The study was carried out in 3 phases:

• 1)

  Development of a foldable educational tool by a group of professionals.

• 2)

  Testing of the foldable tool by a group of experts by means of a Likert scale.

• 3)

  Pilot study in 120 schoolchildren, with 1 experimental and 1 control group (Fig. 1). The study adhered to the principles of the Declaration of Helsinki and was approved by the Ethics Committee of the School of Education and Sports Science of the Universidad de Vigo, Spain (code 14-2802-18).

  
  

  Figure 1.

  Flowchart of the study.

  (0.55MB).

A team of 7 professionals in the fields of education, art, paediatrics and nursing worked together to design the learning aid for use in the school setting, which was based on character drawings. The content was developed taking into account international CPR guidelines and based on the links of the “chain of survival”.11 The development process followed the Delphi method, which is a systematic approach to obtain expert opinions with 4 key characteristics: expert input, interaction with feedback, statistical group responses and confidentiality.13,14 In the study presented here, we used the first 2, and defined the experts as “specialists in their fields” and “individuals with experience in BLS education”. The resulting tool was a folding picture cube with 6 faces that can completely unfold into 2 rectangular mini-posters (Fig. 2).

Figure 2.

Text and images of the Rescube folding educational tool.

(1.12MB).

After the initial design was completed, we sent the Rescube to 11 experts, none of which had participated in the development of the tool or of the study. The inclusion criteria were: (a) to be an active professional in education or health; (b) to have experience in teaching first aid and (c) to be a professor in the field of medicine, nursing or sports science at the college level. The content of each item and the suitability of each of the associated illustrations were assessed by means of a Likert scale with 4 answer options (not relevant, somewhat relevant, quite relevant, very relevant) (Fig. 1), as was done in a previous study.15 We followed the recommendations of Yusoff for the validation process.16

After validating the contents of the cube, we carried out a pilot study in 120 schoolchildren aged 5–8 years. The study had a quasi-experimental design with block randomization and a control group aimed at assessing learning and long-term retention in 2 groups of 60 children each: the experimental group (RG) was trained with the Rescube folding cube, while the control group (CG) was trained with a conventional training programme specific for children.17,18

We obtained informed consent from the parents or legal guardians of participants and from the school. The inclusion criteria were the following: (a) obtention of informed consent from parent/legal guardian; (b) lack of previous knowledge of BLS and C) absence of physical or cognitive impairment.

As described in previous studies conducted in schoolchildren,19,20 the intervention consisted of 3 tests for assessment of practical skills (T0), learning (T1) and long-term assimilation of learned contents/skills (T2).

Taking into account the cognitive development and attention span of children in this age group, the training had a duration of 15minutes with a 1:5 teacher-to-student ratio. A training protocol was developed so that the same information would be delivered to every group, focused on the first step of the “chain of survival” (recognition of cardiac arrest and activation of emergency response system). The training followed the sequence explanation-demonstration-practice.

In the CG, the explanation part consisted of direct instruction by the teacher using the manikin, after which children practiced on the same manikin. In the RG, the explanation part involved going over each of the illustrations in the Rescube (Fig. 2) with a guided discovery approach (to engage students mentally), after which children practiced on the teddy bear.

• •

  Materials. In the CG, practical skills were tested using a paediatric manikin (Prestan® professional child manikin, Ohio, USA) and a toy cell phone. The materials used in the RG were the Rescube, a large teddy bear and a toy cell phone.

• •

  Directions. For every test, children were given the same directions (previously tested in another pilot sample to ensure comprehension) by the same individual (a teacher in the school, trained in first aid and with experience in experimental research): “Imagine that you find someone laying on the ground like this manikin/teddy bear, and does not seem to be moving or breathing. What would you do?”

• •

  Testing timeline. T0 was administered before the training, T1 a week after the training and T2 one month after the training.

Each participant was assessed individually under the supervision of an expert using a checklist that included each of the variables with a dichotomous yes/no answer format (Fig. 1).

We performed the statistical analysis with the software IBM SPSS Statistics version 20 for Windows.

To calculate content validity indices, we followed the recommendations published by Yussof.16 First we dichotomised the evaluation data, transforming “very relevant” and “quite relevant” choices in the Likert scale to “1” and the “somewhat relevant” and “not relevant” choices as “0”.21 Afterward, for each item, we counted the number of experts that agreed with the content and the images, and we documented the items for which the experts reached full consensus. With the resulting data, we calculated the following indices for the written content and images of the folding cube:

• •

  Item-level content validity index (I-CVI): proportion of experts that considered the item relevant out of the total experts.

• •

  Scale-level content validity index (S-CVI): the proportion of all items considered relevant out of all items in the instrument.

• •

  Scale-level content validity index based on the universal agreement (S-CVI/UA): proportion of items considered relevant by all experts.

The CVI values considered acceptable when more than 9 experts are evaluating an item or instrument are those of .78 or greater.22

We expressed demographic variables (sex and age) by means of absolute and relative frequencies. To assess learning in participants, we first described the variables under study as relative frequencies. We calculated frequencies for each group under study (Rescube vs conventional) and for each of the administered tests (T0, T1 and T2). We also graphed the assessment data by study group, time point and age group (5–6 years and 7–8 years). Subsequently, we compared the distributions of each study group at different time points. We did the same for each age group. To compare results at different time points we used the McNemar test for paired data. To compare the results in different groups we used the chi square test if the conditions for its application were met (at least 80% of the expected counts in the contingency table equal to or greater than 5) and otherwise the Fisher exact test. We measured the effect size (ES) by means of Cramer’s V. For all tests, we defined statistical significance as a p-value of .05 or less.

Table 1 presents the obtained content validity indices. All experts agreed that the items were very relevant with one exception. One of the experts considered that the item referring to the step “gives precise information to 112” was not relevant, which resulted in an I-CVI of .91. There was universal agreement in the remaining content and pictographic items, resulting in I-CVIs of 1.00.

In the validation of the written content of the Rescube, we obtained a S-CVI of .98 and a S-CVI/UA of .80, while in the validation of the images we obtained a S-CVI and a S-CVI/UA of 1.00. All the values exceeded the threshold recommended by Lynn for more than 9 experts.22

Table 2 presents the demographic characteristics of the participants. The age distribution was similar in the 2 groups.

Table 3 presents the results concerning the assessment of the OHCA recognition and calling the emergency response team. We did not find statistically significant differences in any of the variables under study (P>.05). All participants remained calm during testing at all 3 time points. Both groups performed significantly better at 1 week (T1) and 1 month (T2) post training compared to baseline (T0) in every variable (P<.001; ES≥.71). A lower proportion in the RG checked the responsiveness and breathing (P=.040; ES=.19) in T2 (83%) compared to T1 (95%), a trend that did not occur on the CG, although the proportion of the CG that performed correctly in this item was somewhat lower compared to the RG at both time points (T1, 88%; T2, 83%; P=.43). When we compared the groups, we did not find significant differences between T1 and T2 in any of the variables under study:

• •

  Calls 112: T1 (RG, 93%; CG, 98%); T2 (RG, 82%; CG, 93%).

• •

  Gives precise information to 112: T1 (RG, 85%; CG, 87%); T2 (RG, 90%; CG, 87%).

• •

  Correct full sequence: T1 (RG, 80%; CG, 87%); T2 (RG, 67%; CG, 77%).

Fig. 3 presents the assessment results for the variables concerning the recognition of OHCA and the call to emergency services by age (5–6 years and 7–8 years). When we compared the participants in the RG based on age we did not find differences at 1 week (T1), although we did find improved results in the 7–8 years group at 1 month (T2) in the checking of responsiveness and breathing (P=.002), providing precise information to 112 (P=.047) and performing the full sequence correctly (P=.028). When we compared the participants in the CG, we found greater proportions of participants in the group aged 7–8 years that checked responsiveness and breathing and performed the full sequence correctly in both T1 and T2 (P≤.004). When we compared the subsets of participants aged 5–6 years in the RG versus the CG, we found a higher proportion that checked responsiveness and breathing in the 5–6 age group at T1 (P=.022), and found no differences at T2. In the group aged 7–8 years, a higher proportion of participants in the CG gave precise information to the 112 dispatcher (P=.010) and performed the full sequence correctly at T1 (P=.005), and there were no differences at T2.

Figure 3.

Comparison of results by age group.

(0.42MB).

There are limitations to our study. We carried out the evaluations in the context of simulated scenarios, so we cannot be sure that individuals will act the same way in a real-life situation. Although calling 112 is easily done even from a phone that is not registered with a carrier or does not have a SIM card, the simulation was performed with a toy phone. Similarly, the evaluation was conducted using the training equipment. We did not assess the skills of participants in real situations or with real phones. This was a pilot study conducted in a specific sample, which precludes generalization of the results to the entire population of schoolchildren in the same age range, and further studies are required with more realistic simulations to confirm our results.