Table 2. Association Between Patient and Hospital Characteristics and 30-Day Ambulatory Follow-up Among Medicare Beneficiaries Aged 65 Years and Older Treated and Discharged From the ED From 2011 to 2016^a.

Characteristic	HR (95% CI)b
Year of ED visit	1.00 (0.999-1.001)
Age, y	0.997 (0.997-0.997)
Sex
Women	1 [Reference]
Men	0.90 (0.898-0.904)
Race/ethnicity
White	1 [Reference]
Black	0.82 (0.81-0.83)
Hispanic	0.96 (0.94-0.98)
Asian	1.07 (1.04-1.09)
North American Native	0.87 (0.84-0.90)
Otherc	1.01 (0.99-1.03)
Unknown	1.00 (0.98-1.02)
Medicaid eligible
No	1 [Reference]
Yes	0.77 (0.77-0.78)
Hospital size, No. of beds
Large, ≥400	1 [Reference]
Small, 1-99	0.87 (0.85-0.89)
Medium, 100-399	0.99 (0.98-1.01)
Control type
Nonprofit	1 [Reference]
For profit	1.00 (0.99-1.02)
Government, nonfederal	0.93 (0.92-0.95)
Teaching status
Major	1 [Reference]
Minor	0.99 (0.97-1.01)
Nonteaching	0.98 (0.95-0.998)
Urban/rural
Urban	1 [Reference]
Rural	0.75 (0.73-0.77)
Safety-net status
No	1 [Reference]
Yes	0.94 (0.93-0.96)
Region
Northeast	1 [Reference]
Midwest	0.96 (0.94-0.97)
South	1.01 (0.99-1.02)
West	1.01 (0.99-1.03)

Abbreviations: ED, emergency department; HR, hazard ratio.

^aCox proportional hazards model with time to ambulatory follow-up as the outcome and beneficiary age, sex, race, and Medicaid eligibility as covariates.

^bThe multivariable model incorporated year of the visit, principal diagnosis category, beneficiary demographic characteristics and chronic conditions, and hospital characteristics as covariates, including clustering by hospital in a single Cox regression model. Mortality was accounted for as a competing risk. An HR less than 1 indicates a longer time until follow-up visit.

^cThis group includes individuals who self-reported race/ethnicity as Asian, Hispanic, North American Native, and other.