Table 2.

Characteristics of Baltimore NEP clients at enrollment associated with multi-visit usage (N=12,388)

	Single Visit		Multiple visit
Characteristics	N	%	N	%	OR of Multi- Visit (95% CI)	Adjusted Odds Ratio of Multi- Visit (95% CI)**
Total	4458		7930
Predisposing
Female	2489	31%	5441	31%	1.03 (0.95–1.11)
< Age 30	774	17%	949	12%	0.65 (0.59–0.72)*	0.69 (0.61–0.77)*
Homeless	176	4%	261	3%	0.83 (0.68–1.01)
High school degree/GED	2634	59%	4849	62%	1.10 (1.02–1.18)*	1.10 (1.02–1.18)*
Race
White	1001	23%	1792	23%	Ref	Ref
Black/African-American	3221	72%	5813	73%	1.01 (0.92–1.10)	0.80 (0.72–0.89)*
Other	232	5%	311	4%	0.75 (0.62–0.90)*	0.65 (0.53–0.79)*
Marital Status
Single	3066	69%	5166	66%	Ref	Ref
Widowed/Separated/Divorced	947	21%	2010	26%	1.26 (1.15–1.38)*	1.15 (1.04–1.26)*
Married	416	9%	693	9%	0.99 (0.87–1.13)	0.92 (0.80–1.05)*
Enabling
Same Zip Code as Enrollment Site	1351	31%	3113	40%	1.50 (1.38–1.62)*	1.49 (1.38–1.62)*
Employed	640	14%	1089	14%	0.96 (0.86–1.06)
Need
Drug treatment (ever)	1418	57%	2675	61%	1.06 (0.96–1.18)
>20 years of injection drug use	1565	35%	3243	42%	1.28 (1.19–1.38)*	1.19 (1.10–1.30)*
>4 injections per day	1144	46%	2231	51%	1.08 (0.97–1.20)

Numbers do not sum to total population due to missing data for some variables. Multivariate models contained year of enrollment as a categorical variable to control for the association of external variables with the log odds of multi-visit usage.

^*significant at p < .05

^**GEE was used to account for possible clustering effects by site; the exchangeable correlation (rho) for this model was .02.