Determinants of Aedes mosquito density as an indicator of arbovirus transmission risk in three sites affected by co-circulation of globally spreading arboviruses in Colombia, Ecuador and Argentina

Table 5 Mixed-effects regression IRR values, the two boundaries of the 95% CI (IRR 2.5% and 97.5% CI) and the P value, for each predictor included in the final model of household female Aedes mosquito density for the dataset from Manta, Ecuador

Predictor^a	Level^a	IRR	IRR 2.5% CI	IRR 97.5% CI	P
Humidity		1.39	1.10	1.76	0.005*
Arbovirus knowledge		1.11	0.89	1.39	0.364
Presence of container in the yard	No	Reference
Presence of container in the yard	Yes	0.69	0.50	0.96	0.025*
Presence of water tank	No	Reference
Presence of water tank	Yes	1.04	0.73	1.47	0.839
Problem obtaining water	No	Reference
Problem obtaining water	Yes	2.07	1.41	3.05	< 0.001*
Store water to wash	No	Reference
Store water to wash	Yes	1.44	0.91	2.26	0.115
Store water to cook	No	Reference
Store water to cook	Yes	0.95	0.68	1.31	0.733
Insecticide use	No	Reference
Insecticide use	Yes	2.05	1.47	2.85	< 0.001*
Washing containers	No	Reference
Washing containers	Yes	1.11	0.53	2.29	0.782
Killing insects	No	Reference
Killing insects	Yes	0.79	0.45	1.35	0.386
Frequency of garbage collection	Every other day	Reference
	Daily	1.59	0.98	2.62	0.064
	Unpredictable	0.95	0.40	2.09	0.894
	Weekly	0.36	0.12	0.92	0.041*

Estimation of model coefficients was conducted on a total of 176 households (53% of the full dataset). Average proportion of variance explained by random-effects factors, i.e. month and neighborhood, was 0.26 and 0.02, respectively
^aPredictors that are underlined have an IRR that is significantly different from 1
*Significant at a threshold α = 0.05

ISSN: 1756-3305