Expert opinions on the authenticity of moulage in simulation: a Delphi study

What does the literature say about moulage?

Until recently there has been no attempt to explore the place of moulage in simulation other than accepting the status quo; evidence for moulage mainly focuses on instructional “how-to” guides and historical accounts in dermatology [14]. A recent systematic review explored the effectiveness of moulage on engagement in simulation, highlighting the paucity of credible research on moulage [12]. The research highlighted the disparities in use across disciplines, industries and simulation centres, identifying there was no evidence to support the need for moulage in simulation [12, 15]. Since the publication of this systematic review and after a moulage “call to arms” [14], Mills et al. (2018) demonstrated that the use of “highly realistic” moulage versus no moulage improved immersion in paramedicine students [16]. Novel research in the field of radiography identified students’ preferences for realistic moulage, with students commenting “… couldn’t tell it was makeup at first. Which I thnk [sic] is better as it looks more realistic …” [17]. Other research suggests that moulage improved realism; however, the quality of moulage was purportedly challenged by the lack of time to apply moulage [18].

The term highly realistic, or realistic, is used frequently to create a distinction in simulation realism and in simulation manikin marketing [16, 19–21]. Despite this, there is no clear definition of what constitutes realistic. In fact, many educational outcomes papers that discuss the comparison of highly realistic or high-fidelity versus their low counterparts do not provide measures to define these classifications. For example, in Mills et al.’s (2018) study described earlier, there was no measurement to quantify “highly realistic”; however, in a study on burns training, Pywell et al. (2016) assessed the face validity of moulage in a comparison study for burns training [22]. In dermatology, authenticity of moulage rates highly due to the high-stakes conditions of diagnosing melanoma and other skin conditions [23–28]. In the field of gynaecology, researchers developed a rating tool to determine realistic simulator design [19, 22]. This is particularly interesting, given the inconsistency in industry approaches to designing simulators appearance—for example, Laerdal SimMan 3G versus LifeCast Adult Male Manikin versus CAE Healthcare Apollo (See Fig. 1). These manikins have varied levels of authenticity in their appearance—LifeCast, for example, employs SPFX teams to replicate the life-like appearance of manikins, including realistic skin and other details.

Some simulation designers and users argue that a high level of accuracy in the portrayal of moulage is necessary for the effectiveness of simulation; however, absolute authenticity is questioned in the current literature [9, 29, 30]. How does the use of moulage change the approach to instructional design of simulation? Before exploring its impact on instructional design, and then simulation participants, we posit that we must first understand what constitutes moulage authenticity. What is good moulage? And what, therefore, is bad moulage? How much time and money can one justify spending on moulage, when the cost-effectiveness of simulation is a priority for institutions and education [31]. We identified authenticity as a priority area for research in the use of moulage in health professions simulation [14], prior to exploring the many other possibilities in engagement, instructional design, and other areas of interest.

Panellist selection

We sought to recruit a representative sample of experts with first-hand knowledge of moulage. An individual was classified as an expert if they worked with direct involvement in simulation instructional design and implementation (in particular, moulage), or as an experienced SPFX artist, involved in creating moulage or arts in anatomy. This wide-ranging sampling approach is supported by literature on Delphi technique to achieve high-quality outcomes [33, 34]—we chose to keep this broad due to the lack of evidence for moulage in the literature. Individuals were excluded from the study if they had no expertise in moulage/SPFX design or application, or, they had no expertise in the instructional design of simulation. The researchers screened participant responses for exclusion.

Element generation

We used the typical Delphi method whereby the authors presented a series of open-ended questions to the experts to generate the initial elements [33]. The authors developed the round 1 questions using Dieckmann’s Theory of Realism as a conceptual framework [7]. In response to the need to understand moulage, we identified Dieckmann’s theory as an appropriate fit for developing a theoretical framework for moulage. This is because realism is contextual to both the environment and learning objectives, yet is subject to the participants’ judgement. Realism, in this theory, is composed of three domains—physical, semantic, and phenomenal. Physical realism refers to the actual physical representation, such as the characteristics of moulage, its textures, and colours; that is, how persuasive is the authenticity in your perception of reality [7, 14]. The semantical mode of realism refers to a conceptual type of realism, it is more about a participant’s relationship with the activity or story. That is, can s/he relate to the story? Is the representation plausible, could it occur in real life? In this instance, the moulage may not be authentic, but by way of the ‘fictional contract’ it is enough to help you regard it as authentic and predict what might happen next. For example, “if A occurred, B will happen”; Dieckmann et al. (2009) use the example of haemhorrhage—bleeding occurred; therefore, the blood pressure will decrease. How the information is shared is irrelevant, if interpretable information is shared. Finally, the phenomenal mode of realism is an emotional realism; engagement is reliant on your involvement in the situation and how persuasive it is to you, overall. Participants engage with the activity as if it were the real experience because they are emotionally engaged. For example, the moulage authenticity may be variable, but they engage with the narrative and situation because it makes sense to them.

This framework of realism creates the basis for starting our discussion on moulage in simulation, with attention to how moulage and engagement might interrelate. The use of Dieckmann’s Theory of Realism presents a novel approach to exploring moulage, operationalising an otherwise theoretical approach to realism in a practical way. Throughout this research, we framed our questions, interpretations, and analysis on this theory.

Data collection tool

Data was collected using Survey Monkey, a secure, encrypted web-based electronic questionnaire system [36].

Delphi procedure

The survey was presented to the experts in multiple iterative rounds via the survey host (Survey Monkey) in late 2016. In the first round, designed to identify the basis for generating the elements of authentic moulage. Instructions for completing the Delphi and the surveys were shared with the panellists, followed by demographic questions regarding the participants, as well as a series of open-ended questions, and a request to list at least three elements that contribute to the appearance of moulage being real (Additional file 4). The questions were grouped into the three categories of Dieckmann’s realism—physical, semantic, and phenomenal. Statements regarding physical, semantic, and phenomenal realism were continually presented to the panel in subsequent rounds. In round 2, the elements identified in round 1 were thematically analysed by the researchers together (JSP, BJ) to group common themes. The themes were presented in round 2 in question format, seeking agreement via 5-point Likert scale (strongly agree to strongly disagree) as to whether the elements contribute to the appearance of moulage being real (e.g. “Colour is an element that contributes to the appearance of moulage being real”). Participants were requested to rank their top 5 elements out of the 14 physical elements. In round 3, providing feedback from round 2, the questions were re-framed to seek a rating of importance. Once the consensus threshold was reached on an element, it was removed from further rounds of questioning. The same process was repeated in round 4 and round 5. Total consensus on every element was not reached; therefore, the Delphi closed after the fifth round as per the priori consensus. In all rounds, experts were given the opportunity to add additional background, suggest revisions or additions via the use of free text boxes in the survey instrument. A flowchart of the process followed can be seen in Fig. 2.

Priori consensus

In line with recommendations for Delphi technique methodology, a priori consensus was set at 80% or five rounds of surveys, whichever was sooner. The data was analysed using Microsoft Excel (Version 15.33). The results were shared in each round with the respective question.

Initial elements generated

In addition to the listing of elements contributing to physical authenticity, panellists were asked to comment on what level of authenticity is required for the moulage to be conceptually believable (semantic realism) and how moulage contributes to choices (phenomenal realism). Seventeen elements presented were in relation to semantic realism, and 16 elements were in relation to phenomenal realism. After combining common elements, there were 15 and 9 resulting elements, respectively.

Subsequent rounds

Throughout the rounds of questioning, the experts were invited to provide further comments to explain their answers or add to the discussion. With little theoretical underpinnings for moulage in simulation, the discussion amongst the experts contributed new ideas and concepts to the approach to the use of moulage. Early on the panellists identified the priority for realistic moulage. Examples of responses include

“the more realistic a situation or environment is, the more believable it will be for the learner and the easier it will be to engage”. (Participant 15: Intermediate Moulage Expertise, USA).

“There definitely needs to be a relatively high level of authenticity so that the learners ‘buy in’ to the scenario. If the learners deem the scenario unrealistic, they will not engage fully, and get nearly as much out of it, as they would if it was more realistic.” (Participant 2: Intermediate Moulage Expertise, Australia).

“… Obviously the higher the level of realism the more engagement of the learner ….” (Participant 3: Advanced Moulage Expertise, USA).

In describing the necessity for authenticity, experts identified that realism is learner dependent:

“… For example a red cloth could symbolise blood in a scenario if the learner is willing to believe that the red cloth is to be interpreted as blood. …” (Participant 18: Advanced Moulage Expertise, USA).

“Some novice students can engage in low fidelity simulations as they are not yet equipped to process information under pressure. For the more experienced, scenarios resembling reality as closely as possible are advantageous.” (Participant 16: Intermediate Moulage Expertise, Australia).

The thoughts from the experts regarding the necessity of authenticity presented some interesting tensions. Firstly, they felt that the accuracy of moulage directly relates to the relevance of the scenario, contributing to diagnoses and behaviours. Secondly, they felt that moulage that was not authentically portrayed can be confusing.

“the visual effect must be similar to the lived experience as this allows the student to connect the dots and transfer concepts and theory to practice”, (Participant 4: Intermediate Moulage Expertise, Canada).

“Bad or excessive moulage can be worse than none as it can remove the focus away from learning to ‘gazing’ at the theatrical aspects as an independent entity. Can be used by some learners as an excuse for poor performance - ‘it was naff and not anything like real life so I couldn’t apply myself like real life’”, and (Participant 7: Intermediate Moulage Expertise, Australia).

“The more realistic looking and feeling the moulage allows trainees to become more invested” (Participant 13: Advanced Moulage Expertise, Australia). However, the issue of cost and time became a central point of discussion—some experts described the use of cue cards (with written cues) instead of moulage. The final instrument can be viewed as Additional file 2.

Limitations

The research did not include experts from South America, Africa, or Asia. We are unsure if this was due to the methods employed or the patterns of research in those continents. Purposive sampling was selected as the method due to the increased likelihood that panellists selected possessed the necessary expertise. Perhaps this reflects the level of maturity of simulation in these continents, whereby research centres, journals, and attendance is not as predominant as in western societies. Additionally, it is worth considering the bias that might be inadvertently demonstrated by publishing bodies—it is known that the majority of authors in high-ranking medical education journals, for example, are from 5% of the world [40, 41].

Despite planned administrative processes and clear objectives, the administration of this Delphi was time-consuming [42]. This led to a delay in some of the survey deployments, which may have contributed to the attrition rate. Despite this, the attrition rate is not too dissimilar to other Delphi work—the larger the panel, the greater the attrition [43]. Some of the attrition may be attributed to a lack of interest or expertise (for example, the SPFX artist dropped out after round 1).

Additionally, the lack of information regarding moulage may have contributed to the need for the full five rounds of questioning and the lack of consensus. Perhaps the paucity of research around moulage and the absence of a theoretical framework contribute to the contradictory views and the lack of consensus on some issues. We would argue that, similar to Mullen (2003), a multiple “correct” answers are better than a “unanimously agreed wrong answer” [43].

As for other limitations, we do not present findings of a content validity study in this paper. This work is not included as we felt it would detract from the discussions raised by the experts and falls outside the scope of this study.

Future directions and conclusion

The potential for authenticity and moulage remains largely unexplored. Experts on moulage present thoughtful ideas as to how moulage might contribute to realism in simulation and what authentic moulage might mean in this context. The development of this instrument presents an opportunity to measure the impact of authentic moulage on various aspects in simulation. To verify the instrument’s representativeness, content validity should be assessed by means of an expert survey. Additionally, there is a need to assess the reliability could be verified in a series of trials, whereby moulage elements with a range of authentic and inauthentic features would be rated for authenticity by simulation experts, clinicians, and students. Additionally, the fit of this instrument across other domains, such as augmented, virtual, or mixed realities, could benefit the wider simulation community.

Further areas of exploration might include the comparison of junior versus senior learners, exploring how the approach to simulation context and cues differs. This could be done applying the moulage instrument to create clear distinction between high and low authentic moulage.

This paper presents novel work in the field of both authenticity and moulage. Tensions remain present in regard to the necessity of moulage, how it is applied and the level of authenticity required to be portrayed in a variety of settings. There is a clear imperative to explore authentic moulage further to benefit the simulation community.