Systematic review of controlled trials of interventions to promote smoke alarms

• Published in: Archives of disease in childhood 2000-05, Vol.82 (5), p.341-348
• Main Authors: DiGuiseppi, Carolyn, Higgins, Julian P T
• Format: Article
• Language: English
• Subjects: accident prevention; Alarm systems; Biological and medical sciences; Burns; Burns - prevention & control; burns prevention; Child Health; Community Child Health, Public Health, and Epidemiology; Control Groups; Controlled Clinical Trials as Topic; Educational Resources; Fatalities; Fires; Fires - prevention & control; Health Promotion; Humans; Injuries; Medical sciences; Meta Analysis; Odds Ratio; Protective Devices - statistics & numerical data; Randomized Controlled Trials as Topic; Smoke; Smoke - adverse effects; smoke alarms; Smoke detectors; Smoke Inhalation Injury - prevention & control; Studies; systematic review; Traumas. Diseases due to physical agents
• ISSN: 0003-9888; 1468-2044
• DOI: 10.1136/adc.82.5.341

AIMS To evaluate the effects of promotion of residential smoke alarms. METHODS Electronic databases, conference proceedings, and bibliographies were systematically searched, and investigators and organisations were contacted, in order to identify controlled trials evaluating interventions designed to promote residential smoke alarms. The following were assessed: smoke alarm acquisition, ownership, and function; fires; burns; and fire related injuries. Odds ratios (OR) were estimated by meta analysis of randomised trials. RESULTS A total of 26 trials were identified, of which 13 were randomised. Overall, counselling and educational interventions had only a modest effect on the likelihood of owning an alarm (OR = 1.26; 95% confidence interval (CI): 0.87 to 1.81) or having a functional alarm (OR = 1.19; 95% CI: 0.85 to 1.66). Counselling as part of primary care child health surveillance had greater effects on ownership (OR = 1.93; 95% CI: 1.04 to 3.58) and function (OR = 1.72; 95% CI: 0.78 to 3.78). Results were sensitive to trial quality, however, and effects on fire related injuries were not reported. In two non-randomised trials, direct provision of free alarms significantly increased functioning alarms and reduced fire related injuries. Media and community education showed little benefit in non-randomised trials. CONCLUSION Counselling as part of child health surveillance may increase smoke alarm ownership and function, but its effects on injuries are unevaluated. Community smoke alarm give away programmes apparently reduce fire related injuries, but these trials were not randomised and results must be interpreted cautiously. Further efforts to promote smoke alarms in primary care or through give away programmes should be evaluated by adequately designed randomised controlled trials measuring injury outcomes.