. 2016 Oct 1;25(10):1044–1051. doi: 10.1089/jwh.2015.5522

Table 3.

Logistic Regression Models of the Effects of Discrimination on Women's Healthcare Utilization and Perceived Health Status

	Health service use models^b						Health status models^c
	Every 3 years or less		Every 1–2 years		More than once a year		Very good/excellent
Discrimination variables^a	OR	95% CI	OR	95% CI	OR	95% CI	OR	95% CI
Mean discrimination score	Base outcome		1.21	0.89–1.64	1.36^*	1.01–1.83	0.65^***	0.54–0.80
Individual discrimination items
“treated with less courtesy”	Base outcome		1.09	0.89–1.26	1.12	0.91–1.31	0.80^***	0.70–0.91
“receive poorer service”	Base outcome		1.09	0.84–1.41	1.13	0.88–1.45	0.70^***	0.67–0.93
“you are not smart”	Base outcome		1.18	0.96–1.46	1.31^**	1.08–1.60	0.81^***	0.71–0.92
“people are afraid of you”	Base outcome		1.15	0.91–1.47	1.32^*	1.03–1.68	0.80^***	0.68–0.93
“threatened or harassed”	Base outcome		1.19	0.90–1.59	1.23	0.92–1.66	0.71^***	0.59–0.85

p-Values significant at ^*<0.05, ^** <0.01, and ^*** <0.001. Point estimates are from models with each discrimination indicator mean score modeled as primary independent variables in separate models. Models controlling for age, race, education, income, marital status, employment situation, political affiliation, insurance type, and perceived health status. Results are presented as OR and 95% CI.

^{^a}

Discrimination variables derived from the short version of the Everyday Discrimination Scale⁷

^{^b}

Multinomial logistic regression models with frequency of health service use as outcome.

^{^c}

Multiple logistic regression models with very good/excellent perceived health status as outcome.

CI, confidence interval; OR, odds ratio.