Table 4. Multivariate logistic regression analysis of risk factors for scrub typhus infection, Beijing^*.

Item	Partial regressioncoefficient	Partial regression coefficientstandard error	P value	AdjustedOR	Adjusted OR95.0% CI
Gender	0.987	0.494	0.046	2.7	1.0∼7.1
Education≤ 6 years	2.773	1.104	0.012	16.0	1.8∼139.2
Working in vegetable fields	1.302	0.601	0.030	3.7	1.1∼11.9
Working in woodland and hilly area	2.101	0.919	0.022	8.2	1.4∼49.5
Harvesting in autumn	2.842	0.621	<0.001	17.2	5.1∼57.9
Sowing	−1.414	0.611	0.021	0.2	0.1∼0.8
House yard without cement floor	1.437	0.711	0.043	4.2	1.0∼17.0
Piling weeds in the house and yard	2.669	0.566	<0.001	14.4	4.8∼43.7
Living in houses near grassland, vegetable field or ditch	1.906	0.543	<0.001	6.7	2.3∼19.5
Raising animals	−0.938	0.550	0.088	0.4	0.1∼1.2
Constant	−45.743	2.188*104	0.998	<0.0	–

^*The significance level for entering the multivariate logistic regression model was set as 0.10 and for staying in it was set as 0.05.

^*Negatively correlated factors: Sowing.

^*Positive correlation factors: Gender, education≤ 6 years, working in vegetable fields, working in hilly area, harvesting in autumn, house yard without cement floor, piling weeds in the house and yard, living in houses near grassland, vegetable field or ditch.