Table 2.
Multivariate logic regression examining the association between the use of powder only, crack only, powder and crack and durable viral suppression in the following 12 months.
| Variables | OR | 95 %CI | p value |
|---|---|---|---|
| Powder only | 3.88 | 1.87–8.03 | 0.0003 |
| Crack only | 3.25 | 1.56–6.74 | 0.0016 |
| Powder and crack | 1 | ||
| Age (45 or older vs younger) | 0.73 | 0.40–1.34 | 0.3112 |
| Education (above high school vs high school or below) | 1.67 | 0.94–2.98 | 0.0818 |
| Unemployment (yes vs no) | 1.11 | 0.56–2.19 | 0.7661 |
| Low household income (yes vs no) | 1.57 | 0.87–2.85 | 0.1347 |
| Homeless (yes vs no) | 0.28 | 0.14–0.55 | 0.0002 |
| When polysubstance use and frequency of cocaine use were additionally adjusted X1 | |||
| Powder only | 4.18 | 1.93–9.03 | 0.0003 |
| Crack only | 3.96 | 1.77–8.86 | 0.0008 |
| Powder and crack | 1 | ||
| Polysubstance use (number of other substances used) | 1.22 | 0.95–1.56 | 0.1167 |
| Frequency (daily/weekly vs less than weekly) | 1.33 | 0.73–2.40 | 0.3501 |
X1: Age education, unemployment, household income and homelessness were included in the model.