Quantitative.

We collected baseline demographic data, including age and sex. In addition, we administered the following questionnaires:

1. The state questions on the State-Trait Anxiety Inventory (STAI): [20], made up of multiple questions on a 4-point Likert scale, responses are summed to get a discrete measure of anxiety. This is a common well-validated anxiety inventory used in equivalent types of research.
2. Simulation Effectiveness Tool [19]. This validated measure addressed our aim to examine participant’s perceptions of the simulation course, centering on key clinically relevant learning objectives, including effectiveness of simulation; how well learning objectives, e.g., in assessment and treatment were met; applicability to clinical practice; and confidence. These were measured via a questionnaire consisting of 3-point Likert scale questions.

Quantitative data was analyzed using SPSS 29 (IBM SPSS Statistics, New York). Numeric variables were presented using mean and standard deviation as data was normally distributed. The linear associations between numeric variables were presented using Pearson’s correlation coefficient. Linear regression models were used to quantify associations between STAI anxiety and simulation effectiveness score after controlling for confounders.

Sample size and power calculations.

Previous research from Gosselin [16] into the relationship between state anxiety and simulation effectiveness found a moderate correlation of r = 0.42 corresponding to a coefficient of determination (r-squared) of approximately 0.18. Setting the univariable coefficient of determination to be 0.15 (assuming a weaker relationship) in our study, we estimated that after accounting for anxiety and other sociodemographic factors, the coefficient of determination would be increased to 0.2. To detect a coefficient of determination of 20% with a power of 95% and type I error probability of 0.05, a sample size of 55 would be required. We expected a dropout rate of 20%, so we needed to recruit 66 participants to account for these possible missing observations. Preliminary data from participants showed a smaller effect size than anticipated and a greater ease in recruitment than anticipated, leading to a larger sample size target of approximately 80 (allowing for approximately 15–20% attrition).

Qualitative.

A subset of candidates also completed qualitative interviews. A semi-structured interview including broad open preliminary questions (‘How did you find the simulation?’) was designed with the aim of eliciting students’ experiences in a non-directive manner. In line with grounded theory, early interviews informed prompts for subsequent interviews [21]. Interviews took place either face to face in a confidential space at the hospital, or online via secure video call. Thematic analysis of interview data was informed by the principles outlined by Braun & Clark [22]:

1. Data immersion
2. Identifying initial codes
3. Identifying preliminary themes
4. Identifying relationships between themes
5. Identifying overarching themes
6. Revisiting previous stages to ensure themes are grounded in the data, with representative quotes being used in the final report.

No sample size was predetermined for qualitative data collection; instead, based on the information power doctrine [23] interviews continued until sufficient detail emerged to answer the research questions. Power was increased, and therefore required sample size decreased, by having clear and circumscribed research aims, studying a specific participant population (5^th year medical students), basing qualitative analysis on established grounded theory and thematic analysis methodologies, having trained interviewers who were able to establish rapport with candidates and increase the quality of the dialogue (GD is a clinical psychologist by background), and using an appropriate analysis strategy as detailed above. This process meant that most of the interviewed candidates were students who were recruited earlier in the study timeline; however, candidates in later phases of recruitment were also interviewed to ensure representativeness of themes across the set.

NS transcribed and identified initial codes. GD separately identified codes. Comparison and discussion was undertaken until a consensus was reached. Only subsequently were higher level themes examined. GD explored a subset of qualitative data for quality assurance and again highlight differences. Final themes were compared with original data as a further check of quality. Data was stored digitally on secure institutional servers, with access limited to the research team. The consolidated criteria for reporting qualitative research (COREQ) [24] checklist can be found in supplementary material, which includes quality assurance comments including on research reflexivity.

Anticipation of the unknown.

At the beginning of their course, many participants reported a build-up of anxiety:

I think the nerves were building up before and then they kind of peaked just before you step into the room. (Participant (p) 25)

Some responded to this by trying to make the unknown more certain and felt reassured when more certain:

My first thought was I need to know where all the stuff is, like what drawers contain what, and I also want to know like I want to guess what the station is. (p. 1)

I think going second was nice because we knew more or less what to expect, and, anyway, we had time to like think about what to do. (p. 17)

As well as it being emotionally difficult to tolerate, uncertainty was seen by many participants as potentially increasing cognitive load.

It’s like a new room, new setup. You just don’t really know where things are. (p. 5)

The increase in anxiety also impaired the memory recall.

One of the [candidates] came out after and basically said he knew the guidelines inside out, but he was kind of like stunned into silence – so the other person basically took control. (p. 3)

The uncertainty was not always viewed as aversive as it offered an opportunity to learn:

I think it’s good actually to go in with a kind of blank knowledge base… I think the areas I gained from most were things that I knew very little about. (p. 2)

Many candidates linked this ‘blank knowledge base’ to what might be faced in real clinical practice, leading to our second overarching theme.

Simulation as real.

The realism of the simulation scenarios contributed to the anxiety for many participants:

I felt as though I was a little bit out of my depth and nervous as to what the next steps would be, especially because it did feel quite real. The noises that were being made and the beeping and all of that and monitors on the screen, so it did feel a little bit overwhelming at times. (p. 13)

It also set the simulation apart from other educational experiences, and elicited more spontaneous relational and emotive responses:

Even though we do a lot of stuff in practice like in [a medical exam] situation, it is very different to being in a more realistic setting. (p. 3)

I feel like part of me was starting to get a little bit snappy; mostly because the situation had dramatically changed, one of the members of the team was on the phone, so he couldn’t communicate with all of us. (p. 1)

These more emotive responses also tied to a realisation among the students of what might be required of them professionally:

You’re like, “OK, yeah I do need to think about this” – and I think like when you’re studying in books and when you’re doing questions a lot of it is like “Oh that will be decisions that I’m going to have to make in the future,” but you don’t really take in and think about how that’s actually going to feel. I especially thought it was good that the nurse questioned me because that’s what’s going to happen. Having to think about what I was saying and back up my reasoning in like a time-pressured situation was really valuable. (p. 8)

Several candidates reported that unlike previous classroom learning, simulation brought with it an insight into the potential responsibilities they would hold as a doctor. These included the challenges of sensitively managing leadership roles, something they had not previously covered in their training, and the thought of how others would perceive their attempts to step into the leadership role.

So, then you’re pretending to be the decision-maker for the first time and then this person is going, “How would you like me to help you?” which is actually a very loaded question…you definitely don’t want to be like the most assertive person, you don’t want to be like: “No I want you to do this.” (p. 1)

I would hope that I’m not like the doctors that just talks and dictates what people should be doing, but I think I had also like been really conscious of not wanting to look silly in front of the nurse. (p. 8)

The demands of communication and teamworking were commented on by many candidates as contributors to anxiety, but also as revealing professionally useful insights into their relational tendencies:

I think, personally, I didn’t know when I should step forward or step back. Normally I just want to stand on the fence and let other people do it, but then, when it’s silence, sometimes you feel like you should step forward but you don’t know if anyone else will. (p. 24)

The contrast between intellectual understanding and its application under pressure in a team setting also informed the assignment of roles, and students described the novel experience of having to adjust to working within a group:

When it’s a group, let’s say three of you, one has to take a back seat and you follow someone else, so then you have to kind of match their way of thinking, which I found a little bit difficult because I remember things at my own pace. (p. 5)

The students spoke of new insights into situational awareness: the nature of decision-making in real time, and the tension between this and the demands and realities of the situation.

At one point I think we asked [the nurse] to do like two or three things at the same time and she’s just like, “Yeah, I’m doing this but there’s only one of me!” (p. 3)

In the classroom or in exams, students have to verbalise or transcribe their decisions. Here, they had to navigate real-time multi-tasking within a clinical space where the time and resource required to prepare medicines, for example, had to be factored in:

[The nurse said] “You’ve just said like four things that I can’t do at the same time” – so that was like one where you can think out loud but you have to differentiate when you’re asking people to do things, as opposed to thinking it. (p. 11)

Several participants reported that they had not appreciated this reality previously, describing simulation as a stepping-stone to the real clinical world. Similarly, asking for help was mentioned by several students as a useful but stressful new skill that simulation exposed them to:

After the first SIM, I was like, “You know what, this time we’re going to go in, we’re going to get the guideline book, we’ll call for help early.” (p. 28)

In some cases, it wasn’t just that the simulated environment helped students realise the gulf between classroom recall and clinical acumen, some of the skills required were in direct contrast to that required in an exam or classroom. In the classroom or exam setting, for example, a student may be rewarded for their declarative knowledge, which is often imperfect and prone to attrition. In a real clinical setting it is often more effective to know where and how to find help, and which sources of information are reliable and which are not – abilities which drive the prime imperative of providing the right care at the right time. The habit of feeling that one needs to demonstrate knowledge above all else can be an impediment to optimal care.

I feel like maybe passing you examines is not necessarily that useful for starting [as a doctor], like learning for [medical exams] I think is of limited use compared to examining or learning in a simulated environment. (p. 2)

Being observed.

A related, though distinct, source of anxiety came from being observed by peers and facilitators. Participants reported a potential sense of exposure or scrutiny from the unavoidable fact of being observed.

I did feel quite nervous, and so, then I was like, “OK, actually my skills are going to be exposed.” So maybe there is a bit of a personal imposter syndrome…I guess, and I should want to expose my skills and show the things I’ve learnt...to show I have learned something in four years in medical school. I enjoyed the exposure, I guess. (p. 15)

Participants recognised that they too could create anxiety for their peers:

[My peers] might feel criticised or judged and then it might make them feel worse about trying in the sim because, you don’t want to feel like someone’s watching you and telling you what you did wrong. (p. 4)

Some linked this scrutiny to an awareness that their practice would be observed in real clinical work, thus circling back to the idea of professional responsibility and their identity as doctors. Several candidates recognised however, that observation leads to constructive criticism and feedback that could improve their skills:

I found [the feedback] useful and it highlighted some things that I need to go over. (p. 24)

Further, several candidates felt observation could be supportive, and that this could be aided by potentially assigning roles to observers. Perhaps the wish for observers to have roles was related to a more unnerving sense of scrutiny, elicited when the nature of the observation remained ambiguous. One candidate recognised this ambiguity as rife for projection of his own standards onto others:

Seeing my colleagues do it as well, having known that I had also done it, was like: you know, it’s okay, we’re all learning here – but maybe if it was the other way around…had I seen the other group do it first and been like, “Oh, you should have done this and that,” and then when I couldn’t do it in the second round maybe I would have really felt: why am I not getting this straight or why am I not getting this right? (p. 8)

Others appreciated that a difference may exist between self-reported performance and peer-reported performance, suggesting an awareness of an overly critical perspective:

It was interesting to see the contrast between how it felt I’ve done and how [the group] saw us working together and being systematic. So it was interesting to see that actually those feelings might not be rooted in reality or rooted in logic. (p. 13)

Thus the act of benign observation corrected an overly critical self-assessment. Further, being both observer and participant facilitated greater empathy for something that might appear easier from an observer role:

We watched other people while they were doing it, and, when they were doing it, it actually seemed a lot easier from the outside; but, once you’re there you see how stressful it is. (p. 17)

The most pedagogically useful sense of observation was the capacity to step back and reflect kindly but critically on one’s own practice. This key aim of simulation was described by several candidates:

It was very interesting to see how my system kind of goes out the window initially when I’m faced with something [real], rather than faced with a textbook example… I think that’s something to reflect on, like how I could be better at managing pressure and just going through my normal process. (p. 19)

Participants reported valuing the protected space for reflection in the debrief:

When [in the debrief] we were reflecting, hearing opinions, it was really interesting because I’ve never really spoken to these people before. (p. 1)

Participants commented that, though the expertise of the staff could make them feel nervous, the stance of the educators helped with the process of making their anxiety manageable and facilitating a more benign sense of observation:

I think once I got out of the simulation I felt: “Oh, I messed up!”. During the debrief, I think just going through it you realise that you actually did a lot! (p. 17)

Thus, educators appeared to be able, via their stance and instruction, to navigate some participants towards a more benign and pedagogically useful framing and experience of observation and anxiety during simulation.