Skip to main content
. Author manuscript; available in PMC: 2022 Sep 6.
Published in final edited form as: Drug Alcohol Depend. 2022 Jun 30;238:109550. doi: 10.1016/j.drugalcdep.2022.109550

Table 4.

Multinomial Logistic Regression Model of Correlates of Latent Classes for Substance Use Treatment.

Class 2 (“Comorbid MH/SU Programs”; n = 3715; 23.1%)
Class 3 (“Syndemic-Informed and SGM-Tailored Programs”; n = 2804; 17.6%)
Class 4 (“Comorbid MH/SU and Trauma-Specific Programs”; n = 2195; 13.7%)
AOR (95% CI) AOR (95% CI) AOR (95% CI)

U.S. region
Northeast ref ref ref
Midwest 0.59 (0.51, 0.69) 0.50 (0.43, 0.59) 0.62 (0.52, 0.74)
South 0.67 (0.57, 0.78) 0.84 (0.73, 0.97) 0.78 (0.66, 0.93)
West 0.81 (0.70, 0.95) 1.05 (0.91, 1.22) 1.26 (1.07, 1.48)
Other 0.50 (0.25, 0.99) 1.10 (0.62, 1.92) 0.99 (0.52, 1.87)
Offers Pay Assistance
No ref ref ref
Yes 1.71 (1.51, 1.95) 1.77 (1.58, 1.99) 1.43 (1.25, 1.63)
Type of Organization
Private Non-Profit ref ref ref
Private For-Profit 0.59 (0.52, 0.67) 1.52 (1.36, 1.70) 0.95 (0.83, 1.08)
Non-Federal Government 0.92 (0.76, 1.10) 0.58 (0.46, 0.73) 0.75 (0.60, 0.94)
Federal Government 1.40 (0.98, 1.99) 1.04 (0.68, 1.58) 1.64 (1.15, 2.36)

Note. MH = mental health; SU = substance use; AOR = adjusted odds ratio; CI = confidence interval; ref = reference group. Boldface type indicates a significant AOR. Omitted (reference) category is Class 1 (“No Tailored Programs”; n = 7290; 45.5%), which was characterized by low probabilities of offering any tailored programs. Class 2 (“Comorbid MH/SU Programs”) was characterized by high probabilities of offering programs for clients with comorbid mental health/substance use concerns; Class 3 (“Syndemic-Informed and SGM-Tailored Programs”) was characterized by high probabilities of offering all types of tailored programming, and was the only class characterized by high probability of offering SGM-tailored programming; Class 4 (“Comorbid MH/SU and Trauma-Specific Programs”) was characterized by offering tailored programs for clients with comorbid mental health/substance use concerns and clients with a history of sexual abuse, intimate partner violence or domestic violence, and/or other trauma. All models utilized the bias-adjusted maximum likelihood approach.