Two-arm pilot community-based randomised control trial for the evaluation of a training intervention.
Sample and sampling method
Professionals from the primary healthcare teams working in the healthcare facilities of the Catalan Health Institute of the Central Catalonia Territorial Management enrolled in the BLS certification courses that were scheduled in the different health care centres. The University Institute for Research in Primary Health Care Jordi Gol i Gurina (Barcelona, Spain) ethics committee approved the trial study protocol (approval code: 20/195-P).
The study is based on two arms: a control arm, which followed training in the usual role-playing methodology, and a second intervention arm, which followed BLS training with a VR programme.
In the total of 20 Basic Life Support certification training activities in the different territories organised by the training unit during the study period (March-December 2021), the intervention and control groups were randomly distributed as follows:
8 intervention groups in the regions of Bages, Berguedà and Moianès.
8 control groups in the regions of Bages, Berguedà and Moianès.
2 intervention groups in the region of Anoia.
2 control groups in the region of Anoia.
A total of 60 students were expected to attend each the intervention and control groups (6 students per group), with a maximum loss of 5% of enrolled students due to unforeseen circumstances, usually of a healthcare nature.
Actions carried out within the control arm. The subjects who underwent BLS training according to the usual methodology received a 15-min MS PowerPoint presentation prepared by the CCR, followed by a role-playing simulation using a sensorized torso and with the trainees performing the functions of simulation training. The total duration of this part was 5 h.
Actions carried out within the intervention arm. The subjects received BLS training based on a self-training strategy that consisted of using technological and computerised VR material. In this first phase, the training was carried out in group format. Once the training was completed, the students, individually, carried out a practical test using the virtual reality goggles. This second phase had a duration of 30 min per simulation scenario, using a minimum of two scenarios. The training was conducted at the workplace (health centres) and each student was assisted by a qualified instructor. The duration of the training was also of 5 h. The VR equipment consisted of a high-performance personal computer, a professional simulation torso with sensors, HTC Vive Cosmos® VR goggles and a carrying case. The software used was the one available in the virtual reality platform LUDUS (https://www.ludusglobal.com) that had a contract with our company during the year 2021. This platform provides of a repository of 5 virtual scenarios in simulation, located in a company, on the street, in a health centre, in a commercial space and in a public site with adverse weather factors. Each student could choose two scenarios to be used in the simulation. This platform has proven experience in the field of virtual reality training content edition.
The main outcome in the two arms of the training intervention is the learning curve of the students. This learning will be assessed by means of the Catalan Resuscitation Council's knowledge questionnaire and will be administered on three occasions: at time zero – before starting the training – to assess participants' baseline knowledge, immediately after completion of the training and six months after finishing the course. For this third evaluation, a reminder will be sent to their corporate email with a link to the intranet page of the teaching unit where they can answer the questionnaire.
The study also aims to assess aspects of satisfaction with the training activity. For this reason, students will be asked to answer a questionnaire on the training received (Annex 1).
Economic aspects will also be evaluated. The researchers will record the time that each student in each arm dedicated to the training, the time dedicated by the teachers and their remuneration and, if applicable, the rental of the equipment used for the training. Once this record has been made, the costs will be compared between the two groups.
Different sociodemographic, academic and occupational variables were recorded for each of the participants in each of the two arms of the study in order to stratify the learning outcomes by their different values. The variables collected were: sex, age, professional profile (clinical assistant, social worker, citizen care or other), job position (primary care centre certified for teaching or not certified) and time (in years) spent working in healthcare.
For descriptive analysis, we expressed continuous variables as median (1r-3r quantiles) and we summarised categorical variables as absolute frequency (percentage). A bivariate analysis of the characteristics of the participating students was carried out to verify that their sociodemographic, academic and occupational data were homogeneously distributed between the two arms.
The bivariate analyses were carried out with the statistical tests that corresponded to the nature of the variables analysed. For dichotomous or polytomous qualitative variables were used Chi Squared test, and for quantitative variables that were distributed or not according to the normal distribution were used Mann–Whitney or Independent T-Test. The normality of the main outcome variable (knowledge scores) was checked and the evolution of its curve (time 0, time 1, time 6 months) was analysed.
A paired comparison was performed for the results of the baseline knowledge test and the score obtained at the end of the training and between this and the score obtained in the questionnaire at six months. The results obtained in each of the test arms were compared. For independent comparisons were used Mann Whitney test, and for paired comparisions were used Wilcoxon test.
The data were analysed with R software, version 4.1.2 for statistical analysis in its most recent version, and the alternative hypotheses were accepted when the probability of an error was less than 5% (p > 0.05).