TABLE 2.

Results of Mixed Logistic Regression Model for State Benefit Design and Program Features

Variable	Adjusted Odds Ratio	Confidence Interval
		Lower	Upper
Benefits Design
Benefits restrictiveness	*0.65	0.43	0.97
Covers oral naltrexone	1.87	0.34	10.30
Covers methadone	0.67	0.15	2.93
Number of utilization limits	0.95	0.84	1.06
Percent residents Medicaid eligible	1.03	0.86	1.22
Program Capacity
Use of electronic health records	**2.75	1.37	5.52
Ability to bill insurance	*2.38	1.16	4.89
Percent staff with computers	*1.02	1.01	1.03
Physician employed on staff	0.96	0.41	2.23
Physician required for treatment	0.73	0.19	2.78
Control Variables
Nonprofit	**3.33	1.35	8.21
Publicly-owned	*4.02	1.21	13.33
Offers inpatient treatment	1.33	0.36	4.91
Offers residential treatment	**0.32	0.13	0.77
Opioid treatment program (OTP)	0.72	0.19	2.68
Accredited	2.06	0.86	4.93
Average Caseload	0.98	0.71	1.34
Number of Tx Programs in county	0.10	0.99	1.00
Methadone* OTP interaction	*3.21	1.02	10.15

^*p<0.05;

^**p<0.01