Evaluation of three smoke detector promotion programs

Am J Prev Med. 1998 Oct;15(3):165-71. doi: 10.1016/s0749-3797(98)00071-3.


Context: Seventy percent of U.S. residential fire deaths occur in homes without a working smoke detector. To help prevent residential fire deaths, many programs have distributed or installed detectors in unprotected homes. Because persons receiving a detector may not install it and because detector batteries require annual replacement, the enduring effectiveness of these programs may be questioned.

Objective: We evaluated the long-term functional status of smoke detectors distributed to high-risk households in eight areas of Minnesota, Cherokee County (North Carolina), and Oklahoma City (Oklahoma).

Design: Cross-sectional.

Setting: Home visits were made to check the detectors that were distributed 3 to 4 years earlier.

Participants: Randomly selected households from the three detector promotion programs. MAIN OUT-COME MEASURE: At least one working smoke detector.

Results: Participation rates ranged from 72% to 82%. The percentage of evaluation households with at least one working detector ranged from 58% in Oklahoma to 73% in North Carolina. In 76% of households with nonworking detectors, the batteries were either missing or disconnected. When batteries in nonworking detectors were replaced, 83% of the detectors regained function.

Conclusions: Future programs should consider distributing detectors that do not require annual battery changes or find effective ways to ensure that batteries are routinely replaced. Programs should also provide each household with the number of detectors needed to meet the most current recommended standard of the National Fire Protection Agency. The evaluation's participation rates support the practicality of unannounced home visits to evaluate home injury prevention programs in high-risk groups.

MeSH terms

  • Accident Prevention*
  • Accidents, Home / prevention & control*
  • Cross-Sectional Studies
  • Evaluation Studies as Topic
  • Fires / prevention & control*
  • Humans
  • Minnesota
  • North Carolina
  • Oklahoma
  • Protective Devices / statistics & numerical data*