Table 2.

Regression coefficients confinement conditions for various health outcomes (N = 493)

Confinement condition models	Good overall health (odds ratios)	Worse health in detention (odds ratios)	Diagnosed mental health condition (odds ratios)	Stress index (incidence rate ratios)
Experienced sleep deprivation	0.41***	2.87***	2.41*	1.77***
	(0.11)	(0.65)	(0.87)	(0.09)
Difficulty seeing family	1.00	1.92**	1.00	1.22***
	(0.21)	(0.39)	(0.27)	(0.05)
Experienced verbal or physical abuse, formal discipline	0.76	1.16	1.62+	1.17***
	(0.16)	(0.23)	(0.44)	(0.05)
Witnessed verbal or physical abuse	0.79	1.13	1.16	1.16***
	(0.17)	(0.22)	(0.31)	(0.05)
Difficulty accessing medical care	0.44***	2.19***	1.16	1.24***
	(0.09)	(0.44)	(0.30)	(0.05)
Difficulty accessing psychological/mental health services	0.41***	2.13**	2.68***	1.39***
	(0.10)	(0.54)	(0.79)	-0.06
Index of conditions of confinement	0.76***	1.39***	1.25*	1.15***
	(0.05)	(0.10)	(0.11)	(0.01)

Source: authors’ analysis of data from the 2013–2014 Rodriguez Survey. Note: standard errors are in parentheses. Each model controls for sex, age, education, ethnicity, English language ability, immigration status, pre-detention health insurance, detention facility, length of ICE detention, and length of total incarceration prior to detention. Models 1–3 use logistical regression and Model 4 uses Poisson regression. Full set of regression results with covariates in Tables 3, 4, 5 and 6 in the Appendix. ***p < 0.001, **p < 0.01, ^*p < 0.05, and ⁺p < 0.1