Skip to main content
. Author manuscript; available in PMC: 2022 Oct 1.
Published in final edited form as: Drug Alcohol Depend. 2021 Jul 28;227:108927. doi: 10.1016/j.drugalcdep.2021.108927

Table 2:

Initiation Models

Model Number 1 2 4 5
Race 1 + Controls 2 + Mediators 4 + Interactions

Coefficient Coefficient Coefficient Coefficient
(95% CI) (95% CI) (95% CI) (95% CI)
Black race (vs. White) −0.209** −0.182** −0.139** −0.145**
(−0.242 - −0.177) (−0.215 - −0.148) (−0.172 - −0.106) (−0.194 - −0.095)

Controls:
Ages 30-39 (vs. 18-29) 0.033* 0.02 0.02
(0.003 - 0.064) (−0.010 - 0.050) (−0.010 - 0.050)
Ages 40-49 (vs. 18-29) 0.047* 0.031 0.031
(0.010 - 0.084) (−0.005 - 0.067) (−0.005 - 0.068)
Ages 50-64 (vs. 18-29) 0.053** 0.016 0.016
(0.014 - 0.092) (−0.023 - 0.054) (−0.022 - 0.055)
Male (vs. Female) −0.035** −0.01 −0.01
(−0.060 - −0.010) (−0.035 - 0.015) (−0.035 - 0.015)
SSI Medicaid eligibility −0.149** −0.148** −0.147**
(−0.182 - −0.116) (−0.180 - −0.116) (−0.179 - −0.115)
Expansion Medicaid eligibility 0.021 0.017 0.017
(−0.008 - 0.050) (−0.011 - 0.046) (−0.012 - 0.046)

Possible Mediators:
Non-OUD SUD Diagnosis −0.049** −0.039*
(−0.077 - −0.021) (−0.070 - −0.007)
Days in Allegheny County Jail −0.003** −0.003**
(−0.003 - −0.002) (−0.003 - −0.002)
Months with Housing Support −0.009 −0.006
(−0.017 - 0.000) (−0.017 - 0.005)
Days in ED −0.010** −0.012**
(−0.015 - −0.006) (−0.018 - −0.007)
Intensive SUD treatment utilization (non-MOUD) −0.080** −0.082**
(−0.106 - −0.054) (−0.110 - −0.054)

Black race (vs. White) Interacted with… §
Non-OUD SUD Diagnosis −0.048
(−0.116 - 0.021)
Days in Allegheny County Jail 0.001
(−0.000 - 0.002)
Months with Housing Support −0.005
(−0.023 - 0.013)
Days in ED 0.006
(−0.003 - 0.015)
Intensive SUD treatment utilization (non-MOUD) 0.016
(−0.051 - 0.083)
Constant 0.506** 0.524** 0.602** 0.603**
(0.493 - 0.520) (0.495 - 0.552) (0.571 - 0.632) (0.572 - 0.635)

Observations 6,374 6,374 6,374 6,374
R-squared 0.025 0.043 0.085 0.086
R-squared Adj. 0.025 0.043 0.085 0.086
**

p<0.01,

*

p<0.05

Mediators with which race had a significant relationship in models 3a (appendix) and a significant association with initiation in models 3b (appendix) were included in model 4, in which we regressed initiation on all relevant mediators and control variables.

§

Interaction terms between possible mediators and race to identify any moderation effect by race.