Table 3.

Multivariable generalized estimating equations (GEE) for predicting the odds of county location in high-high cluster.

	High-High Vs. Non High-High Counties1					High-High Vs. Low-Low Counties1
	Beta	Lower 95% CI	Upper 95% CI	Exp (Beta)	P-value	Beta	Lower 95% CI	Upper 95% CI	Exp (Beta)	P-value
% Hispanic	−0.52	−1.13	0.09	0.59	0.093	−0.54	−1.28	0.20	0.58	0.155
% Black (non-Hispanic)	0.94	0.52	1.36	2.55	<0.001	4.14	2.32	5.96	62.74	<0.001
% American Indian (Non-Hispanic)	−0.37	−0.67	−0.08	0.69	<0.05	−0.60	−1.15	−0.05	0.55	<0.05
% Asian (non-Hispanic)	−0.07	−0.30	0.16	0.93	0.540	0.50	−0.19	1.19	1.65	0.156
% Other (non-Hispanic non-White)	0.05	−0.06	0.17	1.05	0.373	−0.16	−0.55	0.23	0.85	0.415
Socioeconomic deprivation index	0.46	0.20	0.90	1.59	<0.05	1.10	0.11	2.10	3.02	<0.05
% Persons aged 65 and older	−0.72	−1.07	−0.37	0.49	<0.001	−0.95	−1.88	−0.03	0.39	<0.05
% Civilian non-institutionalized with a disability	0.32	−0.13	0.78	1.38	0.164	0.80	−0.12	1.72	2.23	0.087
% Civilian non-institutionalized with no insurance	0.60	0.18	1.03	1.83	<0.01	1.38	0.43	2.34	3.99	<0.01
Population density	−0.09	−0.49	0.32	0.92	0.674	−0.31	−0.77	0.15	0.73	0.181
Intercept	−2.68	−3.32	−2.04	0.07	<0.001	−0.05	−1.43	1.33	0.95	0.944
QIC		1708.3					328.9
N (counties)		3107					865

*GEEs are based on binomial logit specification and an independent correlation matrix.