Table 6.
Logistic regression models for homeless women’s past-year hospitalization, stratified by drug abuse
| Drug abusers (n=454) | Non-drug abusers (n=444) | |||
|---|---|---|---|---|
| Variable | OR | 95% CI | OR | 95% CI |
| Predisposing factors | ||||
| Ethnicity (vs. white) | ||||
| African American | 0.43 | 0.20, 0.95* | 0.76 | 0.33, 1.74 |
| Latina | 0.63 | 0.25, 1.55 | 0.45 | 0.15, 1.30 |
| Other | 0.20 | 0.07, 0.60** | 1.20 | 0.26, 5.49 |
| Usual place to stay, past 30 nights: traditional housing | – | – | 4.59 | 1.90, 11.06*** |
| Physically abused, past 12 months | 2.15 | 1.17, 3.94* | – | – |
| Socialization frequency | 0.80 | 0.66, 0.98* | – | – |
| Mental health hospitalization, ever | 1.94 | 0.93, 4.04 | 3.18 | 1.45, 6.97** |
| Enabling factors | ||||
| Health insurance, past year | – | – | 2.05 | 0.89, 4.73 |
| Need factors | ||||
| Pregnant, past year | 2.89 | 1.32, 6.35** | 12.0 | 4.95, 29.08*** |
Covariates in regression models differed by drug abuse. No statistics were reported for covariates not in the models.
En dash variable not used in the model
p<0.05
p<0.01
p<0.001.