Predictors of Participation in a Fire Department Community Canvassing Program

Research output: Contribution to journalArticlepeer-review


An urban fire department has been distributing free smoke alarms for more than 30 years. A community-academic partnership was developed to conduct a community intervention trial as part of the fire department's home visiting program. The trial comprised 170 canvassing events held across 12 census tracts; half of the census tracts were assigned to the treatment condition and received prepromotion of the home visit events. The objectives of this analysis were to identify environmental and programmatic predictors of 1) whether someone would be at home at the time of a visit, and 2) if at home, whether the resident would participate. A separate multilevel analysis was conducted to address each objective. The canvassing event served as the first level to account for variation in implementation of the program, with the census tract as the second level. All environmental and program characteristics were included as fixed effects in both models. Throughout 170 events, 8080 eligible residential addresses were visited, of which 3216 had someone at home, and 2197 homes participated in the program. Canvassing events held on weekends and during the evening hours was associated with higher odds of a resident being at home. Canvassing events without rain and held in the treatment census tract areas was associated with higher odds of resident participation. Environmental and programmatic factors can impact the reach of home visiting programs. These findings can contribute to emerging best practices for fire department home visiting programs.

Original languageEnglish (US)
Pages (from-to)225-229
Number of pages5
JournalJournal of Burn Care and Research
Issue number4
StatePublished - 2017

ASJC Scopus subject areas

  • Surgery
  • Emergency Medicine
  • Rehabilitation


Dive into the research topics of 'Predictors of Participation in a Fire Department Community Canvassing Program'. Together they form a unique fingerprint.

Cite this