Skip to main content
. 2012 Mar 5;7(3):e32625. doi: 10.1371/journal.pone.0032625

Table 2. Association of parasitaemia risk with environmental/climatic factors, socio-economic status and malaria interventions resulting from raw data summaries and geostatistical Zero-Inflated Binomial models.

Variable Raw Data Geostatistical model Ia Geostatistical model IIb
Prevalence OR 95% BCIc OR 95% BCIc
Night LST 1.16 (0.66, 1.86) 0.83 (0.53, 1.26)
NDVI 1.48 (0.88, 2.48) 0.91 (0.61, 1.83)
Area type
Rural 8.47% 1 1
Urban 1.30% 0.19 (0.07,0.45) 0.43 (0.16, 1.06)
Wealth Index d
Most poor 13.75% 1
Very poor 6.51% 0.77 (0.57, 1.03)
Poor 1.51% 0.22 (0.08, 0.51)
Less poor 0.96% 0.12 (0.05, 0.41)
Least poor 0.65% 0.09 (0.01, 0.26)
Age
0–1 3% 1
1–2 4.54% 1.20 (0.70, 2.43)
2–3 8.07% 2.93 (1.62, 5.33)
3–4 7.95% 2.96 (1.66,5.74)
4–5 8.11% 2.77 (1.44, 5.21)
ITNs e
<1 6.84% 1
Inline graphic1 1.41% 0.14 (0.03, 0.7)
a

Model I includes only environmental/climatic factors.

b

Model II includes ITN coverage, children's age and wealth index.

c

Bayesian Credible intervals.

d

Household wealth index.

e

Number of available ITNs per every two household members.

f

The range parameter (degrees), defined as Inline graphic indicates the distance above which the spatial correlation becomes negligible.