Skip to main content

Table 3 Logistic regression models of predictors associated with support for harm reduction programs in British Columbia (n = 1834)

From: A cross-sectional study of public attitudes towards safer drug use practices in British Columbia, Canada

Variables Unadjusted p-value1 Adjusted p-value2
OR 95% CI OR 95% CI
Sex       
 Female (referent) vs. Male 0.76 0.61, 0.94 0.010 0.74 0.59, 0.92 0.006
Age       
 19 – 34 (referent)       
 35 – 54 0.90 0.69, 1.20 0.481 0.89 0.67, 1.18 0.412
 55+ 0.64 0.49, 0.83 <0.001 0.63 0.48, 0.83 0.001
Education       
 = University graduate (referent)       
 < High school graduation 0.49 0.32, 0.76 0.001 0.51 0.33, 0.79 0.002
 = High school 0.46 0.46, 0.86 0.004 0.70 0.51, 0.97 0.030
 = Some post secondary 0.60 0.60, 1.16 0.271 0.88 0.63, 1.23 0.443
 = Certificate/ diploma program 0.52 0.52, 0.93 0.014 0.69 0.51, 0.92 0.012
Health authority n (%)       
 VCH (referent)       
 IH 0.91 0.64, 1.29 0.597 0.99 0.69, 1.41 0.934
 FH 0.57 0.43, 0.75 <0.001 0.59 0.44, 0.78 <0.001
 VIH 1.02 0.72, 1.45 0.914 1.12 0.78, 1.60 0.545
 NH 0.92 0.56, 1.53 0.757 1.02 0.61, 1.69 0.948
  1. 1p-values from single d.f., Wald χ2-tests for each parameter in single predictor logistic regression models.
  2. 2p-values from single d.f. Wald χ2-tests for each parameter in a multiple predictor logistic regression model including Sex, Age, Education, and Health Authority.