Increase in homes with at least one working alarm at follow-up relative to increase with Current Norm
| Independent variable | OR (95% CI) | p Value |
|---|---|---|
| Full Education or Practice | 2.77 (1.09,7.03) | 0.03 |
| Baltimore | 1.83 (0.58 to 5.80) | 0.30 |
| Pilot test home | 0.69 (0.18 to 2.67) | 0.60 |
| Spanish materials | 0.22 (0.02 to 2.06) | 0.18 |
| Black race/ethnicity | 1.60 (.052 to 4.97) | 0.42 |
| More than 3 children | 0.19 (0.03 to 1.04) | 0.06 |
| Education | 0.91 (0.62 to 1.34) | 0.62 |
| Income group | 1.45 (0.79 to 2.65) | 0.23 |
| Income question refused | 1.98 (0.81 to 4.82) | 0.14 |
| GE 1 working alarm pre | 1.80 (0.72 to 4.51) | 0.21 |
| Resident says fire risk is large | 1.38 (0.99 to 1.93) | 0.06 |
| Resident says alarms save lives | 0.49 (0.18 to 1.37) | 0.17 |
| Resident says alarms reduce worry | 1.57 (0.86 to 2.85) | 0.14 |
| Resident says batteries are costly | 2.27 (0.91 to 5.63) | 0.08 |
| Number of perceived barriers | 0.80 (0.58 to 1.10) | 0.16 |
| Cases | 263 | |
| Pseudo R2 | 0.14 |
The coefficient for Full Education or Practice shows the increased likelihood of a treatment arm home having working alarms at follow-up relative to the comparison arm. Education runs from 0 for elementary school to 6 for graduate school. Income group is set to 0 if refused and separated into 4 income groups above that. The count of perceived barriers was computed from KAB items 8–14.
KAB, knowledge, attitudes and behaviour.