Impact on Performance of the Use of a Digital Cognitive Aid in Simulated Crisis and Stress Situations

The epidemiology of battlefield mortality highlights the need for safety and speed of care in an environment under very high stress. Indeed, it has been found that about 90% of those killed in action die before reaching the first hospital or care unit. Management algorithms such as ABCDE, SAFE MARCHE RYAN and local protocols are among the many doctrines to be applied by military health professionals when dealing with serious trauma in emergencies in different practice settings. During combat, their actions are carried out under enemy fire, in darkness, in extreme environments, for a significant number of victims and with limited means. As a result, on the battlefield, this multitude of choices and specific conditions of practice can make it difficult to apply the guidelines for the management of war-wounded in critical and stressful situations and can be a source of medical error. Health simulation is one of the main tools for training and preparation in a safe environment for unsafe situations ("never the first time on the patient"). A new decision-making and protocolization tool is available for health professionals: the MAX (Medical Assistant eXpert) cognitive aid, which has already proven its effectiveness in reducing the number of errors in the management of vital distress in anesthesia and in basic combat casualty care.

Materials and Methods:

Population studied: military doctors and nurses from the forces, both male and female, of varying ages and operational experience, planned to be deployed on External Operations (OPEX) or Short-term Mission (MCD), designated by the Val-de-Grâce School as learners for the " Survival Conditioning of the War Wounded " (MCSBG) continuing training course.

Study design (conduct): analytical, prospective, randomized, controlled, unblinded study conducted during the MCSBG training courses as part of the operational preparation of military doctors and nurses of the forces within the Center for Education and Simulation in Operational Medicine (CESimMO) in Toulon. The investigators had four pairs of learners work on one afternoon during each week-long training session. Each pair of learners completed two different scenarios of similar difficulty and complexity. Four blocks of scenarios were defined on which the pairs were able to work. The pairs were initially randomly assigned to two groups: Group 1 where the first scenario was performed with the MAX CA and Group 2 where the CA was used in the second scenario. Each group therefore performed a simulated scenario with and without MAX. The scenarios used either high-fidelity mannequins (SimMan 3G / ALS, Laerdal, Stavanger, Stavanger, Norway) or hybrid standardized patients (standardized subjects equipped with anatomical prostheses that allowed their management including technical procedures).

MAX CA Description: MAX is a digital AC, developed by J.C. CEJKA, and initially intended for the management of anesthesia and resuscitation emergencies. For the MARCHE RYAN protocol, each action was presented sequentially by the CA and had to be validated before the next one was displayed (READ and DO mode). It was possible to go back in order to review the validated actions and at the end of the procedure, the user had the possibility to make a final check of all his actions. The evaluation of the effectiveness of MAX was previously carried out at the Lyon center for teaching through health simulation (CLESS, UCBL1 - Claude Bernard University Lyon 1), but also during the training of Medicalization in hostile environments (MEDICHOS) and the continuing training in 2nd level combat rescue / casualty care (SC).

Use of the AC MAX in the study: Each participant disposed of a MAX application when dealing with war wounded in simulated situations. The presentation of the application and its handling was done on the first morning of each course, and a license for downloading and using the application was granted to each participant. The instructors were present to answer any questions about this CA. The anonymity of the study participants was ensured and their volunteering as well as the authorization to take pictures and use the videos collected at the beginning of each course.

Participants were requested to absolutely use the application for one of the two scenarios (randomized choice). The fact of being able to use MAX was communicated to them each time during the scenario briefing. More precisely, the one running the MARCHE RYAN algorithm had to use the MAX AC for a complete sequence, from the beginning of the scenario and follow MAX's prompts until the end of the protocol. The participants decided who in the pair will use MAX, however this person had to use both MAX+ and MAX- scenarios according to the MARCHE RYAN protocol. One of the instructors/facilitators who followed this process was to remind them of this requirement during the care of war wounded when necessary, and to impose the use of CA, if applicable.

Randomization by block allocation was performed using open source software R (randomizr library, R Foundation for Statistical Computing, Vienna, Austria, version 3.4.1), to have a homogeneous distribution of participants in Group 1 (MAX first) and Group 2 (MAX second), according to four blocks (1 block per scenario passage). The types of scenarios, the locations where the training took place and the distribution of trainers were the same for all MCSBG courses included in the study. The designation of the pairs was made on the first day of each course by the trainers in title of the CESimMO of Toulon, from the list of trainees participating.

The calculation of the number of subjects required was carried out a priori on the basis of data from the work of R. Lelaidier and M. Truchot studying the effect of MAX CA on technical and non-technical performance. In this work, the scenarios being of similar difficulty, investigators do not expect any significant effect from this factor, so the differences in means will be influenced by the use of MAX. The size of the expected effect was calculated from a mean difference of 13 points, and a standard deviation of sd=11, with n=8 groups (4 scenario blocks with or without MAX), which equates to a MAX CA effect size of 0.4 between the eight groups. The recruitment of 32 pairs will allow investigators to have a power of 0.88 with a first species risk of 0.05. In other words, a population of 32 pairs will allow investigators to show a minimum effect size of f=0.35, with a power of 0.8 and a first species risk of 0.05. There was no interim analysis planned for this study.

Data collection and analysis:

All performances were filmed (using GoPro Hero 4 cameras, San Mateo, California, USA) and anonymized. Each video recording of the MAX+ and MAX- scenarios was remotely rated by two independent evaluators (NP and PR), using the technical and non-technical performance rating grids. The scales used have already been published in international journals. If a difference of more than one point out of 10 was found between the two ratings, a third assessor (AP) would act as mediator.

A questionnaire (quiz) was also distributed to participants at the end of each course to find out how they felt about the influence of the use of the AC MAX on their performance.