To measure the effect of TraumaFlow as unaffected as possible by participant expertise, 11 participants with limited experience in treating polytrauma patients were included and as compared to the performance of three senior residents. Inexperienced participants were 11 medical students in their last year. Of these 11 students, eight had experienced ≤ 5 polytrauma managements, while 3 had experienced 6 to 10 cases before the study. The basic principles of ATLS® are part of the curriculum in medical school. Personal data collected on the participants included age, gender, and information on medical training level and experience in the treatment of polytrauma patients. All participants used the same prototype version of TraumaFlow, a computer-based decision support system developed by the authors that is run on a tablet computer.

The protocol of this study was approved by the local ethics committee (reference 118/22-ek).

In the trauma room of a level 1 trauma centre, two scenarios of polytrauma patients were simulated successively on a simulation manikin (Resusci Anne-Simulator, Laerdal Medical, Stavanger, Norway. Figure 2 together with virtual patient monitoring (Sim.Care EASY 2, deltaplex UG, Hamburg, Germany with TruMonitor-App, TruCorp Ltd., Northern Ireland).

Performance with and without computer-assisted support by the level of expertise of the participants (scale from 0 to maximum 12 points)

Case A was a 32-year-old male victim of a violent crime who was attacked with blunt weapons. He presents with tension pneumothorax, a relevant bleeding scalp wound, and a splenic injury with a resultant class III haemorrhagic shock. Case B was a 27-year-old man who fell from a scaffold from a height of 6 m. The patient had a severe traumatic brain injury with intracranial haemorrhage, unstable pelvic injury, and free fluid in the abdomen in the Koller pouch and retroperitoneally with resulting class III haemorrhagic shock.

Participants were asked to assume the role of trauma leader and were provided with a nurse and a resident.

The scenario was steered by a certified trauma simulation instructor, and changes in the patient’s condition were reflected via a simulation monitor or verbal instructions.

The first scenario was performed by the participants without support, the second scenario was performed with the assistance of the “TraumaFlow” workflow management system via a tablet. The sequence of cases changed with each participant (A-B, B-A, A-B, …).

During each scenario, a standardized assessment form ranging from 0 to 12 points was used to evaluate the participant on how the patient’s injuries and critical conditions were recognized and whether treatment measures were initiated in accordance with the principles and priorities of ATLS® according to the ATLS® Student Manual 10th edition [4] and the Guideline on the Management of Severly Injured Patients of the German Trauma Society [3]. Before the testing, twelve critical tasks were defined for each scenario, which needed to be completed during the training scenario. One point was given for completing each task resulting in a maximum score of 12 points. Manual skills, e.g., intubation or chest drain insertion, were not assessed.

After each scenario, the participants were asked to complete questionnaires on perceived workload during the simulated cases. For this purpose, the NASA Raw Task Load Index (NASA-RTLX) was utilized [5]. The NASA-RTLX is a multidimensional score with six subscales: mental demand, physical demand, temporal demand, frustration, effort, and performance. After the completion of each scenario, the performed tasks were rated by the participant in the subscales within a 100-points range. In addition, the overall workload is calculated by a combination of all six dimensions.

Statistical analyses were performed using SPSS 27.0 (SPSS Inc., Chicago, IL, USA). The data were summarized as mean with standard deviation (SD). Where applicable, nominal variables crosstabs were associated using Chi-square or Fisher’s exact tests. Student’s t test was used to detect differences in means of normally distributed continuous data. Paired tests were used for comparison of the two scenarios among the same participant. The level of significance was defined as p < 0.05.Assuming an effect of at least 2 performance points (SD 2) when using TraumaFlow and aiming for a power of at least 0.80, the minimum sample size was calculated to be n = 10.