Table 3 Logistic regression analyses: final backward elimination model of determinants of mental health service utilization
| Â | Â | OR | (95% C.I.) | Â | Wald | df | P |
|---|---|---|---|---|---|---|---|
Psychiatric disorder: | No | 1 | Â | Â | Â | Â | Â |
| Â | Yes | 5.27 | 2.40 | 11.6 | 17.09 | 1 | <0.001 |
Disability from mental illness (≥ 1 month) | No | 1 |  |  |  |  |  |
| Â | Yes | 79.9 | 19.3 | 331.7 | 36.42 | 1 | <0.001 |
Self-rated mental health: | Excellent or very good | 1 | Â | Â | 6.997 | 2 | 0.030# |
| Â | Good | 0.84 | 0.38 | 1.85 | 0.183 | 1 | 0.67 |
| Â | Fair of poor | 2.63 | 0.94 | 7.33 | 3.408 | 1 | 0.065 |
Gender | Male | 1 | Â | Â | Â | Â | Â |
| Â | Female | 2.106 | .953 | 4.655 | 4.788 | 1 | 0.029 |
Formal education | No | 1 | Â | Â | Â | Â | Â |
| Â | Yes | 2.40 | 1.14 | 5.07 | 5.219 | 1 | 0.021 |
Household income per month | < $1,000 | 1 | Â | Â | 7.396 | 2 | 0.026# |
|  | $1,000 – $2,999 | 0.64 | 0.30 | 1.37 | 1.291 | 1 | 0.256 |
| Â | >$3,000 | 0.31 | 0.13 | 0.72 | 7.336 | 1 | 0.007 |
Limited by financial resources for medical care | No at all | 1 | Â | Â | Â | Â | Â |
| Â | Yes, to some or great extent | 0.38 | 0.19 | 0.76 | 7.479 | 1 | 0.006 |
Caregiver available | No | 1 | Â | Â | Â | Â | Â |
| Â | Yes | 0.20 | 0.07 | 0.59 | 8.396 | 1 | 0.004 |
Mental illness can be cured | Not at all | 1 | Â | Â | 6.165 | 2 | 0.046# |
| Â | To some extent | 4.01 | 0.96 | 16.70 | 3.649 | 1 | 0.056 |
| Â | To a great extent | 6.89 | 1.47 | 32.3 | 5.982 | 1 | 0.014 |
Professionals can do little to help | Disagree or strongly disagree | 1 | Â | Â | 10.91 | 2 | 0.004# |
| Â | Neither | 2.78 | 1.33 | 5.81 | 7.431 | 1 | 0.006 |
| Â | Agree or strongly agree | 3.35 | 1.41 | 7.97 | 7.491 | 1 | 0.006 |