Social Ecological Based Smoking Cessation Intervention in Public Housing Neighborhoods

This was a cluster randomized controlled study evaluating the effect of a multi-level intervention, Sister to Sister, on the smoking cessation in two Southeastern US metropolitan areas. Fourteen subsidized housing neighborhoods, eight in Charleston, SC and six in Augusta, GA, were pair matched based on size; with one neighborhood in each pair randomized to the intervention to ensure balanced recruitment between groups. Neighborhoods ranged in size from 22 to 33 participants with a total sample size of 409 women; neighborhoods were chosen as the primary sampling unit to prevent contamination of the outcome within neighborhoods, with women in the neighborhoods as secondary sampling unit. All statistical analyses were conducted using SAS 9.2 (SAS Institute, Cary, NC) with statistical significance at an alpha level of 0.05. Comparison between groups at baseline used t-tests for continuous variables and chi-square tests for categorical variables. The primary outcome of interest was smoking status which was dichotomized (smoker/non-smoker) at the 6 and 12 month measurements. Smoking status was determined by cotinine levels in saliva with non-smokers defined as those with cotinine levels of 14 ng/mL or less.