Table 5.

Logistic regression—predictors of restart among early discontinuers

	Odds ratio	95% confidence interval		P value
		Lower	Upper
Patient characteristics
Age	1.01	1.01	1.01	<0.001
Femalea	1.07	1.00	1.14	0.041
Northeast regionb	1.06	0.95	1.17	0.301
North Central regionb	1.16	1.07	1.26	<0.001
West regionb	1.03	0.93	1.13	0.574
Health maintenance organizationc	1.07	0.98	1.17	0.152
Point of servicec	1.02	0.90	1.14	0.808
Other health insurancec	0.97	0.86	1.10	0.619
Pre-period
Charlson comorbidity index	0.98	0.96	1.01	0.160
Hospitalized	0.68	0.62	0.75	<0.001
Emergency room visit	0.88	0.81	0.95	0.002
Total costs	0.99	0.98	1.00	<0.001
Microvascular complication	1.20	1.05	1.37	0.006
Macrovascular complication	1.01	0.91	1.11	0.921
Depression	1.18	0.98	1.42	0.077
Obesity	0.87	0.78	0.97	0.015
Dementia	0.98	0.37	2.57	0.960
Alcohol use disorder	0.90	0.67	1.21	0.485
Mental health disorder	0.92	0.83	1.02	0.099
Use of oral antidiabetic medication	1.19	1.07	1.31	<0.001
Number of antidiabetic medications classes	1.11	1.07	1.16	<0.001
Number of classes of all-cause medications	1.04	1.03	1.05	<0.001
Use of antidepressant	1.12	1.01	1.24	0.034
Use of other mental health medications	0.95	0.87	1.05	0.343
Index prescription
Analogd	0.90	0.80	1.00	0.058
Viale	0.70	0.66	0.75	<0.001
Mixturef	0.80	0.73	0.89	<0.001
Copayment	1.00	1.00	1.00	0.001

Reference categories: ^amale, ^bSouth, ^cpreferred provider organization, ^dhuman insulin, ^epen