Design
A mixed methods research design was employed. The aim was to achieve a more comprehensive approach to the investigated topic, measurements, and the analysis and interpretation of findings [19]. A cohort research method was conducted to obtain quantitative data and a focus groups research method was employed to obtain qualitative data. Both methods were combined in the interpretation stage.
Setting and participants
Cohort study
The cohort study included Slovenian elementary schools (15) which offered CPR training in April and May 2018. Information on which schools and classes actually offered CPR training in that period were obtained from the preventive medicine and health promotion services (part of community health centers); they informed us in advance of the planned CPR trainings. CPR training is not mandatory in Slovenian school curricula. Schools make the decision themselves whether or not to include CPR training in their programs. For research purposes, instructors covered the same educational content and used the same teaching methods; they followed the national CPR program based on the European Resuscitation Council guidelines. Thus, the cohort study included all schoolchildren of the seventh, eighth, and ninth grades of 15 elementary schools which provided CPR training in the observed timeframe. Before CPR training with the use of AED was offered, 893 elementary school children were invited to participate in the study and 764 responded (85.6% response rate). One to two months after CPR training, only those schoolchildren who received the training were invited to participate (n=764), and 566 responded (74.1% response rate).
No significant differences were established in the sample structure prior to and after CPR training according to gender, age distribution, parents’ education, and body mass index (BMI) of schoolchildren, which means that both samples were uniform. The gender distribution in the sample is fairly equal, with just under 50% of participants being boys and just over 50% being girls. Children in the final three grades of elementary school were aged 12-15 years. The mean age of seventh-graders, eighth-graders and ninth-graders was 12.5, 13.5 and 14.5 years, respectively. Distribution of schoolchildren among the three grades was equal prior to and after training.
Focus group
Out of 12 invited experts, eight participated in focus groups: instructors and CPR program developers working as nursing professionals in primary health care. Six participants were women, two were men, and their average length of service was 21.4 years (SD = 12.8). One participant held a master’s degree in nursing, six were registered nurses, one was an emergency care assistant nurse. To start the debate, semi-structured guiding questions were used.
Instrument
Cohort research instrument
Knowledge of CPR was measured using a structured questionnaire consisting of four sections with 27 nominal-level binary questions based on previous relevant research [15, 20, 21, 5, 6]. The sequence of items was designed according to the structure of the research problem. Images were used to increase schoolchildren’s interest in filling out the questionnaire. Validity of the instrument was ensured by setting clear evaluation criteria and by making sure the questions were clearly formulated. Each correct answer was awarded one point; the point total was provided to compare results prior to and after CPR training. In the regression models, the independent variable ‘Knowledge’ was used with a score range of 0 to 15. According to the final point count, respondents were categorized: excellent (15-13 points), very good (12-11 points), good (10-9 points), satisfactory (8-7 points), and unsatisfactory (6 points or less). The level of prosocial behavior was assessed according to the level of knowledge. Respondents were divided into two groups: the first included those with excellent knowledge of CPR and the second those with less than excellent CPR knowledge.
Prosocial behavior was gauged with an ordinal scale of opinions consisting of 22 items. Items were designed based on a literature review [6, 15, 17, 21]. The level of agreement was measured on a 5-point scale (1 = Strongly Disagree, 2 = Disagree, 3 = Neither Agree Nor Disagree, 4 = Agree, 5 = Strongly Agree). The reliability of items on prosocial behavior was good (α = 0.82). The demographic part of the questionnaire included questions on age, grade, month and year of birth, body weight, body height, and parental educational level.
Pilot study
Two pilot studies were carried out. The first one was conducted prior to CPR training and included 66 schoolchildren. For 26 items on prosocial behavior, Cronbach’s alpha was .78. We excluded four items with low reliability. Additional instructions were provided for nominal-level binary questions used to test schoolchildren’s CPR knowledge in order to clarify the question and the possible number of answers. With these improvements to the questionnaire, the second pilot study was carried out after CPR training. Cronbach’s alpha for 22 items on prosocial behavior was .81, with 63 schoolchildren from the same cohort participating. Participants did not indicate any possible unintelligibility of items on CPR knowledge.
Focus group instrument
Semi-structured guiding questions were used for the focus groups and expanded as the discussion developed. Research reliability was ensured by following the prescribed methodology: the appropriate number of focus groups (n = 3) and the number of participants in each group (n = 8), which ensured data saturation in the third focus group.
Data collection
Cohort study data was collected twice from April to June 2018: prior to CPR training and one to two months following the training (May, June 2018).
Focus groups were convened after the implemented CPR training, in September and October 2018. The focus groups were led by a moderator employing a semi-structured questionnaire. All selected research participants received an invitation with a brief description of the research, general topics, and partial results of the quantitative research. The average duration of each focus group was 60 minutes. Discussions were recorded with participants’ written consent.
Data analysis
Data was analyzed using IBM software SPSS Statistics v. 22.0. Descriptive statistics was used to present the results of CPR knowledge test and the answers to the questions on prosocial behaviour, chi-square test was used to analyse the differences in the percentage of correct answers prior to and after CPR training. For establishing the differences in the level of knowledge and prosocial behaviour components prior to and after CPR training, we used a) t-test for comparing two samples, and b) one-way Anova for testing the differences between three groups (age groups) for a particular point in time. The common variables were defined using principal axis factoring (PAF) with orthogonal Varimax rotation. The level of statistical significance was set at p ≤ 0.05.
For qualitative data, the method of thematic content analysis was employed. All recordings were transcribed verbatim and the texts were read several times. After coding units were identified, coding was conducted and categories and key themes were defined. Each focus group was ascribed a corresponding code. The nominal identity of a transcription was lost while the traceability of content was ensured.