. 2009 Jun;44(3):926–945. doi: 10.1111/j.1475-6773.2009.00964.x

Table 2.

Multinomial Logit Regression of Health Status in t+2

	Good versus Exc/VG		Fair/Poor Versus Exc/VG		Dead versus Exc/VG
	Odds Ratio	p-value	Odds Ratio	p-value	Odds Ratio	p-value
Health status
Good	7.40	<.001	10.49	<.001	4.56	<.001
Fair/poor	11.38	<.001	131.67	<.001	66.40	<.001
Uninsured	1.35	.005	1.90	<.001	1.35	.440
Post (Medicare)	0.87	.047	0.84	.162	0.59	.026
Uninsured × health status
Good	0.70	.041	0.67	.062	1.18	.746
Fair/poor	0.58	.027	0.42	.003	0.74	.522
Post (Medicare) × health status
Good	0.86	.060	0.98	.868	1.58	.085
Fair/poor	0.94	.698	1.04	.839	1.49	.124
Uninsured × post (Medicare)	0.70	.026	1.03	.903	2.83	.032
Uninsured × post (Medicare) × health status
Good	1.55	.087	0.89	.711	0.44	.208
Fair/poor	1.24	.546	0.72	.413	0.24	.022
Age	1.05	<.001	1.07	<.001	1.14	<.001
Age × Age	1.00	.003	1.00	.561	0.99	.026
Male	1.08	.041	1.16	.002	1.84	<.001
Race/ethnicity
Black	1.34	<.001	1.59	<.001	1.46	<.001
Hispanic	1.31	.002	1.59	<.001	0.82	.244
Other race	1.28	.127	1.07	.667	0.79	.399
Education
High school graduate	0.77	<.001	0.54	<.001	0.66	<.001
Some college	0.76	<.001	0.47	<.001	0.65	.001
College graduate	0.60	<.001	0.31	<.001	0.51	<.001
Region
Northeast	0.96	.479	0.94	.388	1.01	.951
South	0.98	.656	1.16	.013	1.13	.210
West	0.83	.001	1.05	.552	1.03	.820

p-value of the χ² Tests on the Set of Coefficients Representing the Following Null Hypotheses:

Hypothesis						p-value

U_pre=I_pre						.001
U_post=I_post						.000
U_pre=U_post						.021
I_pre=I_post						.024
(U_post–U_pre)=(I_post–I_pre)						.093

Reference group: Exc/VG, Insured, Pre, Female, White, High school drop-out, and Midwest.