Access to housing subsidies, housing status, drug use and HIV risk among low-income U.S. urban residents

Table 2 Results of logistic regression analysis of access to housing subsidies and supportive housing

	Access to housing subsidy (n = 346)^b				Access to supportive housing (n = 337)^c
Predictor ^a	Odds Ratio	95% Wald Confidence Limits	Wald Chi-Square ^d	Pr > ChiSq	Odds Ratio	95% Wald Confidence Limits	Wald Chi-Square ^d	Pr > ChiSq
Drug Use	0.94	0.55 - 1.60	0.06	.82	0.60	0.33 - 1.10	2.77	.096
Female	2.68	1.52 - 4.72	11.63	.0006	0.74	0.38 - 1.45	0.76	.38
Education	0.85	0.58 - 1.24	0.71	.40	1.16	0.75 - 1.78	0.44	.51
African-American	2.41	1.01 - 5.79	3.89	.049	1.98	0.71 - 5.55	1.69	.19
Latino	1.72	0.72 - 4.08	1.50	.22	1.60	0.56 - 4.56	0.78	.38
Monthly Income ($100s)	1.07	1.01 - 1.14	4.49	.034	1.03	0.96 - 1.11	0.64	.42
Told have HIV/AIDS by doctor	2.00	1.13 - 3.54	5.63	.018	5.15	2.75 - 9.66	26.13	<.0001
Mental Illness	1.34	0.79 - 2.27	1.21	.27	1.84	1.00 - 3.39	3.84	.049
Criminal Conviction	3.43	0.45 - 26.4	1.40	.24	0.21	0.01 - 3.58	1.16	.28
Time Convicted	0.76	0.44 - 1.31	0.99	.32	1.37	0.65 - 2.89	0.69	.41

^a Predictors: drug use, female, African-American, Latino, told have HIV/AIDS by doctor, mental illness and criminal conviction are coded Yes = 1 and No = 0; education is coded "Less than high school diploma" = 1, "High school diploma or GED" = 2 and "More than high school diploma" = 3; and time convicted is coded "No time served" = 0, "One to 6 days" = 1, "One week or more but less than a month" = 2, "One month or more but less than one year" = 3 and "One year or longer" = 4).
^b Access to housing subsidy (coded yes, n = 92; no, n = 254). Hosmer and Lemeshow Goodness-of-Fit Test Chi-Square = 4.37, df = 8. p = .82.
^c Access to supportive housing (coded yes, n = 67; no, n = 270). Hosmer and Lemeshow Goodness-of-Fit Test Chi-Square = 7.01, df = 8. p = .53.
^d PROC Logistic (SAS v9.2, SAS Institute)--Logistic regression--was used to fit logistic regression models to data.

ISSN: 1747-597X