Table 5. Model parameters for the best model using weather, land cover, mosquito infection, and demographic factors to predict the occurrence of WNV human cases in Chicago region.

Variable	Parameter estimate	F-value	P-Value	Odds ratio (95% CI)
Fixed effects
Year	-	17.33	<0.001	-
Temperature of two weeks before	0.06963	22.36	<0.001	1.08 (1.049–1.112)
Temperature of three weeks before	0.1085	42.99	<0.001	1.128 (1.092–1.165)
Temperature of four weeks before	0.1628	116.47	<0.001	1.197 (1.162–1.234)
Average January temperature	0.3613	16.65	<0.001
Mosquito infection rate of one week before	0.003199	21.53	<0.001	1.003 (1.002–1.004)
Mosquito infection rate of two weeks before	0.003938	38.79	<0.001	1.004 (1.002–1.005)
Mosquito infection rate of three weeks before	0.004003	37.83	<0.001	1.004 (1.002–1.005)
Mosquito infection rate of four weeks before	0.003958	34.63	<0.001	1.004 (1.002–1.005)
Total population	0.000225	Infinity	<0.001	1.009 (1.006–1.012)
Open water percentage	-0.05527	9.58	0.002	0.954 (0.921–0.988)
Developed light intensity percentage	0.01848	80.65	<0.001	0.990 (0.985–0.994)
Deciduous forest percentage	-0.02401	4.66	0.0309	0.985 (0.980–0.991)
Grassland percentage	-0.04603	3.14	0.0763
Post 1990 built housing percentage	-0.00546	4.28	0.0386
Random effect
Subject	Estimate	Standard error	Z-value	P-value
Hexagon ID	1.1769	0.1636	7.19	<0.0001