Table 4. Multiple logistic regression model of independent associations between problematic substance use and unmet mental health care needs.
| Have unmet needs formental health care inpast 12 months (n = 278) | No unmet needs formental health care inpast 12 months (n = 906) | Adjusted OR for having unmetneeds for mental health care inpast 12 months (95% CI) (n = 1115) | |
| Problematic drug use1, n (%) | 143 (52) | 198 (22) | 3.06 (2.17, 4.30) |
| Problematic alcohol use2, n (%) | 57 (21) | 129 (14) | 1.27 (0.85, 1.91) |
| Housing status, n (%) | |||
| Vulnerably housed (ref) | 135 (49) | 456 (50) | 1.00 |
| Homeless | 143 (51) | 450 (50) | 1.45 (1.06, 1.98) |
| Lifetime duration of homelessness (in years), mean (SD) | 5.4 (6.2) | 5.0 (6.0) | 0.99 (0.96, 1.02) |
| Gender, n (%) | |||
| Female (ref) | 103 (37) | 285 (32) | 1.00 |
| Male | 167 (61) | 607 (67) | 1.11 (0.78, 1.58) |
| Transgender | 6 (2) | 12 (1) | 1.39 (0.42, 4.56) |
| Age (in years), mean (SD) | 40.9 (9.7) | 42.6 (10.8) | 0.99 (0.98, 1.01) |
| Racial/cultural group, n (%) | |||
| White (ref) | 180 (67) | 538 (61) | 1.00 |
| Black/African-Canadian | 13 (5) | 93 (10) | 0.73 (0.37, 1.45) |
| First Nations/Aboriginal | 52 (19) | 150 (17) | 0.96 (0.62, 1.47) |
| Mixed ethnicity | 12 (5) | 51 (6) | 0.69 (0.33, 1.43) |
| Other | 11 (4) | 48 (6) | 1.15 (0.52, 2.55) |
| Employed in past 12 months, n (%) | 107 (39) | 362 (40) | 1.07 (0.77, 1.50) |
| Incarcerated in past 12 months, n (%) | 102 (37) | 235 (26) | 1.23 (0.88, 1.73) |
| City, n (%) | |||
| Vancouver (ref) | 108 (39) | 286 (32) | 1.00 |
| Toronto | 66 (24) | 330 (36) | 0.71 (0.47, 1.06) |
| Ottawa | 104 (37) | 290 (32) | 0.88 (0.61, 1.23) |
| Number of chronic health conditions3, n (%) | |||
| 0 (ref) | 19 (7) | 131 (15) | 1.00 |
| 1 | 34 (12) | 214 (24) | 0.91 (0.46, 1.82) |
| 2 | 42 (15) | 154 (17) | 1.64 (0.82, 3.28) |
| ≥3 | 183 (66) | 407 (45) | 2.67 (1.43, 4.98) |
| Ever diagnosed with a mental health problem, n (%) | 206 (75) | 398 (45) | 2.98 (2.12, 4.20) |
| Has a primary care provider, n (%) | 175 (63) | 541 (60) | 0.80 (0.57, 1.12) |
Note: Coefficients are based on a multiple logistic regression model. Bolded coefficients are significant at the p<0.05 level. Nagelkerke R2 = 0.23.
Due to missing data, there were slightly varying denominators for reported frequencies.
OR: Odds ratio; CI: Confidence interval.
1
DAST-10≥6.
2
AUDIT≥20.
3
For a definition of chronic health conditions please see footnote of table 1.