Frederick 2000.

Methods	CBA
Participants	1292 children aged between 10 and 11 years from Oxfordshire UK. Number of participants: 657 students in the intervention group and 635 students in the control group.
Interventions	Intervention: IMPS. Teachers were given a resource pack, available for 1 academic year, which covered basic life support training, interactional videos illustrating a range of accidents such as burns and how to respond. This was then followed by a hospital visit, whereby children were given a tour of the accident and emergency department by IMPS trainers. Control: schools with no prior exposure to IMPS. Normal curriculum.
Outcomes	Safety knowledge assessed using a specially developed quiz 5 months after the intervention. A hypothetical basic life support scenario was used to measure observed safety skills and behaviour retained after the intervention. Self‐reported behaviour and safety practices assessed using a validated 'draw and write' test.
Injury mechanisms	Road safety, accidents in the home, fire, electricity, poisons, waterways.
Notes	Control schools were matched on location, size and Standard Assessment Test results. Intervention schools were those that were already enrolled in the IMPS programme.
Risk of bias
Bias	Authors' judgement	Support for judgement
Allocation to intervention/control (selection bias) (for non‐RCT and CBA studies)	High risk	Schools self‐selected an intervention.
Blinding of participants and personnel (performance bias) All outcomes	High risk	No blinding.
Blinding of outcome assessment (detection bias) All outcomes	High risk	For observed outcomes, the trainers were unblinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	Attrition was < 20% for all outcomes and in both intervention and control groups.
Selective reporting (reporting bias)	Unclear risk	Some inconsistencies in the reporting of findings (e.g. between tables and the text).
Other bias	Unclear risk	Tables comparing the characteristics of schools were not included.
Risk of bias due to confounding (for non‐RCTs and CBA studies)	Low risk	Control schools were matched on location, size and Standard Assessment Test results.