TABLE 3—
Logistic Regression Models of Predictors of Community-Based Behavioral Health Treatment Engagement: 8 US Midwestern Jails, 2017–2018
| AOR (95% CI) | |
| Model 3: MH treatment engagementa | |
| Race/ethnicity: White | 1.94 (1.11, 3.38) |
| Male | 1.23 (0.72, 2.10) |
| Housing insecurity | 1.56 (0.95, 2.55) |
| Metropolitan county | 0.99 (0.57, 1.71) |
| Prior MH treatment or current Rx | 3.33 (1.95, 5.66) |
| CMH client | 8.33 (4.97, 13.99) |
| Positive K6 score | 1.33 (0.80, 2.23) |
| Substance misuse | 1.02 (0.61, 1.71) |
| Past y jail | 0.85 (0.52, 1.40) |
| Length of stay | 0.99 (0.99, 0.998) |
| Model 4: SUD treatment engagementb | |
| Race/ethnicity: White | 1.87 (1.09, 3.19) |
| Male | 1.62 (0.96, 2.72) |
| Housing insecurity | 1.47 (0.91, 2.38) |
| Metropolitan county | 0.67 (0.39, 1.16) |
| Prior MH treatment or current Rx | 1.14 (0.67, 1.94) |
| CMH client | 9.33 (5.53, 15.74) |
| Positive K6 score | 1.46 (0.89, 2.40) |
| Substance misuse | 1.04 (0.63, 1.69) |
| Past y jail | 0.85 (0.52, 1.38) |
| Length of stay | 0.99 (0.99, 0.998) |
| Model 5: COD treatment engagementc | |
| Race/ethnicity: White | 4.47 (2.082, 9.61) |
| Male | 1.17 (0.638, 2.13) |
| Housing insecurity | 1.66 (0.93, 2.97) |
| Metropolitan county | 2.06 (1.09, 3.89) |
| Prior MH treatment or current Rx | 2.25 (1.12, 4.54) |
| CMH client | 6.95 (3.22, 15.00) |
| COD | 0.70 (0.37, 1.31) |
| Past y jail | 0.96 (0.53, 1.73) |
| Length of stay | 0.99 (0.98, 0.998) |
Note. AOR = adjusted odds ratio; CI = confidence interval; CMH = community mental health; COD = co-occurring disorder; Rx = medications; SUD = substance use disorder. Sample size was n = 418 for COD and n = 413 for MH and SUD.
a
χ2(10, n = 413) = 161.564, P < .001; Nagelkerke R2 = 43.4%, predicted cases = 76.3%.
b
χ2(10, n = 413) = 132.509, P < .001; Nagelkerke R2 = 37.0%, predicted cases = 74.6%.
c
χ2(9, n = 418) = 82.463, P < .001; Nagelkerke R2 = 29.9%, predicted cases = 83.7%.