3.1 Study identification and study characteristics
For research question one, 58 studies were identified (see figure 1). Of these, 43% (25 studies) was performed in Europe, whereas only 4% (two studies) were conducted in Africa. For the first aid topics other than ‘resuscitation’, evidence was scarce or even non-existent. Detailed characteristics of the 58 included studies are listed in Additional file 7.
For research question two, 819 references were screened and finally two systematic reviews were included (see figure 2). The first systematic review included 238 (quasi-)experimental studies, conducted in LMIC, studying a large range of interventions aimed at improving school enrolment, attendance, drop-out, completion and learning as primary outcomes . Of interest to our research question were four studies looking at the provision of instructional materials (e.g. textbooks, flip-charts), and 19 studies investigating structured pedagogy interventions (as defined above). The second systematic review was a PhD thesis of 66 (quasi-)experimental studies, conducted in Sub-Saharan Africa, that looked at the same interventions as listed above, as well as interventions focused on improving repetition and retention rates . Of these, we included 16 studies on the use of alternative pedagogical methods, of which three were also included in the category of structured pedagogy interventions. Detailed study characteristics are presented in Additional file 8.
3.2 Best available evidence on the effectiveness of first aid training to children
In the paragraphs below, the findings concerning the effectiveness of first aid training interventions for ‘burns’, ‘bleeding’, and ‘skin wounds’ are discussed in detail. The detailed findings on the other first aid topics, as well as the risk of bias and certainty of evidence, can be consulted in Additional file 7.
In total, we identified six studies that included burns in their training programme (see Table 1 for study characteristics, and Table 2 for study findings) [25-30]. A statistically significant increase in first aid knowledge was shown after attending a first aid training course, compared to the baseline situation or control group without training, in children of 6-7 years old (MD(%): 27, 95% CI [11;40]) , 10-11 years old (MD: 0.2, 95% CI [0.08;0.32]) , 10-15 years old (OR: 1.83, p=0.026) , and 11-15 years old (RR: 19.80, p<0·001, immediately after the course; RR: 20.00, p<0.001, after two months follow-up) . Two studies also measured the children’s skills, by providing an audio-recorded scenario with a severe burn injury in a toddler or a description of a situation requiring first aid for burns [25, 30]. In a pre-post study with 11- to 15-year-olds, a significant increase in first aid skills was found (RR: 189.00, p<0·001, immediately after training; RR: 149.00, p<0·001, after two months follow-up) . However, in a study with 11- to 16-year-olds that used a placebo training (on tobacco and alcohol prevention) as a control, no increase in skills concerning the order of first aid responses and listing the correct procedures, could be demonstrated (RR: 7.52, 95% CI [0.89;63.69] and RR: 1.05, 95% CI [0.91;1.21], respectively) .
Three studies reported on the effectiveness of first aid training concerning bleeding and skin wounds [25, 29, 30]. A statistically significant increase of knowledge concerning first aid for bleeding or skin wounds was found in children of 11 years onwards. The study by Campbell et al. in 11- to 16-year-olds showed a significant improvement in skills concerning the order of first aid responses in case of bleeding or skin wounds, but not in listing the correct procedures . The certainty of the evidence was very low for the three topics.
3.3 Best available evidence on the effectiveness of educational interventions in LMIC
We identified four studies on the provision of instructional materials to primary schools or their individual students in three different LMIC (India, Kenya and Sierra Leone), from the meta-analyses by Snilstveit et al. (2015) . It could not be demonstrated that the provision of textbooks, flip-charts, or grants used directly for the purchase of materials, results in a statistically significant increase in composite test scores (SMD: 0.01±0.01, 95% CI [-0.01;0.02], p=0.23), language arts test scores (SMD: 0.00±0.01, 95% CI [-0.02;0.02], p=0.78) or mathematics test scores (SMD: -0.02±0.02, 95% CI [-0.06;0.02], p=0.26). The final certainty of evidence was low.
In total, 17 (quasi-)experimental studies provided 41 effect sizes on using alternative pedagogical methods on learning or testing outcomes of students attending primary or secondary schools across seven African countries (Nigeria, Kenya, Ghana, South Africa, Uganda, Liberia and Mali) . Compared to the use of conventional teaching methods (mostly lecturing), the use of alternative pedagogical methods, such as problem-solving instruction, small-group instruction, guided-inquiry instruction, cooperative instruction and constructivist instruction, was shown to significantly increase the students’ learning or testing outcomes (Cohen’s d: 0.918±0.314, 95% CI [0.25;1.59], p<0.05). The certainty of evidence was rated as low.
Concerning structured pedagogy interventions, we obtained evidence from 19 studies in primary and secondary school students in 12 LMIC (Sudan, Kenya, Uganda, South Africa, Liberia, Mali, India, Cambodia, The Philippines, Chile, Brazil and Costa Rica) . When comparing these interventions to no or other small educational interventions, a statistically significant increase in composite test scores (SMD: 0.06±0.01, 95% CI [0.03;0.08], p<0.0001), language arts test scores (SMD: 0.23±0.05, 95% CI [0.13;0.34], p<0.001) and mathematics test scores (SMD: 0.14±0.03, 95% CI [0.08;0.20], p<0.001) was observed. Significant changes in cognitive test scores could not be demonstrated (SMD: 0.01±0.03, 95% CI [-0.04;0.07], p=0.66). The certainty of evidence was downgraded to low.
Study findings are provided in Table 3, whereas study characteristics and determination of the certainty of evidence can be found in Additional file 8.
3.4 Educational pathway on first aid for Sub-Saharan Africa
The best available evidence on the effectiveness of first aid training to children was used to draft the educational pathway, which was then discussed with the expert panel for context adaptation. Since the certainty of evidence was low to very low, the expert panel had a very important role in formulating and approving training objectives and age ranges. Overall, the panel did not often disagree on the specific age ranges at which certain learning objectives should be encouraged, known or repeated. However, for some topics there was more disagreement among the panel members and a more extensive discussion was necessary, e.g. the inclusion of major incidents/disaster management as a topic, the inclusion of psychological first aid as an intervention, the management of fever, the competence of putting on gloves, the objective of willingness to touch a stranger. In each case of disagreement, arguments were listed and discussed, and a proposal was made by the chair until full consensus was achieved.
An example of how the pathway was adapted to the African context deals with seeking help from a medical care provider. In the draft pathway, it was proposed that children should know how to seek help from a medical care provider at the age of 7-8 years. Since medical care is less accessible in Africa than in high income countries, the expert panel decided to postpone this to the age of 9-10 years, and to keep on repeating until the age of 18 years. A second example deals with the general first aid competence of handwashing before and after administering first aid. Because of the higher prevalence of infectious diseases in the African context, the panel proposed to repeatedly train children until the age of 18 years.
For the topic of burns, evidence showed that children of 6-7 years old can be taught how to correctly provide first aid . The expert panel extrapolated the evidence on burns knowledge in a consistent way to the topics of bleeding and skin wounds, since evidence on the latter topics for children under 11 years was lacking, hence concluding that children should have acquired the basic knowledge at the age of 7-8 years. More advanced knowledge should be attained at the age of 11-12 years (e.g. children know the link between tetanus and skin wounds), or at the age of 13-14 years (e.g. children know the different types of burns). Based on the opinion of the experts, only one knowledge item should already be acquired at the age of 5-6 years: the most common causes of burns (i.e. hot water, fire, flames). Skills competences were set in accordance with the knowledge items.
The final version of the pathway can be found in Table 4 (for the topics ‘burns’, ‘bleeding’, and ‘skin wounds’) and Additional file 9 (for all topics).
3.5 Teaching materials for first aid training to children in Sub-Saharan Africa
The panel members agreed to cluster the 2-year age ranges of the educational pathway into 3 broader age groups for the development of separate teaching materials (5-8 years old, 9-12 years old and 13-18 years old). Also, they agreed on a number of recommended educational methods, applicable to the African context. An overview of these methods, the age groups for which they are (most) appropriate, as well as their strengths and weaknesses, is presented in Table 5. The top three of most appropriate and successful teaching methods for each age group is presented in figure 3. Text messaging and the use of individual worksheets for children were perceived as non-feasible or non-desirable educational methods.
Following the panel meeting, teaching materials were developed for the age group of 9- to 12-year-olds, incorporating the recommended teaching methods for this age group. The competencies to be achieved when following a training using these materials were based on the content of the educational pathway, again for this specific age group.