Increasing Smoke Alarm Operability thru Health Education: A Randomized Trial

Background: Although 96% of American homes have smoke alarms, only about 80% have alarms with working batteries. Fire departments and community groups that install alarms provide some education intended to promote fire safety and alarm operability, but it rarely is theory based. Methods: Alarms were installed in more than 500 homes in low income areas, with random cluster assignment to typical fire department education, a health belief model that incorporated the observational learning, reinforcement, and modeling precepts of social cognitive theory, and a compressed form of the theory-driven model that took the same amount of time as existing education but retained hands-on practice installing a replacement battery. Follow-up visits checked alarm operability 500-700 days after initial installation. After dropping cases because the original batteries still were working or all working alarms were replaced with hard-wired ones, 351 cases entered the regression analysis. Results: Relative to the typical fire department approach, the intervention increased the likelihood that homes had at least one working smoke alarm at follow-up (odds ratio 3.3, logistic regression, p=0.98), the average number of working alarms per home by 0.25 (Poisson regression, p=0.99) (or estimated another way, the growth in working alarms post versus pre by 0.4; linear regression, p=0.99). The reductions achieved with the condensed and full education programs were virtually identical with each independently significant at the 90th percentile or above. Conclusion: With the education time available, fire departments and community groups can achieve greater smoke alarm operability.