3% change from 22.6% to 30.1%). The percentage of control respondents who gave the same answer showed a slight decrease (?0.5% change from 25.9% to 25.4%). Our final step in analyzing the behavioral variables focused on those variables that showed a significant difference pre- www.selleckchem.com/products/BAY-73-4506.html to postintervention for both groups in the bivariate analysis (Table 4). The question ��In the last month, have you smoked on public transportation?�� that was targeted at current and former smokers revealed a significant difference (p = .002) between groups when adjusted for age, marital status, education, work status, and smoker status: the intervention group was 1.65 times more likely to have smoked on transportation before the intervention when compared to after.

The next two outcome variables, targeted at the whole study population, were adjusted for the same factors plus gender. There was no significant difference between the groups in the number who implemented a smoking ban in the home (p = .205). There was a significant difference (p = .011) in the number who had asked someone on public transportation to stop smoking: intervention respondents were 1.24 times more likely to have asked after the intervention. The last question asked whether respondents avoided exposure to smoking. Again there was a significant difference (p = .004) between groups (intervention: 1.14 times more likely). This model adjusted for all of the above factors except smoker status��when included as a predictor variable, the difference disappeared since smokers from neither group avoided other smokers. Table 4.

Multivariate Analysis Given the changes in smoking behavior, the question arises, why did respondents change their behavior? Two questions in both the pre- and postintervention addressed this issue (data not shown). The first asked, ��Why would you quit smoking?�� Options included: to improve one��s health, to save money, on doctor��s advice, and for children��s health. Children��s health had the most significant increase from pre- to postintervention for both groups: the control group increased from 19.8% to 30.5% (p = .007) and the intervention group from 13.7% to 40.7% (p < .001). Also in the preintervention period, significantly more control respondents (p = .043) listed children��s health as a reason compared to intervention respondents. Postintervention, this trend reversed itself, although not significantly (p = .

128). The other responses to motivation for quitting were not significantly different between the two groups. The second question asked, ��Who advised you to stop smoking?�� Of the possible sources (doctors, other health professionals, friends, family, and religious clerics), only ��other health professionals�� Drug_discovery showed a significant increase from pre- to postintervention in both the control (2.0%�C24.6%, p < .001) and intervention villages (3.7%�C38.3%, p < .