Teamwork is known to impact on patient care and safety [1] and effective teamwork can improve patient outcomes [2, 3]. Evidence supports the need for ongoing teamwork training to reduce avoidable errors and health care costs [3, 4] and simulation is well-recognised as a teaching method to improve teamwork [5, 6] and patient safety [7, 8].

It has been suggested that organizations could benefit from building a critical mass of staff who have acquired ‘transportable’ teamwork competencies relevant across many settings and teams [9]. Health professionals need to function in dynamic, ad-hoc teams with multiple team memberships [10]. Tannenbaum et al. [9] coined the term ‘flash team’, referring to teams that form without members having worked together. These ‘flash teams’ are common in healthcare [11]. Using immersive simulation to teach teamwork in large undergraduate programs presents the dilemma of large numbers of students with limited faculty resources.

Teamwork training has now been established in many undergraduate curricula. However, several issues arise. Firstly, most teamwork measurement tools have been designed for practising clinicians in demanding clinical contexts, rather than for students. Secondly, most published studies have relied on external faculty observing and assessing the team behaviours of students to facilitate feedback in debriefing sessions following the simulation [12–14].

The latter is resource intensive and the high cost of small group simulation and expert debriefing may limit the exposure that most students have to experiential teamwork training. There is some evidence that other approaches that facilitate reflection may be useful. Interprofessional within-team debriefing, using video self-review, has been shown to be as effective as traditional facilitator-led debriefing [15]. At the least, this suggests that simulation participants are able to debrief when facilitators are not experienced or available. Self-debriefing is dependent on participants ability to engage in self-assessment, which is critical to the learning process [16]. A teamwork rating tool with explicit descriptors of the behaviours of effective teamwork could assist this self-assessment.

There are numerous tools that assess teamwork performance, however, these have been developed for practising health professionals and for specific contexts. The Mayo High Performance Teamwork scale which assesses the effectiveness of team training has been used extensively in practicing health professional populations [17]. TeamSTEPPS is a team training intervention which developed a self-report that was shown to have construct validity in multidisciplinary healthcare professionals [18]. A valid teamwork measure for intensive care teams after simulated critical events was developed [19] and then validated later as a self-assessment tool [20]. However, teamwork tools for self-assessment that are designed specifically for health professional students are lacking. Furthermore, teamwork tools designed for complex and highly specialised clinical domains such as critical care or the operating room may not be appropriate for health professional students, with their limited clinical knowledge and lack of experience working in healthcare teams. As such, the authors developed a self-assessment teamwork tool, based on the Weller et al. [19, 20] teamwork tool, appropriate for health professional students and in a specific clinical context more relevant to the knowledge and skills of students.

The aim of this study was to develop a self-assessment teamwork tool for medical and nursing students and to explore its reliability and validity.

To engage large numbers of students in a stressful clinical environment with time pressures of managing trauma patients, a mass casualty simulation was utilised.

202 students were enrolled into the study: 128 medical students and 74 nursing students, with 145 students attending the first simulation scenario, and 57 students completing the second scenario. There were more female students (66.2 %) and a minority of students had participated in a previous interprofessional learning experience with other health students (16.8 %).

The authors used several well-agreed criteria to fit the exploratory factor analysis [27]. Two items (item 16: when faced with a problem, external assistance was sought and item 17: situational updates were given when the situation changed) were removed from the analysis as they had inter-item correlations less than 0.30. Kaiser-Meyer-Olkin measure of sampling adequacy was deemed acceptable at 0.912, greater that 0.5 scores. Bartlett’s test of sphericity was statistical significant (χ² (136) = 1425.67, p < 0.001). These measures were at an acceptable level to continue with the EFA.

Over 60 % of the cumulative variance was explained within a 2-factor solution. The first factor had an eigenvalue of 4.32, accounting for 48.0 % of the variance, with the second factor having an eigenvalue of 1.10, with a further 12.3 % of the variance.

Confirmatory factor analysis was performed using the 9 items identified from the EFA (Fig. 1). The overall Goodness of Fit model was deemed acceptable (χ² min/df = 1.768 (p = 0.009); GFI = 0.970; RMSEA = 0.062; CFI = .970; TLI = .948) [26]. Despite the chi-square goodness of fit being above 1.0, the GFI, RMSEA, CFI and TLI all indicated a satisfactory 2-factor solution fit.

The SATTS was shown to have good reliability and construct validity following analysis using EFA and CFA in the context of a large-scale mass casualty simulation which provided challenging teamwork situations that required students to work collaboratively [21].

The EFA revealed 2 factors in the SATTS which related to the constructs, team coordination and communication and information sharing and support. These are different from Weller et al. [19] which had 3 factors relating to leadership and team coordination, shared situational information, and mutual performance monitoring. The difference in factors is understandable following deliberate removal of items from the leadership and mutual performance monitoring construct. However, the CFA model validated these findings, as the goodness of fit model for a 2-factor solution was suitable. These results demonstrate that the SATTS is a valid tool that measures teamwork, specifically for student cohorts.

There were six items that were cross-loaded on the two factors and not included in the CFA model (see italicized items from Table 1). Four of the items (no. 3, 7, 10, and 11) were from the Weller et al. [19] tool, and the remaining 2 were new items developed by the simulation experts. The EFA and CFA findings provide evidence that removal of the Weller et al. tool items was appropriate for the student cohort. The mass casualty simulation was a complex, challenging teamwork situation and some of the items relevant to critical care teams were not suitable for undergraduate health professional students. Student teams were not experienced like existing health professional teams who normally work together. Further, the mix of clinical skills, expertise and leadership behaviours appropriate for real teams did not fit the study scenario teamwork context. Nevertheless, two of the new items related to teamwork communication (no. 8 and 9) did not fit the 2-factor solution. These 6 items had similar factor loadings for the 2 factors, indicating that they were not distinguishable between the Teamwork coordination and communication and Information sharing and support constructs and were correctly removed.

Teamwork tools mainly use behavioural rating scales or self-assessment formats to evaluate teamwork performance [7, 29]. While behavioural rating scales require a trained observer(s) to measure teamwork and score team performance, self-assessment tools enable users to gain insights into their own thinking and the reasons behind observed behaviours [7]. The SATTS demonstrated evidence of reliability and validity when used by students, and could potentially promote formative self-assessment of teamwork prior to facilitator-led debriefing. This tool may also enhance the educational value of future teamwork-focused simulation for students, especially when large student numbers limit the availability of small group debriefing by expert facilitators. Use of the SATTS prior to debriefing may facilitate students’ critical self-assessment of teamwork. The SATTS could function as a structured guide to reflect on the teamwork behaviours experienced by the students following simulation scenarios and potentially promote the learning process. This could potentially reduce the need for an expert to facilitate the debrief [30]. However, further research is required to establish this paradigm.

Whilst the SATTS was developed for self-assessment of teamwork in student cohorts, the context of the tool needs to be considered. The original Weller tool was designed to measure teamwork behaviours in critical care health professionals (doctors and nurses) who work in these settings. The authors chose to adapt this for undergraduate health professional students as the tool contained items indicating requisite behaviours for optimal teamwork. The use of the SATTS following the mass casualty simulation was likely to have influenced the student responses as the clinical scenario was one that had not been encountered by students previously. Therefore, it was appropriate to adapt the tool for both student teamwork and the simulation environment. Further, the validation of the SATTS using both EFA and CFA provided construct validity for teamwork communication and information sharing which are similar to other tools [17, 28].

This tool could potentially be used in other contexts, where students are exposed to different forms of teamwork training. The SATTS could be further developed for use by expert observer raters who assess teamwork performance, similar to the original critical care tool development. The tool could then be used to assess student teamwork in assessment formats, such as Team Objective Structured Clinical Examinations.

To date, most teamwork tools focus on practising health professionals with few designed specifically for undergraduate students. The authors modified an existing self-assessment teamwork tool (developed for critical care teams) for use with undergraduate health professional medical and nursing students. The results of the present study provide evidence to support the validity and reliability of the SATTS for self-assessment of teamwork in the context of simulated mass casualties. The SATTS could be used by students to promote learning of important teamwork behaviours and potentially well suited for large scale interprofessional student teamwork activities.