The impact of recent changes in smoke alarm legislation on residential fire injuries and smoke alarm ownership in New South Wales, Australia.

In 2006, New South Wales (NSW) state legislation changed from requiring smoke alarms in new houses only to all houses. We evaluated the impact of this legislative change on residential fire injury and smoke alarm ownership characteristics. Residential fire injuries for 2002 to 2010 were identified from hospitalization data for all hospitals in NSW. Data relating to smoke alarm ownership and demographic factors were obtained from the NSW Population Health Survey. Negative binomial regression analysis was used to analyze trends over time. Prior to the introduction of universal legislation, hospitalization rates were increasing slightly; however, following the introduction of legislation, hospitalization rates decreased by an estimated 36.2% (95% confidence interval [CI], 16.7-55.8) annually. Smoke alarm ownership increased from 73.3% (95% CI, 72.5-74.2) prelegislation to 93.6% (95% CI, 93.1-94.2) 18 months postlegislation. Thirty percent of households reported testing their alarms regularly. Speaking a language other than English (relative risks [RRs], 1.82; 95% CI, 1.44-2.99), allowing smoking in the home (RR, 1.73; 95% CI, 1.31-2.27), and being part of the most disadvantaged socioeconomic group (RR, 1.47; 95% CI, 1.14-1.91) remain major risk factors for nonownership. Broadening the scope of state legislation has had a positive impact on residential fire-related hospitalizations and smoke alarm ownership. However, it is of concern that the legislation has been the least effective in increasing smoke alarm ownership among non-English-speaking households, in households where smoking is allowed, in low socioeconomic households, and that a high proportion of householders do not test their smoke alarms regularly. Targeted campaigns are needed to reach these high-risk groups and to ensure that smoke alarms are functional.