Table 4.

Spatially-lagged Multivariate Hurdle Models Examining Correlates of High-cost SLHs

	OR	[95% CI]	p
Logistic regression model examining correlates of any high-cost SLHs (N=2332)
Population density	0.99	[0.61– 1.59]
% Hispanic	0.55	[0.18– 1.66]
% w/o high school diploma	0.75	[0.22– 2.55]
% Below poverty	1.24	[0.74– 2.09]
Median home value	2.44	[1.59– 3.73]	***
Number of bars	1.21	[0.97– 1.51]
Number of off-premises alcohol outlets	1.24	[0.92– 1.67]
Number of treatment facilities	1.27	[1.06– 1.54]	*
Number of AA meetings	0.99	[0.80– 1.24]
Rho (spatially-lagged high-cost SLH indicator variable)	80.66	[11.89–547.10]	***
Constant	0.00	[0.00– 0.01]	***
	IRR	[95% CI]	p
Zero-truncated Poisson model examining correlates of counts of high-cost SLHs among neighborhoods with high-cost SLHs (N=49)
Population density	0.30	[0.10– 0.87]	*
Rho (spatially-lagged high-cost SLH density)	2.73	[1.25– 5.94]	*
Constant	0.00	[0.00– 0.00]	***
Exposure (log of total housing units)	1.00

^*p<0.05;

^**p<0.01;

^***p<0.001