TABLE 3.

Results of risk factor analyses for CHIKV infection and symptomatic status in different participant subgroups using modified Poisson regression^a

Model	Outcome	Study population (no. of individuals with outcome/total no. of individuals)	Variable selected	aRR	95% CI
1	Infection after or slightly before study initiation	157 contacts at risk of infection after or slightly before study initiation (74/157)	High frequency of weekly trash collection	0.38	0.19, 0.76
			Externally piped water	0.52	0.32, 0.84
			Ownership of vehicle	1.58	1.10, 2.27
			Having at least 1 tire in the yard	1.35	0.73, 2.49
			At or below primary education	1.02	0.71, 1.47
			High frequency of yearly abatement useb	0.97	0.58, 1.64
2	Incident infection	104 contacts at risk of an infection after study initiation (21/104)	Ownership of a vehicle	2.10	0.93, 4.76
			Externally piped water	0.40	0.14, 1.10
			At or below primary education	0.58	0.27, 1.21
			Having at least 1 tire in the yard	0.61	0.16, 2.38
3	Symptomatic status	63 contacts with active and incident infections (32/63)	—	—	—
4	Symptomatic status	21 contacts with incident infections (5/21)	—	—	—

^aVariables for inclusion in the model were chosen on the basis of a best-subset variable selection procedure. See the text for the candidate set of predictors. Dashes indicate that the variable selection procedure did not select any variables; hence, no predictors were found for models 3 and 4. aRR, multivariable-adjusted risk ratio from a modified Poisson model; CI, confidence interval.

^bA high frequency of an event was defined as an event that occurred ≥4 times.