu- chose to use only one scenario due to the resource- dent common rooms, library and student-teaching areas. intensive nature of the study. The selected scenario was The invitation included information regarding the dur- a male who was brought in by ambulance to the local ation and location of the study, and the study aims. After Emergency Department (ED) following a mountain bike a student expressed interest, they were sent the full Par- accident. Participants were given an ED Admissions ticipant Information Statement and invited to book in a sheet outlining the presenting complaint and were then session at the simulated laboratory. Participants were called in to the scenario by the confederate Endorsed randomised into control and experimental groups to Enrolled Nurse (EEN) to come and review the new pa- participate in a trauma simulation. The control group tient in ED. All study conditions took place in the was narrative case only, whilst experimental groups were Chameleon Simulation Centre in a well-lit room, quiet both narrative case and moulage. That narrative case and devoid of extraneous props. Each participant com- and moulage are described in more detail below. The ex- pleted the scenario individually. A confederate EEN was perimental groups were further randomised to either in the room providing narrative cues and assisting with highly authentic or inauthentic moulage. All data was nursing tasks. The simulated emergency room was set collected in late semester 2 of 2017 and 2018. The study up to replicate local emergency department rooms. The protocol was approved by the University of Newcastle room consisted of a bed, oxygen/air outlets, suction, Human Research Ethics Committee (H-2017-0214). oxygen delivery devices, emergency resuscitation trolley, bed, intravenous (IV) fluids pump, observations monitor, Randomisation standard equipment trolley and a bed. The stock and Participants were given a unique ID code using random- equipment trolleys included mock medications, fluids, izer.com. The Research Assistant generated the random wound dressing supplies and various other medical codes independent of the Chief Investigator (JSP) and al- equipment relevant to trauma scenarios (Fig. 1). located the codes at random to the participants. The The confederate was given training prior to the com- Chief Investigator was only aware of the randomisation mencement of the study. They were taught to trouble- on the day of the study. Participants were told that they shoot technical issues within the scenario, instructed on would be randomised to one of the three groups, but how to respond to the students and were instructed to were blind to the group they were allocated until the not prompt action or correct clinical decisions that they simulation commenced. perceived as errors. The confederate was equipped with Stokes-Parish et al. Advances in Simulation (2020) 5:23 Page 4 of 12 (ANOVA) for 3-group comparisons and t tests for 2- group comparisons. The results are detailed in Table 1. Measures Immersion Video footage of the simulation was reviewed to identify episodes of engagement or disengagement. Using the Immersion Score Rating Instrument (IRSI) [4], the foot- age was reviewed by JSP at a later date and the results were then discussed with the other authors. The ISRI is a tool to measure participant immersion within the simulation. Despite the use of the word immersion, we interpreted the authors’ intent as to measure engage- ment. Although these are subjective measures, we con- sidered them appropriate for the study at hand, Fig. 1 Simulation room particularly since the engagement of participants was measured by additional outcomes—such as eye-tracking a one-way ear piece in which the scenario manager glasses, engagement self-report and stimulated recall could feed information if required. interviews. Variable Clinical markers The only difference between the groups was the appear- Participants’ performance was assessed by means of clin- ance, i.e. moulage or no moulage. ical performance and time-to-treat (that is, how long it In the control group, the manikin had no moulage ap- took them to achieve expected actions). The expected plied. Instead, the confederate would give the participant clinical performance included physical assessment, ad- a verbal cue describing the areas of injury (e.g. “there are ministration of intravenous fluids, ordering an ultra- some grazes and cuts on the face”, and “he has a bruise sound, administration of oxygen and was verified by on his stomach” and “there is a laceration and grazing to expert clinicians elsewhere [19]. This data was collected the left arm”). In the low- and high-authenticity groups, through the Laerdal LLEAP® program and through ob- the confederate only gave verbal cues if the participant servational measures. JSP extracted the Laerdal Scenario requested further information about the wounds (e.g. actions file and then observed the videos and noted ac- "no, there is no active bleeding”). tions taken by participants, including timestamp. These The authenticity of moulage was rated by independent codes were discussed with the other authors throughout clinicians from a variety of specialties using the Moulage the coding process to ensure representativeness. These Authenticity Rating Scale (MARS) [20] (Full makeup ap- observations were compared across groups, by means of plication description in the Appendix). Following reli- difference in time-to-treat and omission of actions. ability testing, we compressed the elements of the MARS into two categories—the Physical and Cognitive Self-report measures Scales of Authenticity. We completed a Comparison of Immediately following the scenario, participants com- Scale Means utilising a one-way analysis of variance pleted a survey to report their perceived engagement Table 1 Expert rating of authenticity Wound Control Mean (n) LowAuth Mean (n) HighAuth Mean (n) Statistical significance* (p < 0.05) Arm Physical 11.2 (6) 17.3 (4) 16.0 (8) p 0.039* Cognitive 9.9 23.8 12.6 p 0.000* All elements 21.2 41.0 28.6 p 0.000* Abdominal Physical 10.5 (5) 15.0 (7) NA p 0.119 Cognitive 13.5 18.6 p 0.213 All 24.0 33.6 p 0.086 Facial Physical 11.5 (5) 16.1 (6) 15.5 (9) p 0.116 Cognitive 19.6 11.8 16.4 p 0.026* All 31.1 28.0 31.9 p 0.563 Stokes-Parish et al. Advances in Simulation (2020) 5:23 Page 5 of 12 with the scenario and the perceived reality of the visual Results cues (face and content validity). This survey was an A total of 33 undergraduate medical students were re- adaptation of the survey used by Pywell et al. [21]. The cruited in the latter half of semester 2 in 2017 and 2018. adaptation included additional questions regarding per- Of these participants, 15 were year 4 medical students ceived engagement and refocusing the questions on face and 18 were year 5. The participants had good exposure and content validity to be trauma based (see Appendix to simulation-based education, including Advanced Life 3). Support training. Twenty-two (66%) of the participants In addition to this, participants rated the authenticity were females and 11 were males (33%). Nine were ran- of moulage using the Moulage Authenticity Rating Scale domised to the control group, 13 to low authenticity (MARS) [20]. Both self-report measures were compared (LowAuth) and 10 to high authenticity (HighAuth). across groups to determine differences, if any. In this section, we break down the relevant results in their measurement groups in the same categories as the methods description. The full data can be seen in the Interviews supplementary file (Appendix 1). Participants were interviewed following the simulation using video-stimulated recall techniques [22]. This Clinical actions method was selected to explore how the moulage au- Clinical actions completed thenticity impacts on participant engagement (H1) and Data were available from 32 of the 33 participants. Data their perceptions of high and low-authenticity moulage from one participant were lost due to a technical glitch. (H2). Stimulated recall techniques are recommended to Groups were compared on the following indices: enhance recall of events and to complement eye- whether they completed hand hygiene at any point dur- tracking methodologies, aligning thoughts with action ing the encounter, requested an ultrasound, ordered IV [23]. The interview questions were structured with a fluids, exposed the abdomen, examined the abdomen, general framework, however, were flexible enough to ex- called for help and investigated or treated the injury plore areas of deeper focus. The central themes of the cues. We performed chi-squared statistics comparing interview focused on engagement and moulage. The whether groups completed expected actions. guide for interviews can be found in Appendix 2. The In these clinical actions, there was a trend of complet- interviews were audiotaped and transcribed verbatim by ing clinical actions in the high-authenticity (HighAuth) a professional academic transcription service. Drawing moulage group as compared to other groups (neuro- from Grounded Theory techniques, the interviews were logical observations, p = .04) and a trend to complete an analysed using a four-phase process. The first phase was abdominal palpation with the low-authenticity moulage familiarisation with the literature (reading transcripts (LowAuth) group (p = .03). When comparing combina- and listening to the audio recording), followed by an ini- tions, there was a statistically significant difference in tial code, then a categorical coding process, and, finally, the control/LowAuth group to conduct an abdominal making meaning. Using a manual process, JSP coded line palpation (p = .02). Differences between all other indices by line, noting sentences and phrases that described the were not significant. See Table 2 for a visual representa- underlying meaning. This continued until saturation was tion of what clinical actions were completed. reached, at which point they were categorised in to groups. Throughout this process, JSP consulted with the Time-to-treat other authors on the coding (BJ, RD), took memos and To determine any differences between groups on the reflective notes to synthesise the evidence and gradually time-to-treat, we conducted one-way ANOVA. In the in- build meaning. stance that a participant did not complete the action, we treated them as if they would have taken the longest Statistics time to complete the action. We compared the three IBM® Statistical Software Package for Social Science groups by one-way ANOVA and further post hoc tests (SPSS v. 23) was used for all statistical comparisons. (Tukey’s HD) where applicable. Statistical significance was defined as a value of 0.05. We The full analysis can be viewed in Table 3.Inexposingthe used one-way ANOVA to compare groups, dependent abdomen, the LowAuth group took the longest (115.92s, SD on the level of measurement of the data (ISRI), time-to- 97.82) and the control group the shortest (69.77 s, SD 43.55). action, MARS, self-reported engagement and used fur- Participants in the HighAuth group took the longest to call ther post hoc tests (Tukey’s) where appropriate to deter- for help (245.1 s, SD 97.71), while the LowAuth group called mine differences between the three groups. We for help the quickest (197.15 s, SD 95.30). When requesting performed chi-squared tests with Fisher’s exact to com- an ultrasound, the HighAuth group ordered it the quickest pare the clinical actions completed. (193 s, SD 82.28) and the control group the slowest (242.66 Stokes-Parish et al. Advances in Simulation (2020) 5:23 Page 6 of 12 Table 2 Clinical actions completed by participant C n = 9 (% in group) LowAuth n = 13 (% in group) HighAuth n = 10 (% in group) Hand hygiene at commencement of scenario 3 (33%) 6 (46%) 6 (60%) Gloves 1 (11%) 2 (15%) 5 (50%) Abdominal ultrasound 2 (22%) 6 (46%) 5 (50%) Intravenous fluids 8 (89%) 12 (92%) 10 (100%) Neuro observations 5 (56%) 1 (1%) 4 (40%) Pathology 5 (56%) 6 (46%) 5 (50%) Abdominal palp 8 (89%) 13 (100%) 6 (60%) Called for help 6 (67%) 9 (70%) 6 (60%) X-ray 3 (33%) 6 (46%) 4 (40%) Investigated injury cues 5 (56%) 7 (54%) 8 (80%) Treated injury cues 1 (11%) 2 (15%) 3 (30%) Participants did not complete any further hand hygiene throughout the scenario Significant differences chi-square between the 3 groups s, SD 38.44). The HighAuth group took the longest to order genders (M/F). In a t test (with Levene’s test for equality of intravenous fluids (174.9 s, SD 88), whilst the LowAuth variances) comparison of moulage (combined LowAuth and group were the quickest (112.41 s, SD 49.69). There were no HighAuth) versus no moulage (control), there was no statis- statistically significant differences between the groups. tically significant difference (Table 5). Despite this lack of sig- nificance, when observing the scatterplot representation of the means, HighAuth had less variability in immersion scores Immersion (see Appendix 1) as compared to both the control and Low- We ran a one-way ANOVA by group of the ISRI (Table 4), Auth group. where the mean score across all experimental groups was 38.59 (SD 14.45). There was no statistically significant differ- ence between the experimental groups. We drilled down fur- Self-report measures ther to explore if there was a difference between Engagement survey undergraduate year and gender. There was no statistically In all groups, the participants felt they were engaged. significant result between year 4 and 5 students or between The participants rated moulage as important in all Table 3 Mean times to action N Mean Std. Std. 95% confidence interval for mean deviation error Lower bound Upper bound Hand hygiene Control 9 71.4 36.1 12.0 43.6 99.2 LowAuth 13 51.7 43.0 11.9 25.7 77.7 HighAuth 10 41.3 42.1 13.3 11.1 71.4 Exposes abdomen Control 9 69.7 43.5 14.5 36.2 103.2 LowAuth 13 115.9 97.8 27.1 56.8 175.0 HighAuth 10 72.4 51.4 16.2 35.6 109.1 Calls for help Control 9 211.3 89.8 29.9 142.2 280.3 LowAuth 13 197.1 95.3 26.4 139.5 254.7 HighAuth 10 245.1 97.7 30.9 175.1 315.0 Orders fast scan Control 9 242.6 38.4 12.8 213.1 272.2 LowAuth 13 220.7 60.0 16.6 184.4 257.0 HighAuth 10 193.0 82.2 26.0 134.1 251.8 Inspects injuries Control 9 214.0 99.4 33.1 137.5 290.4 LowAuth 13 252.0 95.4 26.4 194.4 309.7 HighAuth 10 200.0 82.1 25.9 141.2 258.7 Stokes-Parish et al. Advances in Simulation (2020) 5:23 Page 7 of 12 Table 4 One-way ANOVA of ISRI N Mean Std. Std. 95% confidence interval for mean deviation error Lower bound Upper bound Control 9 33.4 17.1 5.7 20.3 46.6 Experimental group 1 13 43.2 15.9 4.4 33.5 52.8 Experimental group 2 10 37.4 8.2 2.5 31.4 43.2 groups and felt that the lack of moulage did contribute correlated with a rating of authenticity, as opposed to no to disengagement (p = 0.02). When exploring the real- moulage. The full analyses of results are accessible in ism of the scenario, the participants in the HighAuth Appendix 1. group rated the realism higher (p = 0.01) and as repre- sentative of trauma compared to the other groups (p = Interviews 0.00). The moulage contributed to the participant’s abil- Thematic summary ity to treat the simulation as if it were real and made Four primary themes emerged from the participant in- them feel like they were in a real trauma situation (p = terviews, including (1) the rules of simulation, (2) believ- 0.01). The presence of moulage in both the LowAuth ability, (3) consistency of presentation and (4) personal and HighAuth groups contributed to a positive training knowledge. Within these themes, subthemes appeared: experience (p = 0.03). Full results are presented in Ap- (1) awareness of simulating, (2) making sense of the con- pendix 1. text (3) hidden agendas, (4) between two places, (5) dis- missing, (6) person centred-ness, (7) missing Moulage authenticity rating information (8) level of training and (9) previous When comparing participants’ ratings on the authenti- experiences. city of moulage, there were statistically significant differ- ences between groups across the scales. The ANOVA The rules of simulation identified differences between the groups in the physical, Participants described the process of determining the cognitive and all elements scales. Overall, the partici- rules of simulation and learning how to settle into simu- pants rated the moulage as most authentic in the High- lation. They expressed challenges determining if what Auth group when rating the elements individually they were doing was an actual part of the simulation or (position, p = 0.02; detail, p = 0.00; likeness to real a condition of the simulation. Participants described in- world; p = 0.00; colour, p = 0.00; size, p = 0.04) and in stances of attempting to progress through the simulation the global rating of authenticity (p = 0.00). When com- whereby they needed to make sense of the context of paring the physical and cognitive scales where there was simulation, determine if there were hidden agendas; they little difference between the LowAuth and HighAuth demonstrated an experience of being between two places group. In post hoc analysis of the physical scale, there to make meaning of the rules of simulation. That is, they was a statistically significant difference between control were aware they were simulating, yet they were mentally and HighAuth (p = 0.00) and control and LowAuth (p = processing the conditions of simulation versus reality at 0.02). In post hoc analysis of the cognitive scale, there the same time. was a statistically significant difference between control and HighAuth (p = 0.00) and control and LowAuth (p = Awareness of “simulating” 0.00). In the all elements scale, there was statistical sig- The more participants were aware of the simulation, the nificance between groups (p 0.00) and within control vs less engaged they were; meaning they were not necessar- HighAuth (p 0.00), but not control vs LowAuth or Low- ily engaged in learning, but more focussed on determin- Auth vs HighAuth. The use of moulage was strongly ing the rules of simulation. For example, participant 22 (control group) expressed “as soon as I looked and then Table 5 t test comparison of ISRI scores saw it was like crystal clean…it just like kind of pulls you Item (n) Mean (SD) back in, okay it’s a simulation”. Participant 58 (control Year 4 (14) 34.2 (14.9) group) said regarding the lack of moulage and its contri- Year 5 (18) 42 (13.6) bution to engagement “the engagement in believing it was real was less so. Like I took it as oh this is a simula- Male (10) 38.3 (12.5) tion now, I’m going to be doing a simulation…”. Female (22) 38.7 (15.6) They identified that they had a constant background Moulage (9) 33.4 (17.1) awareness that they were simulating, at varying degrees, No moulage (23) 40.6 (13.2) depending on the level of authenticity presented. Stokes-Parish et al. Advances in Simulation (2020) 5:23 Page 8 of 12 Participants described this type of engagement as more engage with the cues presented; when they are confused of a check-box activity, a “going through the motions” as about the cues presented or unsure of the believability, opposed to meaningful learning activity. participants described needing to “step out” to verify the conditions of the simulation –“I did disengage in the “I guess at the back of my mind there's always this sense that I had to then pull myself out of it and thought idea of that this is just a simulation. Yeah. I think I – all right, let’s just evaluate what’s happened, rather wasn't - I don't know, I think I wasn't having that than keep rolling on” (participant 10, LowAuth). feeling, oh okay this is real, I really have to do some- thing about this patient, yeah. It was more like going Believability through the motions” (participant 20, experimental Throughout the interview analysis, participants repeat- group) edly described a desire or need to be able to “believe what they see”. They identified that they wanted visual Making sense of the context cues to be convincing as they felt the cues contributed Participants described attempting to make sense of the to overall engagement and sense of reality. Participants simulated conditions by verifying cues presented, search- expressed that the lack of reality created confusion, lead- ing for additional cues (that they otherwise would not ing them to not take the scenario seriously. The students look for in a real patient) and questioning their own identified that this is a crucial aspect for their learning, judgements. Participant 14 (control group) identified as they felt there was no point in a simulation if it did feeling confused – “…is this the site or am I just imagin- not allow them to practice an assessment in an authentic ing it…I disengage and went into my own thoughts be- way. Participant 40 (HighAuth) describes, “they look cause…I wasn’t 100 percent sure that what I was…an human-like…it sets you up very well for a clinical sce- issue”. This confusion was echoed by participant 39 nario…” and participant 50 (HighAuth) highlights “we’re (LowAuth), they said “…you can’t visualise so you don’t trained to always be looking at the whole page …looking know whether he is supposed to have a bruise or whether for every little detail about the patient to see what you he really doesn’t have any bruise. So you have to can glean about their clinical situation”. Believable mou- assume…”. lage encouraged them to treat the scenario as if it were real and to physically complete actions instead of pre- Hidden agendas tending to. Participants felt there were hidden purposes to the simu- lation itself. In some instances, they described taking the Dismissing confederates’ cues (instead of visually presented) as if to A consistent theme in the interviews was the idea of dis- mean there was importance to the cue, leading them to missing or ignoring the cues if they were delivered ver- pursue that particular path, participant 14 says “oh okay, bally (C) or represented poorly (LowAuth). Participants I’m missing something again”. In their mind, if a confed- in LowAuth expressed they viewed and they assumed erate voiced a cue, there was hidden meaning behind the moulage was unimportant due to the unidimensional it—“they’re telling me about it so it must be the main im- aspect. portant thing” (participant 50, HighAuth). Participants On the inclusion of moulage, participant 40 (High- expressed an expectation that there was something going Auth) says “it just gives it a good indication of where to happen—the patient would “crash” and require emer- they've been hit which you - we don't have otherwise in gency treatment, mostly because prior scenarios they these trauma cases that we get… otherwise you just have were involved in went down the path of cardiopulmo- to ask everything. You don't know what he has and what nary resuscitation. For example, participant 12 (Low- he doesn't have unless you're specifically told…you'd Auth) noted “I thought you’re going to make him crash never ask that or you wouldn't normally ask that in a on me. It’s like a classic”. normal clinical situation because you can see it”. When the reasons for dismissing where explore further, partici- Between two places pants described feeling overloaded with information, This subtheme describes the degree to which partici- causing them to forget –“I missed that cue. I completely pants were aware that they were “in” a simulation. Mul- forgot that the nurse …said that” (participant 21, tiple participants described “stepping in” and “stepping LowAuth). out” of the simulation, for example, participant 12 (Low- Auth) says: “[I]..have to switch out of the scenario to Person-centredness check things out. In real life you can either see it’s hap- Participants described the impact of authentic moulage pening or it’s not”. When they are fully engaged, partici- in terms of how they approached the patient. For ex- pants are able to progress through the simulation and ample, they valued engaging with the patient verbally, Stokes-Parish et al. Advances in Simulation (2020) 5:23 Page 9 of 12 and the moulage provided a trigger to remind them to [Objective Structured Clinical Examinations]…I say take the simulation seriously; in their view, the inter- everything out loud…It’s the worst, it’s so bad clinically”. action became more patient-focused because of the pres- In addition to this, participants described the lack of au- ence of moulage. For example, participant 14 (control) thenticity in previous simulations (non-OSCE type simu- says “I snapped out of the situation again…thinking more lations) lead them to treat future simulations with less in terms of a manikin than a human”. believability. Consistency of presentation “I’ve done previous simulations before where it’s like Participants valued the consistency of presentation of you’re very much, you look at someone and you say visual cues and how the cues interacted with the rest of what are the obs? How is the heart rate, kind of the story. They repeatedly described that the combined thing and you just go from there? And I sort of just cues contributed to how well they engaged in simulation. went back into that … as opposed to actively search- It was not one single aspect that contributed more. ing for wounds or actively feeling the pulse…” (par- ticipant 58, control) Missing information Participants described missing information as a trigger Discussion for disengaging from the simulation. In these instances, The study described sought to explore the potential rela- they described being reminded that it was a simulation, tionship between the authenticity of moulage and par- and that there were limitations. Additionally, they felt ticipant engagement in undergraduate medical students. that missing information was a limitation to learning To our knowledge, this is the first study of its kind in how to assess patients; in their view, authenticity forced any health professions field. In this discussion, we link independent thinking and assisted them to understand the results described above with the hypotheses pre- how they might behave in real life. Participant 1 (High- sented in the introduction and present the potential Auth) says “the moulage is good and it’s showing what links to simulation practice in medical education. it’s meant to…that would be really good, but if it’s just We predicted that higher levels of authenticity would like a sticker or something that says ‘blood here’, then improve participant engagement in simulation (H1). that might detract from the situation because I’m like I’ll This hypothesis was supported by the self-report results, have to ask heaps of questions about that sort of thing”. whereby students rated highly authentic moulage as less likely to contribute to episodes of disengagement and Personal knowledge lack of moulage was likely to contribute to disengage- Personal knowledge was described as a cause of disen- ment (H2). However, participants in all three groups gagement in the simulation. This was two-dimensional: agreed that they felt engaged throughout the scenario, the level of clinical training the participant had and the which makes H1 less plausible. This finding was sup- previous experiences in simulation. ported by the results of the ISRI, in which there were no significant differences between groups. We are unsure if Level of training this is due to the small study size or a true representa- Participants described being unable to progress in the tion. In the scatterplot representation of the ISRI scores, simulation if they got to a point at which they had no the control group had more widely distributed re- experience. For example, deciding what treatment deci- sponses; the pattern of HighAuth results might suggest sion would come next, participant 13 (HighAuth) de- more consistent engagement with the inclusion of au- scribed feeling at the limit of what they could do after thentic moulage than the other groups. These findings attempting to manage the blood pressure: “I disengaged of the authenticity rating scale (MARS) also suggested a little bit here but this is just my lack of knowledge, ra- that some moulage, as opposed to authentic moulage, ther than the actual situation itself”. was sufficient for engagement (further making H1 less plausible). One explanation for the ability to engage re- Previous experiences gardless of authenticity might be that medical students Beyond this, the participants repeatedly referred to their are known to have high levels of motivation—they may previous experiences in simulation and how that influ- already have motivation to engage within a simulation enced their interaction with the moulage. Participants [24]. This was echoed in the interviews with participants, described confusion between conditions of simulations where they talked about an ability to just continue on and simulated assessments. For example, participant 50 and reset their engagement. However, participants also (HighAuth) described simulated formative assessments discussed constantly searching for something to engage where instead of doing the clinical activity, they talked with, either by responding to visual cues or by way of about what they would do –“I’m used to OSCEs dismissing what they were unable to reconcile within the Stokes-Parish et al. Advances in Simulation (2020) 5:23 Page 10 of 12 simulation. This could describe a sort of disengagement, The HighAuth group administered intravenous fluids supporting H2. However, perhaps the ability to engage slower (at least 50 s slower) than the LowAuth and con- despite the level of authenticity is as a result of extrinsic trol groups—again, this might be attributed to the num- motivator factors, whereby the individual is motivated ber of visual cues that needed processing, signalling their by pressure of others (such as the presence of a confed- active engagement with the simulation. These results dif- erate nurse or the continual flow of the simulation) or fer from Mills et al. (2018) where they found in a com- perhaps this is what Padgett et al. refer to in being “fo- parison of moulage versus no moulage, that the cused towards a task” [7]. paramedicine students in a moulage group were quicker A secondary aspect of the study was to explore stu- to respond in time-to-treat [10]. Although there were dents’ perceptions of the authenticity of moulage in differences in these times-to-treat, we do not interpret simulation. All three groups identified that the authenti- them as clinically significant. A 1-min difference in these city of moulage is important in simulation, and partici- items is unlikely to be life-threatening. pants in the control or LowAuth moulage groups did Interestingly, the participants of the HighAuth group not perceive their encounter to be a realistic representa- were more likely to complete neurovascular observations tion of a trauma scenario in the survey. However, their as compared to the other two groups (p =0.05).However, limited exposure to simulation and real trauma may LowAuth were more likely to complete an abdominal pal- have limited their ability to truly rate this. From an op- pation (p = 0.03). The HighAuth group applied gloves posing perspective, perhaps this reinforces the import- more often as compared to the other groups combined. ance of accurate moulage portrayal for inexperienced The trend to conduct an abdominal palpation continued clinicians. Extending on this idea of perceiving authenti- in the combined LowAuth/control group (p =0.02). We city, the participants highlighted the impact of previous considered that in the case of the high-authenticity group simulation authenticity and design; that is, perhaps the completing neurovascular observations, this might have prior exposure to simulation has a stronger impact on been due to the additional visual stimuli of blood that trig- their perception of reality in simulation, than the design gered the need (signalling adequate conceptual realism) to of this simulation itself? check pupillary response and other neurovascular indica- We anticipated that the moulage groups would act tors. As hypothesised early in regard to the abdominal pal- quicker than the control in the time-to-action index. pation, we felt that the absence of other factors (such as This was not supported by the data—in fact, in some in- lacerations and grazing), the participant focused on the stances the time-to-action was slower in the HighAuth visual and audio cues of the abdominal injury. It was un- group. The HighAuth group took (on average) 245.10 s surprising to us that the HighAuth were more likely to to call for help, approximately 30 s longer than the con- apply gloves, students expressed in the interviews “oh I trol group and 50 s longer than LowAuth. Although the saw the blood and thought, I need to put gloves on”;this results were not significant, we hypothesise that High- could have interesting implications for the role of moulage Auth had more visual items to prioritise and consider as in teaching the use of gloves and personal protective a part of their assessment process. Interestingly, the con- equipment (PPE). trol group exposed the abdomen quicker than LowAuth In considering the comparisons of moulage versus no (around 40 s difference), and HighAuth was very similar moulage (control versus LowAuth/HighAuth) and High- to the control group timing (3 s longer). It is plausible Auth versus LowAuth/control, it was interesting that the that this also is due to cue processing and the focus on significant results existed in the latter comparison as op- audible cues may have prioritised their clinical decisions. posed to the first. We interpret this to mean that high- In the interviews, participants identified they focused on authenticity moulage has a more directive effect than certain verbal cues more than others as they believed the low- or no-moulage conditions. perhaps there was hidden meaning in them or the con- In rating the moulage authenticity, participants rated federate was trying to direct them a certain way. This the moulage, or lack of moulage, accordingly. This con- does not explain why the HighAuth group exposed the firmed the ratings from the other self-report. Students abdomen so quickly, perhaps the visual cues on the face consistently rated the control group moulage as low au- and arms may have triggered a need to investigate, dem- thenticity, LowAuth as medium authenticity and High- onstrating the effects of physical and semantic realism. Auth as high authenticity. There has been no previous Alternatively, maybe they found the authentic moulage exploration of moulage authenticity and participants’ distracting; however, this would appear unlikely given interaction with varied levels of moulage. the participants’ discussion in interviews where they expressed the positive views towards moulage being in- Implications for moulage use cluded and the sense of urgency when it was present Although moulage may not impact clinical decisions (demonstrating phenomenal realism). detrimentally, this might not be enough to consider that Stokes-Parish et al. Advances in Simulation (2020) 5:23 Page 11 of 12 moulage is insignificant. As we have seen in the inter- by statistical power calculations. The study was adver- views, participants identified that they spent significant pe- tised with many weeks in advance and delivered at alter- riods of time trying to determine the conditions of nate times that might be suitable for students study simulation. This is supported by work exploring the schedules, including extending the study for an add- process of suspending disbelief (SOD) in nursing students itional year. The simulation centre was based on the whereby authors state “enhanced environmental fidelity same site where students attended classes and place- promotes SOD” [25]. Expanding on this further, if the ments. This provides limitations for the interpretation of conditions of simulation are not consistent across all ex- results—the data results may have been too low to de- posures in a curriculum, it seems that this has impact on tect sizes of effect. Despite this, we did achieve statisti- their ability to suspend disbelief, spending more time on cally significant results that seemed to be accompanied focusing on deciphering the relevance. The underlying by an adequate effect size. We recognise the limitations message here is that consistency across simulations is key. of a single assessor to determine the clinical actions Beyond this, moulage might contribute as a visual cue completed and the time-to-treat information. A more more significantly than expected—as demonstrated by the robust approach might have been to have two assessors participants’ use of gloves and the completion of neuro- to then confirm the reliability of the judgement. This vascular observations and the students’ views. Simulation limitation is also extended to the coding of the inter- provides an opportunity to rehearse clinical practice and views—although Grounded Theory techniques do not develop the ability to manage complex situations. Students typically use multiple coders, we did not utilise the described not taking the simulation seriously or “faking” whole breadth of Grounded Theory. In this instance, it it; what is the implication for this in transferring learning? may have strengthened the work by having a second Although our primary focus on this study was the impact coder. Unfortunately, time and budgetary restraints lim- on engagement, there is a potential link here. If the lack of ited the feasibility of these approaches. authenticity of moulage prompts participants to take The type of scenario used could be a potential limita- shortcuts, then it is worth questioning if we are contribut- tion. Namely, a trauma situation may have more weight ing to negative learning? What we mean by this is the in- on the importance of engagement, as opposed to, for ex- advertent, incorrect messages that we send to participants. ample, a dermatology scenario. Conversely, the urgency In this scenario, no or poorly authentic moulage reduced of a trauma scenario may have enough impetus to en- the likelihood of applying PPE, sending the message that gage participants regardless of the level of authenticity, gloves are unimportant, thereby leading to “habitual un- whilst the authenticity of dermatology might be more safe behaviour” as described by Weller et al. [26]. The important than the authenticity of a trauma simulation. broader result might be an artificial type of learning, which we feel the students alluded to in their comments Conclusions on “doing it for the sake of doing it”. Another extension of Exploring engagement is an emerging topic in simulation, this negative learning might be the example of the slower with new techniques for measurement becoming available. abdominal palpation in the high-authenticity group—by These methods might provide better guides for measuring not exposing participants to real conditions distracting engagement. Other areas of work that should be explored factors, we might be inadvertently training them to only include investigating how the quality of previous simula- look for the obvious. Creating an authentic environment tions determine engagement with scenario, and how mou- is often limited by cost; however, we would argue that not lage influences on so-called negative learning and taking full advantage of simulation (significant expend- developing good clinical habits. Additionally, further work iture is already there) would be a missed opportunity for could be done to explore the relationship of authentic rehearsing clinical practice. moulage and working memory or cognitive load. This Although not generalisable for all situations, the mou- work would be interesting if replicated in a different clin- lage might be better off being authentic. Moulage added ical environment—for example, obstetrics, and other complexity to the scenario. Highly authentic moulage emergency scenarios. And, finally, it would be beneficial might provide more consistent performance behav- to explore the impact of authentic moulage on fully quali- iours—what are the implications of this for high-stakes fied clinicians or in other health professions groups. assessment versus technical skills? Perhaps low-authenti- This study adds to our understanding of the role mou- city moulage is be more confusing than high- lage can play in the participants engagement in simula- authenticity. tion. Within the context of undergraduate medical students, the use of authentic moulage may provide Limitations more consistent patterns of engagement, as compared to Despite repeated efforts to recruit participants, we had no or poor-quality moulage in simulation. Additionally, no success in recruiting the required number indicated moulage may provide a more realistic process of Stokes-Parish et al. Advances in Simulation (2020) 5:23 Page 12 of 12 prioritising care, thereby contributing to deep learning. Received: 28 October 2019 Accepted: 13 August 2020 We suggest that the authenticity of moulage contributes to learner engagement by highlighting the importance of References the activity, allowing them to fully rehearse an activity 1. Benyon, D., P. Turner, and S. Turner, Designing interactive systems: people, activities, contexts, technologies. 2005: Pearson Education. and minimise instances of determining what is real and 2. Hung, D., S.C. Tan, and T.S. Koh, Engaged learning: making learning an what a condition of the simulated environment is. authentic experience, in Engaged learning with emerging technologies. 2006, Springer. p. 29-48. 3. Dede C. Immersive interfaces for engagement and learning. Science. 2009;323. Supplementary information 4. Hagiwara MA, et al. Measuring participants’ immersion in healthcare Supplementary information accompanies this paper at https://doi.org/10. simulation: the development of an instrument. Adv Simul. 2016;1(1):17. 1186/s41077-020-00142-0. 5. Dieckmann P, Gaba D, Rall M. Deepening the theoretical foundations of patient simulation as social practice. Simul Healthc. 2007;2(3):183–93. 6. Rudolph JW, Simon R, Raemer DB. Which reality matters? Questions on the Additional file 1:. Supplementary data file path to high engagement in healthcare simulation. Simul Healthc. 2007;2. Additional file 2:. Stimulated recall interview instructions 7. Padgett J, et al. Engagement: what is it good for? Adv Health Sci Educ Additional file 3:. Participant survey on perceived engagement Theory Pract: The role of learner engagement in healthcare simulation contexts; 2018. 8. Stokes-Parish JB, Duvivier R, Jolly B. Investigating the impact of moulage on simulation engagement — a systematic review. Nurse Educ Today. 2018;64:49–55. Abbreviations 9. Stokes-Parish JB, Duvivier R, Jolly B. Does appearance matter? Current issues LowAuth: Low-authenticity moulage; HighAuth: High-authenticity moulage; and formulation of a research agenda for moulage in simulation. Simul ISRI: Immersion Scale Reporting Instrument; SR: Stimulated Recall; JSP: Jessica Healthc. 2016;22:22. Stokes-Parish; BJ: Brian Jolly; RD: Robbert Duvivier; ED: Emergency 10. Mills BW, et al. Investigating the extent realistic moulage impacts on immersion and Department; EEN: Endorsed Enrolled Nurse; MARS: Moulage Authenticity performance among undergraduate paramedicine students in a simulation-based Rating Scale; INACSL: International Nursing Association for Clinical Simulation trauma scenario a pilot study. Simul Healthc. 2018;13(5):331–40. and Learning; ANOVA: One-way analysis of variance; SPSS: Statistical Software 11. Shiner N, Howard ML. The use of simulation and moulage in undergraduate Package for Social Science; OSCE: Objective Structured Clinical Examination; diagnostic radiography education: a burns scenario. Radiography. 2019. PPE: Personal protective equipment; SOD: Suspension of disbelief 12. Sabzwari S, Afzal A, Nanji K. Mimicking rashes: use of moulage technique in undergraduate assessment at the aga khan university, Karachi. Educ Health. Acknowledgements 2017;30(1):60–3. NA 13. Garg A, et al. The skin cancer objective structured clinical examination (SCOSCE): a multi-institutional collaboration to develop and validate a clinical skills assessment for melanoma. J Am Acad Dermatol. 2015;73(6):959–65. Authors’ contributions 14. Petersen C, et al. Optimization of simulation and Moulage in military-related JSP designed the majority portion of the study, RD and BJ contributed to the medical training. J Spec Oper Med. 2017;17(3):74–80. critique and review of the design. JSP carried out the statistical analysis with the 15. Rystedt H, Sjoblom B. Realism, authenticity, and learning in healthcare supervision of BJ. JSP carried out the qualitative analysis with the supervision of simulations: rules of relevance and irrelevance as interactive achievements. RD. JSP drafted the manuscript. RD and BJ contributed to the review and Instruct Sci. 2012;40(5):785–98. additional content. All authors read and approved the final manuscript. 16. Mills BW, et al. Effects of low- versus high-fidelity simulations on the cognitive burden and performance of entry-level paramedicine students: a mixed-methods Funding comparison trial using eye-tracking, continuous heart rate, difficulty rating scales, This paper is as a result of the work exploring how the authenticity of video observation and interviews. Simul Healthc. 2016;11(1):10–8. moulage influences simulation. The work received an SSH Novice Researcher 17. Hattie, J., Visible learning: A synthesis of over 800 meta-analyses relating to Grant. In addition to the SSH Funding, this work received Seed Funding from achievement. 2008: Routledge. The University of Newcastle, Australia. 18. INACSL. Standards of best practice: simulation<sup>SM</sup> facilitation. Clin Simul Nurs. 2016;12:S16–20. 19. Sarahn Lovett, J.R., Sharyn Hunter, Ian Symonds, Naomi Tomlinson, Robert Availability of data and materials Gagnon, Bernard Charlin, Joerg Mattes, Respective value of the traditional All data generated or analysed during this study are included in this clinical rotation and high fidelity simulation on the acquisition of clinical published article and its supplementary information files. If the data is not reasoning skills in medical students – a randomized Controlled Trial. 2016, available in the supplement, they are available from the corresponding AMEE: MedEdPublish. p. 9. author on reasonable request. 20. Stokes-Parish J, Duvivier R, Jolly B. Expert opinions on the authenticity of moulage in simulation: a Delphi study. Adv Simul. 2019;4(1):16. Ethics approval and consent to participate 21. Pywell MJ, et al. High fidelity, low cost moulage as a valid simulation tool to The study protocol was approved by the University of Newcastle Human improve burns education. Burns. 2016;42(4):844–52. Research Ethics Committee (H-2017-0214). 22. Lyle J. Stimulated recall: a report on its use in naturalistic research. Br Educ Res J. 2003;29(6):861–78. 23. Holmqvist, K., et al., Eye tracking: A comprehensive guide to methods and Consent for publication measures. 2011: OUP Oxford. All participants’ consent forms included consent for publication. 24. Sobral DT. What kind of motivation drives medical students' learning quests? Med Educ. 2004;38(9):950–7. Competing interests 25. Muckler VC. Exploring suspension of disbelief during simulation-based The authors declare that they have no competing interests. learning. Clin Simul Nurs. 2017;13(1):3–9. 26. Weller JM, et al. Simulation in clinical teaching and learning. Med J Aust. Author details 2012;196(9):594. School of Medicine & Public Health, University of Newcastle, Callaghan, New South Wales, Australia. Center for Educational Development and Research in Health Sciences, University Medical Center Groningen, Groningen, The Publisher’sNote Netherlands. School of Rural Medicine, University of New England, Armidale, Springer Nature remains neutral with regard to jurisdictional claims in New South Wales, Australia. published maps and institutional affiliations. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Advances in Simulation Springer Journals

How does moulage contribute to medical students’ perceived engagement in simulation? A mixed-methods pilot study

Loading next page...
 
/lp/springer-journals/how-does-moulage-contribute-to-medical-students-perceived-engagement-v0T6HvOAbV

References (38)

Publisher
Springer Journals
Copyright
Copyright © The Author(s) 2020
eISSN
2059-0628
DOI
10.1186/s41077-020-00142-0
Publisher site
See Article on Publisher Site

Abstract

Introduction: Moulage is used frequently in simulation, with emerging evidence for its use in fields such as paramedicine, radiography and dermatology. It is argued that moulage adds to realism in simulation, although recent work highlighted the ambiguity of moulage practice in simulation. In the absence of knowledge, this study sought to explore the impact of highly authentic moulage on engagement in simulation. Methods: We conducted a randomised mixed-methods study exploring undergraduate medical students’ perception of engagement in relation to the authenticity moulage. Participants were randomised to one of three groups: control (no moulage, narrative only), low authenticity (LowAuth) or high authenticity (HighAuth). Measures included self-report of engagement, the Immersion Scale Reporting Instrument (ISRI), omission of treatment actions, time-to-treat and self-report of authenticity. In combination with these objective measures, we utilised the Stimulated Recall (SR) technique to conduct interviews immediately following the simulation. Results: A total of 33 medical students participated in the study. There was no statistically significant difference between groups on the overall ISRI score. There were statistically significant results between groups on the self- reported engagement measure, and on the treatment actions, time-to-treat measures and the rating of authenticity. Four primary themes ((1) the rules of simulation, (2) believability, (3) consistency of presentation, (4) personal knowledge ) were extracted from the interview analysis, with a further 9 subthemes identified ((1) awareness of simulating, (2) making sense of the context (3) hidden agendas, (4) between two places, (5) dismissing, (6) person centredness, (7) missing information (8) level of training (9) previous experiences). Conclusions: Students rate moulage authenticity highly in simulations. The use of high-authenticity moulage impacts on their prioritisation and task completion. Although the slower performance in the HighAuth group did not have impact on simulated treatment outcomes, highly authentic moulage may be a stronger predictor of performance. Highly authentic moulage is preferable on the basis of optimising learning conditions. Keywords: Moulage, Engagement, Instructional design, Medical education, Realism * Correspondence: Jessica.stokesparish@gmail.com School of Medicine & Public Health, University of Newcastle, Callaghan, New South Wales, Australia Full list of author information is available at the end of the article © The Author(s). 2020 Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data. Stokes-Parish et al. Advances in Simulation (2020) 5:23 Page 2 of 12 Introduction in a significant difference between control and experi- Engagement in simulation is described as a key to suc- mental groups where no moulage versus moulage was cess; if a participant is engaged, the learning/simulation tested in a study on paramedicine students [10]. In this must have “worked”. Grounded in the notion of active study, participants were randomised to two groups (no learning theories such as experiential learning and con- moulage or moulage) and researchers measured task structivism, engaged learners “construct knowledge from immersion, eye-tracking and interviews. Moulage is experience, meaning interpretation and having interac- gaining attention in other fields, such as radiology [11], tions with peers” (Hung et al. 2006). But what is engage- where it has not been explored before, whilst areas like ment? In gaming, engagement is described as being dermatology continue to research the use of moulage as associated with qualities that pull people in [1]. Hung a teaching method for melanoma identification [12, 13]. et al. (2006) describe engaged learning as “authentic”, In other fields of simulation, such as military or defence whereby learners are able to problem-solve, make training, highly authentic moulage is often a de facto in- choices and interact with peers and instructors [2]. clusion that is regarded highly important [14]. Simulation incorporates this in the very nature of its de- We have identified elsewhere the need to explore how livery—participants are given a case they must work moulage contributes to simulation, as opposed to a sort through, often in a group. In simulation, the word of de facto inclusion in simulation instructional design. engagement is often interchanged with the word We propose that moulage fits in the domains of realism “immersion”. Immersion is the “subjective impression suggested by Dieckmann et al. [5]. That is, moulage is that one is participating in a comprehensive, realistic ex- physical (the moulage appears real), semantic (moulage perience” [3]. This highlights the individual part of being is conceptually believable—if A occurs, B will happen, so able to suspend disbelief to participate actively in the therefore I engage) and phenomenal (I emotionally en- simulation. This concept of engagement is echoed by gage with the case because moulage enhances first im- many authors [4–6], yet there has been little discussion pressions). However, we do not understand precisely on what engagement means in the context of simulation. how moulage fits within this framework. A moulage Indeed, Padgett et al. raise this in a critical narrative re- should be believable, make sense to the viewer and not view of the definition of engagement in simulation, in a contradictory manner. We hypothesise that if a agreeing that the term engagement is used loosely and moulaged wound does not match the narrative or if it without clear definition [7]. In their terms: “Learner en- was portrayed inaccurately, this could disrupt the partic- gagement is a context-dependent state of dedicated ipants’ engagement, potentially influencing engagement focus towards a task wherein the learner is involved cog- in learning activity. This hypothesis is supported by lit- nitively, behaviourally, and emotionally” [7]. However, erature where episodes of disengagement occurred in Padgett et al. do not explore gaming literature, the con- simulations where the narrative or setting were not cept of suspending disbelief or the likeness between plausible or factual [15]. immersion and engagement [7]. For the purpose of this The aims of this study were to answer the following study, we have defined engagement as questions: the state in which the participant is observed to be 1. How does the use of moulage authenticity impact actively interacting with the simulation as if it were on engagement of participants in a healthcare real. simulation? 2. What are stakeholders perceptions of the value of With the opposite being true of disengagement, the high and low-authenticity moulage compared to participant is unable to interact as if it were real. none in the educational process? Experts posit strategies to increase engagement through realism. Moulage is increasingly described as a To answer these questions, we had the following way to increase realism in simulation. Defined as “the hypotheses: use of special effects makeup techniques to simulate ill- nesses, bruises, bleeding, wounds or other effects to a Hypothesis 1: Highly authentic moulage causes greater manikin or simulated patient, acting as visual and tactile engagement in simulation participants cues for the learner” [8], moulage is used at varying Hypothesis 2: Poorly authentic moulage causes levels in simulation scenarios. Since the publication of disengagement in simulation participants our commentary, [9], a number of studies have been published to explore its use and benefit in simulation. In the following sections, we describe the methods and One such study by Mills et al. (2018) explored how study design for this work, before moving on to the immersion is influenced by the use of moulage, resulting results. Stokes-Parish et al. Advances in Simulation (2020) 5:23 Page 3 of 12 Methods Orientation to simulation and study Participants Participants were given a standard simulation orienta- We recruited participants from the final 2 years of the tion to the location, including covering the fiction con- undergraduate medical degree (5 years) at the University tract (the process in which the participant agrees to of Newcastle in Australia. Students were eligible to par- interact in the simulation within the set rules of simula- ticipate in the study if they had participated in simula- tion), confidentiality agreement and ground rules for tions previously as a part of their degree. Students were participation in simulation, as per the International not eligible to participate if they had no previous experi- Nursing Association for Clinical Simulation and Learn- ence participating in highly immersive simulations or if ing (INACSL) Standards for Simulation [18] as well as they wore glasses (due to the eye-tracking component of an outline of how the study would flow. At this point, the study, contacts were allowed). the participants signed consent to participate. Following Based on power calculations from previous studies this, eye-tracking equipment was applied and calibrated [16] and the size of a useful or meaningful difference, we (the results of the eye-tracking study will be reported in identified that a sample of 21 participants in the control a later paper). The participant was then familiarised with group and 18 each in the experimental groups would be the manikin, props and surrounding equipment. This in- needed to detect an effect size of 0.8 with a power of cluded talking to the manikin, conducting a physical 90% between control and experimental wings. A slightly examination and meeting the confederate. The partici- larger sample size (n = 23) was required to detect differ- pant was then invited to sit outside the simulation room ences of the same magnitude (0.8) between the two ex- and read the scenario brief. This entire process was perimental conditions. Meta-analysis of over 1500 completed by the Research Assistant. educational interventions suggest that the average effect size for any intervention is 0.4, so effect sizes greater Materials than 0.4 were identified as worth pursuing and reliably Scenario detecting [17]. The scenario chosen was previously assessed for content Recruitment took place via lectures and online post- validity and a peer-reviewed trauma scenario [19]. We ings on the course website. Flyers were placed in the stu- chose to use only one scenario due to the resource- dent common rooms, library and student-teaching areas. intensive nature of the study. The selected scenario was The invitation included information regarding the dur- a male who was brought in by ambulance to the local ation and location of the study, and the study aims. After Emergency Department (ED) following a mountain bike a student expressed interest, they were sent the full Par- accident. Participants were given an ED Admissions ticipant Information Statement and invited to book in a sheet outlining the presenting complaint and were then session at the simulated laboratory. Participants were called in to the scenario by the confederate Endorsed randomised into control and experimental groups to Enrolled Nurse (EEN) to come and review the new pa- participate in a trauma simulation. The control group tient in ED. All study conditions took place in the was narrative case only, whilst experimental groups were Chameleon Simulation Centre in a well-lit room, quiet both narrative case and moulage. That narrative case and devoid of extraneous props. Each participant com- and moulage are described in more detail below. The ex- pleted the scenario individually. A confederate EEN was perimental groups were further randomised to either in the room providing narrative cues and assisting with highly authentic or inauthentic moulage. All data was nursing tasks. The simulated emergency room was set collected in late semester 2 of 2017 and 2018. The study up to replicate local emergency department rooms. The protocol was approved by the University of Newcastle room consisted of a bed, oxygen/air outlets, suction, Human Research Ethics Committee (H-2017-0214). oxygen delivery devices, emergency resuscitation trolley, bed, intravenous (IV) fluids pump, observations monitor, Randomisation standard equipment trolley and a bed. The stock and Participants were given a unique ID code using random- equipment trolleys included mock medications, fluids, izer.com. The Research Assistant generated the random wound dressing supplies and various other medical codes independent of the Chief Investigator (JSP) and al- equipment relevant to trauma scenarios (Fig. 1). located the codes at random to the participants. The The confederate was given training prior to the com- Chief Investigator was only aware of the randomisation mencement of the study. They were taught to trouble- on the day of the study. Participants were told that they shoot technical issues within the scenario, instructed on would be randomised to one of the three groups, but how to respond to the students and were instructed to were blind to the group they were allocated until the not prompt action or correct clinical decisions that they simulation commenced. perceived as errors. The confederate was equipped with Stokes-Parish et al. Advances in Simulation (2020) 5:23 Page 4 of 12 (ANOVA) for 3-group comparisons and t tests for 2- group comparisons. The results are detailed in Table 1. Measures Immersion Video footage of the simulation was reviewed to identify episodes of engagement or disengagement. Using the Immersion Score Rating Instrument (IRSI) [4], the foot- age was reviewed by JSP at a later date and the results were then discussed with the other authors. The ISRI is a tool to measure participant immersion within the simulation. Despite the use of the word immersion, we interpreted the authors’ intent as to measure engage- ment. Although these are subjective measures, we con- sidered them appropriate for the study at hand, Fig. 1 Simulation room particularly since the engagement of participants was measured by additional outcomes—such as eye-tracking a one-way ear piece in which the scenario manager glasses, engagement self-report and stimulated recall could feed information if required. interviews. Variable Clinical markers The only difference between the groups was the appear- Participants’ performance was assessed by means of clin- ance, i.e. moulage or no moulage. ical performance and time-to-treat (that is, how long it In the control group, the manikin had no moulage ap- took them to achieve expected actions). The expected plied. Instead, the confederate would give the participant clinical performance included physical assessment, ad- a verbal cue describing the areas of injury (e.g. “there are ministration of intravenous fluids, ordering an ultra- some grazes and cuts on the face”, and “he has a bruise sound, administration of oxygen and was verified by on his stomach” and “there is a laceration and grazing to expert clinicians elsewhere [19]. This data was collected the left arm”). In the low- and high-authenticity groups, through the Laerdal LLEAP® program and through ob- the confederate only gave verbal cues if the participant servational measures. JSP extracted the Laerdal Scenario requested further information about the wounds (e.g. actions file and then observed the videos and noted ac- "no, there is no active bleeding”). tions taken by participants, including timestamp. These The authenticity of moulage was rated by independent codes were discussed with the other authors throughout clinicians from a variety of specialties using the Moulage the coding process to ensure representativeness. These Authenticity Rating Scale (MARS) [20] (Full makeup ap- observations were compared across groups, by means of plication description in the Appendix). Following reli- difference in time-to-treat and omission of actions. ability testing, we compressed the elements of the MARS into two categories—the Physical and Cognitive Self-report measures Scales of Authenticity. We completed a Comparison of Immediately following the scenario, participants com- Scale Means utilising a one-way analysis of variance pleted a survey to report their perceived engagement Table 1 Expert rating of authenticity Wound Control Mean (n) LowAuth Mean (n) HighAuth Mean (n) Statistical significance* (p < 0.05) Arm Physical 11.2 (6) 17.3 (4) 16.0 (8) p 0.039* Cognitive 9.9 23.8 12.6 p 0.000* All elements 21.2 41.0 28.6 p 0.000* Abdominal Physical 10.5 (5) 15.0 (7) NA p 0.119 Cognitive 13.5 18.6 p 0.213 All 24.0 33.6 p 0.086 Facial Physical 11.5 (5) 16.1 (6) 15.5 (9) p 0.116 Cognitive 19.6 11.8 16.4 p 0.026* All 31.1 28.0 31.9 p 0.563 Stokes-Parish et al. Advances in Simulation (2020) 5:23 Page 5 of 12 with the scenario and the perceived reality of the visual Results cues (face and content validity). This survey was an A total of 33 undergraduate medical students were re- adaptation of the survey used by Pywell et al. [21]. The cruited in the latter half of semester 2 in 2017 and 2018. adaptation included additional questions regarding per- Of these participants, 15 were year 4 medical students ceived engagement and refocusing the questions on face and 18 were year 5. The participants had good exposure and content validity to be trauma based (see Appendix to simulation-based education, including Advanced Life 3). Support training. Twenty-two (66%) of the participants In addition to this, participants rated the authenticity were females and 11 were males (33%). Nine were ran- of moulage using the Moulage Authenticity Rating Scale domised to the control group, 13 to low authenticity (MARS) [20]. Both self-report measures were compared (LowAuth) and 10 to high authenticity (HighAuth). across groups to determine differences, if any. In this section, we break down the relevant results in their measurement groups in the same categories as the methods description. The full data can be seen in the Interviews supplementary file (Appendix 1). Participants were interviewed following the simulation using video-stimulated recall techniques [22]. This Clinical actions method was selected to explore how the moulage au- Clinical actions completed thenticity impacts on participant engagement (H1) and Data were available from 32 of the 33 participants. Data their perceptions of high and low-authenticity moulage from one participant were lost due to a technical glitch. (H2). Stimulated recall techniques are recommended to Groups were compared on the following indices: enhance recall of events and to complement eye- whether they completed hand hygiene at any point dur- tracking methodologies, aligning thoughts with action ing the encounter, requested an ultrasound, ordered IV [23]. The interview questions were structured with a fluids, exposed the abdomen, examined the abdomen, general framework, however, were flexible enough to ex- called for help and investigated or treated the injury plore areas of deeper focus. The central themes of the cues. We performed chi-squared statistics comparing interview focused on engagement and moulage. The whether groups completed expected actions. guide for interviews can be found in Appendix 2. The In these clinical actions, there was a trend of complet- interviews were audiotaped and transcribed verbatim by ing clinical actions in the high-authenticity (HighAuth) a professional academic transcription service. Drawing moulage group as compared to other groups (neuro- from Grounded Theory techniques, the interviews were logical observations, p = .04) and a trend to complete an analysed using a four-phase process. The first phase was abdominal palpation with the low-authenticity moulage familiarisation with the literature (reading transcripts (LowAuth) group (p = .03). When comparing combina- and listening to the audio recording), followed by an ini- tions, there was a statistically significant difference in tial code, then a categorical coding process, and, finally, the control/LowAuth group to conduct an abdominal making meaning. Using a manual process, JSP coded line palpation (p = .02). Differences between all other indices by line, noting sentences and phrases that described the were not significant. See Table 2 for a visual representa- underlying meaning. This continued until saturation was tion of what clinical actions were completed. reached, at which point they were categorised in to groups. Throughout this process, JSP consulted with the Time-to-treat other authors on the coding (BJ, RD), took memos and To determine any differences between groups on the reflective notes to synthesise the evidence and gradually time-to-treat, we conducted one-way ANOVA. In the in- build meaning. stance that a participant did not complete the action, we treated them as if they would have taken the longest Statistics time to complete the action. We compared the three IBM® Statistical Software Package for Social Science groups by one-way ANOVA and further post hoc tests (SPSS v. 23) was used for all statistical comparisons. (Tukey’s HD) where applicable. Statistical significance was defined as a value of 0.05. We The full analysis can be viewed in Table 3.Inexposingthe used one-way ANOVA to compare groups, dependent abdomen, the LowAuth group took the longest (115.92s, SD on the level of measurement of the data (ISRI), time-to- 97.82) and the control group the shortest (69.77 s, SD 43.55). action, MARS, self-reported engagement and used fur- Participants in the HighAuth group took the longest to call ther post hoc tests (Tukey’s) where appropriate to deter- for help (245.1 s, SD 97.71), while the LowAuth group called mine differences between the three groups. We for help the quickest (197.15 s, SD 95.30). When requesting performed chi-squared tests with Fisher’s exact to com- an ultrasound, the HighAuth group ordered it the quickest pare the clinical actions completed. (193 s, SD 82.28) and the control group the slowest (242.66 Stokes-Parish et al. Advances in Simulation (2020) 5:23 Page 6 of 12 Table 2 Clinical actions completed by participant C n = 9 (% in group) LowAuth n = 13 (% in group) HighAuth n = 10 (% in group) Hand hygiene at commencement of scenario 3 (33%) 6 (46%) 6 (60%) Gloves 1 (11%) 2 (15%) 5 (50%) Abdominal ultrasound 2 (22%) 6 (46%) 5 (50%) Intravenous fluids 8 (89%) 12 (92%) 10 (100%) Neuro observations 5 (56%) 1 (1%) 4 (40%) Pathology 5 (56%) 6 (46%) 5 (50%) Abdominal palp 8 (89%) 13 (100%) 6 (60%) Called for help 6 (67%) 9 (70%) 6 (60%) X-ray 3 (33%) 6 (46%) 4 (40%) Investigated injury cues 5 (56%) 7 (54%) 8 (80%) Treated injury cues 1 (11%) 2 (15%) 3 (30%) Participants did not complete any further hand hygiene throughout the scenario Significant differences chi-square between the 3 groups s, SD 38.44). The HighAuth group took the longest to order genders (M/F). In a t test (with Levene’s test for equality of intravenous fluids (174.9 s, SD 88), whilst the LowAuth variances) comparison of moulage (combined LowAuth and group were the quickest (112.41 s, SD 49.69). There were no HighAuth) versus no moulage (control), there was no statis- statistically significant differences between the groups. tically significant difference (Table 5). Despite this lack of sig- nificance, when observing the scatterplot representation of the means, HighAuth had less variability in immersion scores Immersion (see Appendix 1) as compared to both the control and Low- We ran a one-way ANOVA by group of the ISRI (Table 4), Auth group. where the mean score across all experimental groups was 38.59 (SD 14.45). There was no statistically significant differ- ence between the experimental groups. We drilled down fur- Self-report measures ther to explore if there was a difference between Engagement survey undergraduate year and gender. There was no statistically In all groups, the participants felt they were engaged. significant result between year 4 and 5 students or between The participants rated moulage as important in all Table 3 Mean times to action N Mean Std. Std. 95% confidence interval for mean deviation error Lower bound Upper bound Hand hygiene Control 9 71.4 36.1 12.0 43.6 99.2 LowAuth 13 51.7 43.0 11.9 25.7 77.7 HighAuth 10 41.3 42.1 13.3 11.1 71.4 Exposes abdomen Control 9 69.7 43.5 14.5 36.2 103.2 LowAuth 13 115.9 97.8 27.1 56.8 175.0 HighAuth 10 72.4 51.4 16.2 35.6 109.1 Calls for help Control 9 211.3 89.8 29.9 142.2 280.3 LowAuth 13 197.1 95.3 26.4 139.5 254.7 HighAuth 10 245.1 97.7 30.9 175.1 315.0 Orders fast scan Control 9 242.6 38.4 12.8 213.1 272.2 LowAuth 13 220.7 60.0 16.6 184.4 257.0 HighAuth 10 193.0 82.2 26.0 134.1 251.8 Inspects injuries Control 9 214.0 99.4 33.1 137.5 290.4 LowAuth 13 252.0 95.4 26.4 194.4 309.7 HighAuth 10 200.0 82.1 25.9 141.2 258.7 Stokes-Parish et al. Advances in Simulation (2020) 5:23 Page 7 of 12 Table 4 One-way ANOVA of ISRI N Mean Std. Std. 95% confidence interval for mean deviation error Lower bound Upper bound Control 9 33.4 17.1 5.7 20.3 46.6 Experimental group 1 13 43.2 15.9 4.4 33.5 52.8 Experimental group 2 10 37.4 8.2 2.5 31.4 43.2 groups and felt that the lack of moulage did contribute correlated with a rating of authenticity, as opposed to no to disengagement (p = 0.02). When exploring the real- moulage. The full analyses of results are accessible in ism of the scenario, the participants in the HighAuth Appendix 1. group rated the realism higher (p = 0.01) and as repre- sentative of trauma compared to the other groups (p = Interviews 0.00). The moulage contributed to the participant’s abil- Thematic summary ity to treat the simulation as if it were real and made Four primary themes emerged from the participant in- them feel like they were in a real trauma situation (p = terviews, including (1) the rules of simulation, (2) believ- 0.01). The presence of moulage in both the LowAuth ability, (3) consistency of presentation and (4) personal and HighAuth groups contributed to a positive training knowledge. Within these themes, subthemes appeared: experience (p = 0.03). Full results are presented in Ap- (1) awareness of simulating, (2) making sense of the con- pendix 1. text (3) hidden agendas, (4) between two places, (5) dis- missing, (6) person centred-ness, (7) missing Moulage authenticity rating information (8) level of training and (9) previous When comparing participants’ ratings on the authenti- experiences. city of moulage, there were statistically significant differ- ences between groups across the scales. The ANOVA The rules of simulation identified differences between the groups in the physical, Participants described the process of determining the cognitive and all elements scales. Overall, the partici- rules of simulation and learning how to settle into simu- pants rated the moulage as most authentic in the High- lation. They expressed challenges determining if what Auth group when rating the elements individually they were doing was an actual part of the simulation or (position, p = 0.02; detail, p = 0.00; likeness to real a condition of the simulation. Participants described in- world; p = 0.00; colour, p = 0.00; size, p = 0.04) and in stances of attempting to progress through the simulation the global rating of authenticity (p = 0.00). When com- whereby they needed to make sense of the context of paring the physical and cognitive scales where there was simulation, determine if there were hidden agendas; they little difference between the LowAuth and HighAuth demonstrated an experience of being between two places group. In post hoc analysis of the physical scale, there to make meaning of the rules of simulation. That is, they was a statistically significant difference between control were aware they were simulating, yet they were mentally and HighAuth (p = 0.00) and control and LowAuth (p = processing the conditions of simulation versus reality at 0.02). In post hoc analysis of the cognitive scale, there the same time. was a statistically significant difference between control and HighAuth (p = 0.00) and control and LowAuth (p = Awareness of “simulating” 0.00). In the all elements scale, there was statistical sig- The more participants were aware of the simulation, the nificance between groups (p 0.00) and within control vs less engaged they were; meaning they were not necessar- HighAuth (p 0.00), but not control vs LowAuth or Low- ily engaged in learning, but more focussed on determin- Auth vs HighAuth. The use of moulage was strongly ing the rules of simulation. For example, participant 22 (control group) expressed “as soon as I looked and then Table 5 t test comparison of ISRI scores saw it was like crystal clean…it just like kind of pulls you Item (n) Mean (SD) back in, okay it’s a simulation”. Participant 58 (control Year 4 (14) 34.2 (14.9) group) said regarding the lack of moulage and its contri- Year 5 (18) 42 (13.6) bution to engagement “the engagement in believing it was real was less so. Like I took it as oh this is a simula- Male (10) 38.3 (12.5) tion now, I’m going to be doing a simulation…”. Female (22) 38.7 (15.6) They identified that they had a constant background Moulage (9) 33.4 (17.1) awareness that they were simulating, at varying degrees, No moulage (23) 40.6 (13.2) depending on the level of authenticity presented. Stokes-Parish et al. Advances in Simulation (2020) 5:23 Page 8 of 12 Participants described this type of engagement as more engage with the cues presented; when they are confused of a check-box activity, a “going through the motions” as about the cues presented or unsure of the believability, opposed to meaningful learning activity. participants described needing to “step out” to verify the conditions of the simulation –“I did disengage in the “I guess at the back of my mind there's always this sense that I had to then pull myself out of it and thought idea of that this is just a simulation. Yeah. I think I – all right, let’s just evaluate what’s happened, rather wasn't - I don't know, I think I wasn't having that than keep rolling on” (participant 10, LowAuth). feeling, oh okay this is real, I really have to do some- thing about this patient, yeah. It was more like going Believability through the motions” (participant 20, experimental Throughout the interview analysis, participants repeat- group) edly described a desire or need to be able to “believe what they see”. They identified that they wanted visual Making sense of the context cues to be convincing as they felt the cues contributed Participants described attempting to make sense of the to overall engagement and sense of reality. Participants simulated conditions by verifying cues presented, search- expressed that the lack of reality created confusion, lead- ing for additional cues (that they otherwise would not ing them to not take the scenario seriously. The students look for in a real patient) and questioning their own identified that this is a crucial aspect for their learning, judgements. Participant 14 (control group) identified as they felt there was no point in a simulation if it did feeling confused – “…is this the site or am I just imagin- not allow them to practice an assessment in an authentic ing it…I disengage and went into my own thoughts be- way. Participant 40 (HighAuth) describes, “they look cause…I wasn’t 100 percent sure that what I was…an human-like…it sets you up very well for a clinical sce- issue”. This confusion was echoed by participant 39 nario…” and participant 50 (HighAuth) highlights “we’re (LowAuth), they said “…you can’t visualise so you don’t trained to always be looking at the whole page …looking know whether he is supposed to have a bruise or whether for every little detail about the patient to see what you he really doesn’t have any bruise. So you have to can glean about their clinical situation”. Believable mou- assume…”. lage encouraged them to treat the scenario as if it were real and to physically complete actions instead of pre- Hidden agendas tending to. Participants felt there were hidden purposes to the simu- lation itself. In some instances, they described taking the Dismissing confederates’ cues (instead of visually presented) as if to A consistent theme in the interviews was the idea of dis- mean there was importance to the cue, leading them to missing or ignoring the cues if they were delivered ver- pursue that particular path, participant 14 says “oh okay, bally (C) or represented poorly (LowAuth). Participants I’m missing something again”. In their mind, if a confed- in LowAuth expressed they viewed and they assumed erate voiced a cue, there was hidden meaning behind the moulage was unimportant due to the unidimensional it—“they’re telling me about it so it must be the main im- aspect. portant thing” (participant 50, HighAuth). Participants On the inclusion of moulage, participant 40 (High- expressed an expectation that there was something going Auth) says “it just gives it a good indication of where to happen—the patient would “crash” and require emer- they've been hit which you - we don't have otherwise in gency treatment, mostly because prior scenarios they these trauma cases that we get… otherwise you just have were involved in went down the path of cardiopulmo- to ask everything. You don't know what he has and what nary resuscitation. For example, participant 12 (Low- he doesn't have unless you're specifically told…you'd Auth) noted “I thought you’re going to make him crash never ask that or you wouldn't normally ask that in a on me. It’s like a classic”. normal clinical situation because you can see it”. When the reasons for dismissing where explore further, partici- Between two places pants described feeling overloaded with information, This subtheme describes the degree to which partici- causing them to forget –“I missed that cue. I completely pants were aware that they were “in” a simulation. Mul- forgot that the nurse …said that” (participant 21, tiple participants described “stepping in” and “stepping LowAuth). out” of the simulation, for example, participant 12 (Low- Auth) says: “[I]..have to switch out of the scenario to Person-centredness check things out. In real life you can either see it’s hap- Participants described the impact of authentic moulage pening or it’s not”. When they are fully engaged, partici- in terms of how they approached the patient. For ex- pants are able to progress through the simulation and ample, they valued engaging with the patient verbally, Stokes-Parish et al. Advances in Simulation (2020) 5:23 Page 9 of 12 and the moulage provided a trigger to remind them to [Objective Structured Clinical Examinations]…I say take the simulation seriously; in their view, the inter- everything out loud…It’s the worst, it’s so bad clinically”. action became more patient-focused because of the pres- In addition to this, participants described the lack of au- ence of moulage. For example, participant 14 (control) thenticity in previous simulations (non-OSCE type simu- says “I snapped out of the situation again…thinking more lations) lead them to treat future simulations with less in terms of a manikin than a human”. believability. Consistency of presentation “I’ve done previous simulations before where it’s like Participants valued the consistency of presentation of you’re very much, you look at someone and you say visual cues and how the cues interacted with the rest of what are the obs? How is the heart rate, kind of the story. They repeatedly described that the combined thing and you just go from there? And I sort of just cues contributed to how well they engaged in simulation. went back into that … as opposed to actively search- It was not one single aspect that contributed more. ing for wounds or actively feeling the pulse…” (par- ticipant 58, control) Missing information Participants described missing information as a trigger Discussion for disengaging from the simulation. In these instances, The study described sought to explore the potential rela- they described being reminded that it was a simulation, tionship between the authenticity of moulage and par- and that there were limitations. Additionally, they felt ticipant engagement in undergraduate medical students. that missing information was a limitation to learning To our knowledge, this is the first study of its kind in how to assess patients; in their view, authenticity forced any health professions field. In this discussion, we link independent thinking and assisted them to understand the results described above with the hypotheses pre- how they might behave in real life. Participant 1 (High- sented in the introduction and present the potential Auth) says “the moulage is good and it’s showing what links to simulation practice in medical education. it’s meant to…that would be really good, but if it’s just We predicted that higher levels of authenticity would like a sticker or something that says ‘blood here’, then improve participant engagement in simulation (H1). that might detract from the situation because I’m like I’ll This hypothesis was supported by the self-report results, have to ask heaps of questions about that sort of thing”. whereby students rated highly authentic moulage as less likely to contribute to episodes of disengagement and Personal knowledge lack of moulage was likely to contribute to disengage- Personal knowledge was described as a cause of disen- ment (H2). However, participants in all three groups gagement in the simulation. This was two-dimensional: agreed that they felt engaged throughout the scenario, the level of clinical training the participant had and the which makes H1 less plausible. This finding was sup- previous experiences in simulation. ported by the results of the ISRI, in which there were no significant differences between groups. We are unsure if Level of training this is due to the small study size or a true representa- Participants described being unable to progress in the tion. In the scatterplot representation of the ISRI scores, simulation if they got to a point at which they had no the control group had more widely distributed re- experience. For example, deciding what treatment deci- sponses; the pattern of HighAuth results might suggest sion would come next, participant 13 (HighAuth) de- more consistent engagement with the inclusion of au- scribed feeling at the limit of what they could do after thentic moulage than the other groups. These findings attempting to manage the blood pressure: “I disengaged of the authenticity rating scale (MARS) also suggested a little bit here but this is just my lack of knowledge, ra- that some moulage, as opposed to authentic moulage, ther than the actual situation itself”. was sufficient for engagement (further making H1 less plausible). One explanation for the ability to engage re- Previous experiences gardless of authenticity might be that medical students Beyond this, the participants repeatedly referred to their are known to have high levels of motivation—they may previous experiences in simulation and how that influ- already have motivation to engage within a simulation enced their interaction with the moulage. Participants [24]. This was echoed in the interviews with participants, described confusion between conditions of simulations where they talked about an ability to just continue on and simulated assessments. For example, participant 50 and reset their engagement. However, participants also (HighAuth) described simulated formative assessments discussed constantly searching for something to engage where instead of doing the clinical activity, they talked with, either by responding to visual cues or by way of about what they would do –“I’m used to OSCEs dismissing what they were unable to reconcile within the Stokes-Parish et al. Advances in Simulation (2020) 5:23 Page 10 of 12 simulation. This could describe a sort of disengagement, The HighAuth group administered intravenous fluids supporting H2. However, perhaps the ability to engage slower (at least 50 s slower) than the LowAuth and con- despite the level of authenticity is as a result of extrinsic trol groups—again, this might be attributed to the num- motivator factors, whereby the individual is motivated ber of visual cues that needed processing, signalling their by pressure of others (such as the presence of a confed- active engagement with the simulation. These results dif- erate nurse or the continual flow of the simulation) or fer from Mills et al. (2018) where they found in a com- perhaps this is what Padgett et al. refer to in being “fo- parison of moulage versus no moulage, that the cused towards a task” [7]. paramedicine students in a moulage group were quicker A secondary aspect of the study was to explore stu- to respond in time-to-treat [10]. Although there were dents’ perceptions of the authenticity of moulage in differences in these times-to-treat, we do not interpret simulation. All three groups identified that the authenti- them as clinically significant. A 1-min difference in these city of moulage is important in simulation, and partici- items is unlikely to be life-threatening. pants in the control or LowAuth moulage groups did Interestingly, the participants of the HighAuth group not perceive their encounter to be a realistic representa- were more likely to complete neurovascular observations tion of a trauma scenario in the survey. However, their as compared to the other two groups (p =0.05).However, limited exposure to simulation and real trauma may LowAuth were more likely to complete an abdominal pal- have limited their ability to truly rate this. From an op- pation (p = 0.03). The HighAuth group applied gloves posing perspective, perhaps this reinforces the import- more often as compared to the other groups combined. ance of accurate moulage portrayal for inexperienced The trend to conduct an abdominal palpation continued clinicians. Extending on this idea of perceiving authenti- in the combined LowAuth/control group (p =0.02). We city, the participants highlighted the impact of previous considered that in the case of the high-authenticity group simulation authenticity and design; that is, perhaps the completing neurovascular observations, this might have prior exposure to simulation has a stronger impact on been due to the additional visual stimuli of blood that trig- their perception of reality in simulation, than the design gered the need (signalling adequate conceptual realism) to of this simulation itself? check pupillary response and other neurovascular indica- We anticipated that the moulage groups would act tors. As hypothesised early in regard to the abdominal pal- quicker than the control in the time-to-action index. pation, we felt that the absence of other factors (such as This was not supported by the data—in fact, in some in- lacerations and grazing), the participant focused on the stances the time-to-action was slower in the HighAuth visual and audio cues of the abdominal injury. It was un- group. The HighAuth group took (on average) 245.10 s surprising to us that the HighAuth were more likely to to call for help, approximately 30 s longer than the con- apply gloves, students expressed in the interviews “oh I trol group and 50 s longer than LowAuth. Although the saw the blood and thought, I need to put gloves on”;this results were not significant, we hypothesise that High- could have interesting implications for the role of moulage Auth had more visual items to prioritise and consider as in teaching the use of gloves and personal protective a part of their assessment process. Interestingly, the con- equipment (PPE). trol group exposed the abdomen quicker than LowAuth In considering the comparisons of moulage versus no (around 40 s difference), and HighAuth was very similar moulage (control versus LowAuth/HighAuth) and High- to the control group timing (3 s longer). It is plausible Auth versus LowAuth/control, it was interesting that the that this also is due to cue processing and the focus on significant results existed in the latter comparison as op- audible cues may have prioritised their clinical decisions. posed to the first. We interpret this to mean that high- In the interviews, participants identified they focused on authenticity moulage has a more directive effect than certain verbal cues more than others as they believed the low- or no-moulage conditions. perhaps there was hidden meaning in them or the con- In rating the moulage authenticity, participants rated federate was trying to direct them a certain way. This the moulage, or lack of moulage, accordingly. This con- does not explain why the HighAuth group exposed the firmed the ratings from the other self-report. Students abdomen so quickly, perhaps the visual cues on the face consistently rated the control group moulage as low au- and arms may have triggered a need to investigate, dem- thenticity, LowAuth as medium authenticity and High- onstrating the effects of physical and semantic realism. Auth as high authenticity. There has been no previous Alternatively, maybe they found the authentic moulage exploration of moulage authenticity and participants’ distracting; however, this would appear unlikely given interaction with varied levels of moulage. the participants’ discussion in interviews where they expressed the positive views towards moulage being in- Implications for moulage use cluded and the sense of urgency when it was present Although moulage may not impact clinical decisions (demonstrating phenomenal realism). detrimentally, this might not be enough to consider that Stokes-Parish et al. Advances in Simulation (2020) 5:23 Page 11 of 12 moulage is insignificant. As we have seen in the inter- by statistical power calculations. The study was adver- views, participants identified that they spent significant pe- tised with many weeks in advance and delivered at alter- riods of time trying to determine the conditions of nate times that might be suitable for students study simulation. This is supported by work exploring the schedules, including extending the study for an add- process of suspending disbelief (SOD) in nursing students itional year. The simulation centre was based on the whereby authors state “enhanced environmental fidelity same site where students attended classes and place- promotes SOD” [25]. Expanding on this further, if the ments. This provides limitations for the interpretation of conditions of simulation are not consistent across all ex- results—the data results may have been too low to de- posures in a curriculum, it seems that this has impact on tect sizes of effect. Despite this, we did achieve statisti- their ability to suspend disbelief, spending more time on cally significant results that seemed to be accompanied focusing on deciphering the relevance. The underlying by an adequate effect size. We recognise the limitations message here is that consistency across simulations is key. of a single assessor to determine the clinical actions Beyond this, moulage might contribute as a visual cue completed and the time-to-treat information. A more more significantly than expected—as demonstrated by the robust approach might have been to have two assessors participants’ use of gloves and the completion of neuro- to then confirm the reliability of the judgement. This vascular observations and the students’ views. Simulation limitation is also extended to the coding of the inter- provides an opportunity to rehearse clinical practice and views—although Grounded Theory techniques do not develop the ability to manage complex situations. Students typically use multiple coders, we did not utilise the described not taking the simulation seriously or “faking” whole breadth of Grounded Theory. In this instance, it it; what is the implication for this in transferring learning? may have strengthened the work by having a second Although our primary focus on this study was the impact coder. Unfortunately, time and budgetary restraints lim- on engagement, there is a potential link here. If the lack of ited the feasibility of these approaches. authenticity of moulage prompts participants to take The type of scenario used could be a potential limita- shortcuts, then it is worth questioning if we are contribut- tion. Namely, a trauma situation may have more weight ing to negative learning? What we mean by this is the in- on the importance of engagement, as opposed to, for ex- advertent, incorrect messages that we send to participants. ample, a dermatology scenario. Conversely, the urgency In this scenario, no or poorly authentic moulage reduced of a trauma scenario may have enough impetus to en- the likelihood of applying PPE, sending the message that gage participants regardless of the level of authenticity, gloves are unimportant, thereby leading to “habitual un- whilst the authenticity of dermatology might be more safe behaviour” as described by Weller et al. [26]. The important than the authenticity of a trauma simulation. broader result might be an artificial type of learning, which we feel the students alluded to in their comments Conclusions on “doing it for the sake of doing it”. Another extension of Exploring engagement is an emerging topic in simulation, this negative learning might be the example of the slower with new techniques for measurement becoming available. abdominal palpation in the high-authenticity group—by These methods might provide better guides for measuring not exposing participants to real conditions distracting engagement. Other areas of work that should be explored factors, we might be inadvertently training them to only include investigating how the quality of previous simula- look for the obvious. Creating an authentic environment tions determine engagement with scenario, and how mou- is often limited by cost; however, we would argue that not lage influences on so-called negative learning and taking full advantage of simulation (significant expend- developing good clinical habits. Additionally, further work iture is already there) would be a missed opportunity for could be done to explore the relationship of authentic rehearsing clinical practice. moulage and working memory or cognitive load. This Although not generalisable for all situations, the mou- work would be interesting if replicated in a different clin- lage might be better off being authentic. Moulage added ical environment—for example, obstetrics, and other complexity to the scenario. Highly authentic moulage emergency scenarios. And, finally, it would be beneficial might provide more consistent performance behav- to explore the impact of authentic moulage on fully quali- iours—what are the implications of this for high-stakes fied clinicians or in other health professions groups. assessment versus technical skills? Perhaps low-authenti- This study adds to our understanding of the role mou- city moulage is be more confusing than high- lage can play in the participants engagement in simula- authenticity. tion. Within the context of undergraduate medical students, the use of authentic moulage may provide Limitations more consistent patterns of engagement, as compared to Despite repeated efforts to recruit participants, we had no or poor-quality moulage in simulation. Additionally, no success in recruiting the required number indicated moulage may provide a more realistic process of Stokes-Parish et al. Advances in Simulation (2020) 5:23 Page 12 of 12 prioritising care, thereby contributing to deep learning. Received: 28 October 2019 Accepted: 13 August 2020 We suggest that the authenticity of moulage contributes to learner engagement by highlighting the importance of References the activity, allowing them to fully rehearse an activity 1. Benyon, D., P. Turner, and S. Turner, Designing interactive systems: people, activities, contexts, technologies. 2005: Pearson Education. and minimise instances of determining what is real and 2. Hung, D., S.C. Tan, and T.S. Koh, Engaged learning: making learning an what a condition of the simulated environment is. authentic experience, in Engaged learning with emerging technologies. 2006, Springer. p. 29-48. 3. Dede C. Immersive interfaces for engagement and learning. Science. 2009;323. Supplementary information 4. Hagiwara MA, et al. Measuring participants’ immersion in healthcare Supplementary information accompanies this paper at https://doi.org/10. simulation: the development of an instrument. Adv Simul. 2016;1(1):17. 1186/s41077-020-00142-0. 5. Dieckmann P, Gaba D, Rall M. Deepening the theoretical foundations of patient simulation as social practice. Simul Healthc. 2007;2(3):183–93. 6. Rudolph JW, Simon R, Raemer DB. Which reality matters? Questions on the Additional file 1:. Supplementary data file path to high engagement in healthcare simulation. Simul Healthc. 2007;2. Additional file 2:. Stimulated recall interview instructions 7. Padgett J, et al. Engagement: what is it good for? Adv Health Sci Educ Additional file 3:. Participant survey on perceived engagement Theory Pract: The role of learner engagement in healthcare simulation contexts; 2018. 8. Stokes-Parish JB, Duvivier R, Jolly B. Investigating the impact of moulage on simulation engagement — a systematic review. Nurse Educ Today. 2018;64:49–55. Abbreviations 9. Stokes-Parish JB, Duvivier R, Jolly B. Does appearance matter? Current issues LowAuth: Low-authenticity moulage; HighAuth: High-authenticity moulage; and formulation of a research agenda for moulage in simulation. Simul ISRI: Immersion Scale Reporting Instrument; SR: Stimulated Recall; JSP: Jessica Healthc. 2016;22:22. Stokes-Parish; BJ: Brian Jolly; RD: Robbert Duvivier; ED: Emergency 10. Mills BW, et al. Investigating the extent realistic moulage impacts on immersion and Department; EEN: Endorsed Enrolled Nurse; MARS: Moulage Authenticity performance among undergraduate paramedicine students in a simulation-based Rating Scale; INACSL: International Nursing Association for Clinical Simulation trauma scenario a pilot study. Simul Healthc. 2018;13(5):331–40. and Learning; ANOVA: One-way analysis of variance; SPSS: Statistical Software 11. Shiner N, Howard ML. The use of simulation and moulage in undergraduate Package for Social Science; OSCE: Objective Structured Clinical Examination; diagnostic radiography education: a burns scenario. Radiography. 2019. PPE: Personal protective equipment; SOD: Suspension of disbelief 12. Sabzwari S, Afzal A, Nanji K. Mimicking rashes: use of moulage technique in undergraduate assessment at the aga khan university, Karachi. Educ Health. Acknowledgements 2017;30(1):60–3. NA 13. Garg A, et al. The skin cancer objective structured clinical examination (SCOSCE): a multi-institutional collaboration to develop and validate a clinical skills assessment for melanoma. J Am Acad Dermatol. 2015;73(6):959–65. Authors’ contributions 14. Petersen C, et al. Optimization of simulation and Moulage in military-related JSP designed the majority portion of the study, RD and BJ contributed to the medical training. J Spec Oper Med. 2017;17(3):74–80. critique and review of the design. JSP carried out the statistical analysis with the 15. Rystedt H, Sjoblom B. Realism, authenticity, and learning in healthcare supervision of BJ. JSP carried out the qualitative analysis with the supervision of simulations: rules of relevance and irrelevance as interactive achievements. RD. JSP drafted the manuscript. RD and BJ contributed to the review and Instruct Sci. 2012;40(5):785–98. additional content. All authors read and approved the final manuscript. 16. Mills BW, et al. Effects of low- versus high-fidelity simulations on the cognitive burden and performance of entry-level paramedicine students: a mixed-methods Funding comparison trial using eye-tracking, continuous heart rate, difficulty rating scales, This paper is as a result of the work exploring how the authenticity of video observation and interviews. Simul Healthc. 2016;11(1):10–8. moulage influences simulation. The work received an SSH Novice Researcher 17. Hattie, J., Visible learning: A synthesis of over 800 meta-analyses relating to Grant. In addition to the SSH Funding, this work received Seed Funding from achievement. 2008: Routledge. The University of Newcastle, Australia. 18. INACSL. Standards of best practice: simulation<sup>SM</sup> facilitation. Clin Simul Nurs. 2016;12:S16–20. 19. Sarahn Lovett, J.R., Sharyn Hunter, Ian Symonds, Naomi Tomlinson, Robert Availability of data and materials Gagnon, Bernard Charlin, Joerg Mattes, Respective value of the traditional All data generated or analysed during this study are included in this clinical rotation and high fidelity simulation on the acquisition of clinical published article and its supplementary information files. If the data is not reasoning skills in medical students – a randomized Controlled Trial. 2016, available in the supplement, they are available from the corresponding AMEE: MedEdPublish. p. 9. author on reasonable request. 20. Stokes-Parish J, Duvivier R, Jolly B. Expert opinions on the authenticity of moulage in simulation: a Delphi study. Adv Simul. 2019;4(1):16. Ethics approval and consent to participate 21. Pywell MJ, et al. High fidelity, low cost moulage as a valid simulation tool to The study protocol was approved by the University of Newcastle Human improve burns education. Burns. 2016;42(4):844–52. Research Ethics Committee (H-2017-0214). 22. Lyle J. Stimulated recall: a report on its use in naturalistic research. Br Educ Res J. 2003;29(6):861–78. 23. Holmqvist, K., et al., Eye tracking: A comprehensive guide to methods and Consent for publication measures. 2011: OUP Oxford. All participants’ consent forms included consent for publication. 24. Sobral DT. What kind of motivation drives medical students' learning quests? Med Educ. 2004;38(9):950–7. Competing interests 25. Muckler VC. Exploring suspension of disbelief during simulation-based The authors declare that they have no competing interests. learning. Clin Simul Nurs. 2017;13(1):3–9. 26. Weller JM, et al. Simulation in clinical teaching and learning. Med J Aust. Author details 2012;196(9):594. School of Medicine & Public Health, University of Newcastle, Callaghan, New South Wales, Australia. Center for Educational Development and Research in Health Sciences, University Medical Center Groningen, Groningen, The Publisher’sNote Netherlands. School of Rural Medicine, University of New England, Armidale, Springer Nature remains neutral with regard to jurisdictional claims in New South Wales, Australia. published maps and institutional affiliations.

Journal

Advances in SimulationSpringer Journals

Published: Aug 26, 2020

There are no references for this article.