|Logistic regression results for each dependent variable: re-use of needles, re-use of cookers, re-use of filters and re-use of filters|
|Independent variables||Adjusted ORs*||Hosmer Lemoshow Test|
|Logistic regression 1: Re-use of needles|
|Age ≥ 25 years||.280||.087||.902||.033||4.157 (6)||ns|
|Logistic regression 2: Re-use of cookers|
|HCV +ve||0.445||.206||.961||.039||1.566 (4)||ns|
|Logistic regression 3: Re-use of mixing/rinsing water|
|Food insecure||2.591||.982||6.837||.054||0.479 (4)||ns|
|Logistic regression 4: Re-use of filters|
|Food insecure||3.112||1.082||8.956||.035||2.930 (5)||ns|
* Based on bivariate analyses, we entered the following independent variables into the regression analyses - age under 25 years versus 25 years and over, injected outdoors in the past 6 months, self-reported Hepatitis C status positive versus negative, injected opiates in the past 6 months, injected crack cocaine in the past six months and food insecure (i.e., did not have enough to eat because of a lack of money - and above show only variables found to be significant in the logistic regressions. Adjusted odds ratios refer to the contribution of each independent variable after controlling for the contribution of other significant, independent variables in the logistic regression model.
Strike et al.
Strike et al. BMC Public Health 2012 12:1058 doi:10.1186/1471-2458-12-1058