The challenge
Low-income households are more likely to have house fires and less likely to have working alarms.
The case
In a cluster randomised trial from 1997, 40 deprived electoral wards in the London boroughs of Camden and Islington were paired, and 20 were chosen for a giveaway. About 20,050 smoke alarms, fittings and fire safety leaflets were handed out free, with installation on request. Fire-related injuries did not fall: the injury rate ratio was 1.3 (95% CI 0.9 to 1.9), and there was no drop in hospital admissions, deaths or fire brigade call-outs. A household survey explained why: only about 30% of sampled households in the giveaway wards had an alarm installed, much the same as in comparison wards, and only 16% had one that worked.
Why it didn't work
Handing out alarms is not the same as getting working alarms on ceilings. Programmes need to install and maintain alarms, not just distribute them.
Sources
Spot an error or know a better source? Tell us