Table 4.

Mixed-effect logistic regression model of the relationship between telemedicine and missed appointments.

Variable	OR	MV-adjusted ORa, 95% CI	p-Value
Appointment type
Face-to-face appointment	Ref.
Telemedicine appointment	0.87	0.84–0.89	<.001
Age
18–64	Ref.
<18	0.83	0.79–0.88	<.001
>64	0.71	0.65–0.77	<.001
Gender
Male	Ref.
Female	1.01	0.98–1.03	.64
Ethnicity
Non-Hispanic	Ref.
Hispanic	1.19	1.16–1.23	<.001
Race
White	Ref.
Black or African American	1.61	1.54–1.68	<.001
Asian	0.82	0.74–0.90	<.001
American Indian, Alaska Native, other Pacific Is.	1.19	1.05–1.35	.013
Mixed race	1.22	1.13–1.36	<.001
Insurance coverage
Private insurance	Ref.
Medicare	1.62	1.50–1.75	<.001
Medicaid	1.75	1.68–1.83	<.001
Uninsured	1.90	1.82–1.92	<.001
Service line
Family practice	Ref
Mental health	1.20	1.15–1.25	<.001
Obstetrics/gynecology	0.88	0.84–0.92	<.001
Paediatrics	0.94	0.89–1.00	.05
Senior care	0.74	0.67–0.81	<.001
Metropolitan status
Nonmetropolitan	Ref.
Metropolitan	1.06	0.98–1.14	.15
Distance from clinic
<5 miles	Ref.
5–10 miles	1.02	0.93–1.13	.59
10–20 miles	1.01	0.89–1.08	.71
20–50 miles	1.02	0.89–1.09	.80
≥50 miles	1.00	0.91–1.11	.93
MUA status
Non-MUA	Ref.
MUA	0.95	0.93–0.98	.001
Visit history (December 2018 to February 2020)
1–2 visits	Ref.
3–4 visits	1.67	1.61–1.74	<.001
5+ visits	1.61	1.56–1.66	<.001
Intraclass correlation coefficient for random effect
	ICC	95% CI	Std. Err.
Patient	0.20	0.19–0.21	0.003

^a MV-adjusted OR, multivariate adjusted odds ratio.