Table 2.

Model 1 results of binary logistic regression examining post-release heroin use.

	Adjusted Odds Ratio	Standard Error	95% Confidence Interval	p-Value
Age	0.96	0.011	0.94–0.99	.002
White	4.84	2.177	2.00–11.06	.001
Central Appalachia	0.38	0.091	0.24–0.60	.001
Past-year depressive symptoms	1.42	0.290	0.95–2.11	.089
Past-year suicidal ideation	2.55	0.996	1.17–5.42	.017
Abstinence self-efficacy prior to treatment	0.90	0.192	0.60–1.37	.654
Social satisfaction	0.42	0.101	0.25–0.67	.001
Age of drug use initiation	0.99	0.022	0.95–1.04	.903
Ever IDU	1.34	0.296	0.87–2.07	.180
30-day drug use prior to incarceration Prescription sedatives	0.95	0.203	0.62–1.44	.812
Cocaine/crack cocaine	1.43	0.297	0.96–2.15	.080
Amphetamines	0.58	0.121	0.38–0.87	.008
Heroin	5.32	1.191	3.54–8.25	.001
Prescription opioids	2.05	0.468	1.31–3.21	.002
Buprenorphine	0.72	0.168	0.46–1.13	.154
Lifetime drug treatment episodes	.1.02	0.022	0.98–1.06	.285
Nights spent in controlled environment	1.00	0.001	1.00–1.01	.005

Intercept-only: −533.336/Model: −405.742 Pseudo R² = 0.239.

Wald Chi² (df = 18) 220.306 Cox-Snell/ML = 0.151.

Hosmer-Lemeshow Chi² = 2.39, p = .966 Cragg-Uhler/Nagelkerke = 0.305.