Table 2. Final multivariate multinomial regression model of baseline factors associated with opioid prescriptions^*.

Patient Characteristics	Moderate risk opioids			High risk opioids
Demographics	IRR	CI		IRR	CI
Age	1.00	1.00	1.00	0.99*	0.98	1.00
Female	1.12*	1.03	1.21	1.32*	1.14        1.53
Medicaid	1.41*	1.24	1.60	1.68*	1.36	2.06
Race/ethnicity (vs. white)
Hispanic/LatinX	1.29	0.81	2.05	1.05	0.42	2.67
Non-Hispanic
Black	1.02	0.66	1.58	1.56	0.65	3.76
Unspecified	0.44*	0.26	0.74	0.69	0.24	1.98
Other specified	0.64	0.37	1.13	0.30	0.07	1.30
Medical Conditions
Cirrhosis	1.04	0.95	1.14	1.22*	1.04	1.43
NAFLD	0.90*	0.82	0.98	0.58*	0.49	0.68
HCC	1.69*	1.08	2.63	1.69	0.85	3.38
Substance use disorder	1.66*	1.41	1.96	2.00*	1.56	2.56
Chronic pain	2.24*	2.07	2.43	3.40*	2.92	3.94
Depression	1.63*	1.45	1.84	1.87*	1.56	2.24
Anxiety	1.16*	1.03	1.31	1.85*	1.54	2.21
Charlson index	1.17*	1.14	1.20	1.27*	1.22	1.32

*Indicates statistically significant association, p<0.05.

High risk opioids are defined as >90MME/day or co-prescription with benzodiazepines; moderate risk are all other prescriptions opioids.

Abbreviations: IRR, incidence rate ratio; CI, confidence interval; NAFLD, nonalcoholic fatty liver disease; HCC, hepatocellular carcinoma.