Table 1. Comparison of bivariate non-spatial models for S. japonicum infections based on either ELISA or Kato-Katz results with demographic, reservoir and environmental covariates that were significant at a p-value<0.15.
Covariate group | Covariate | Covariate category | ELISA | Kato-Katz | ||||
ORa | CIb | AICc | ORa | CIb | AICc | |||
Demography | ||||||||
Age group (years) | ||||||||
0–10 | 1 | 1 | ||||||
11–20 | 2.06 | 1.68, 2.52 | 1.56 | 0.83, 2.95 | ||||
21–30 | 3.59 | 2.91, 4.43 | 4.41 | 2.36, 8.25 | ||||
31–40 | 4.42 | 3.65, 5.35 | 6.01 | 3.35, 10.76 | ||||
41–50 | 5.27 | 4.34, 6.38 | 8.32 | 4.65, 14.89 | ||||
>50 | 5.56 | 4.60, 6.73 | 33584 | 8.52 | 4.78, 15.19 | 8442 | ||
Sex | ||||||||
Female | 1 | 1 | ||||||
Male | 1.99 | 1.88, 2.11 | 33914 | 3.12 | 2.66, 3.65 | 8474 | ||
Occupation | ||||||||
Herdsman, farmer, fisherman or boatman | 1 | 1 | ||||||
Preschool or student | 0.33 | 0.30, 0.36 | 0.17 | 0.13, 0.23 | ||||
Civil servant or businessman | 0.88 | 0.75, 1.04 | 0.22 | 0.11, 0.47 | ||||
Other | 0.52 | 0.39, 0.71 | 33690 | 0.56 | 0.27, 1.13 | 8456 | ||
Reservoir | ||||||||
Presence of infected buffaloes in the village | 1.76 | 1.66, 1.86 | 34104 | 2.96 | 2.60, 3.40 | 8458 | ||
Prevalence of infected buffaloes in the village | 1.25 | 1.22, 1.28 | 34136 | 1.35 | 1.29, 1.41 | 8571 | ||
Environment | ||||||||
Mean NDVI | 1.20 | 1.16, 1.24 | 34337 | 0.98 | 0.92, 1.05 | 8706 | ||
Mean NDVI during spring, summer and fall | 1.31 | 1.26, 1.35 | 34204 | 1.16 | 1.07, 1.25 | 8691 | ||
Mean NDVI during winter | 1.08 | 1.05, 1.11 | 34439 | 0.86 | 0.81, 0.91 | 8682 | ||
Distance to Dongting Lake (km) | 1.08 | 1.05, 1.11 | 34438 | 1.05 | 0.99, 1.13 | 8703 | ||
Endemic types | ||||||||
Lake embankment | 1 | 1 | ||||||
Inside embankment | 0.69 | 0.61, 0.78 | 0.24 | 0.17, 0.33 | ||||
Lake fork | 1.02 | 0.94, 1.12 | 0.72 | 0.60, 0.85 | ||||
Grassy lake beach/marshland | 0.78 | 0.70, 0.86 | 0.16 | 0.12, 0.21 | ||||
Hills/hilly area | 0.52 | 0.45, 0.61 | 34292 | 0.02 | 0.01, 0.07 | 8324 |
OR: Odds ratio.
CI: Confidence interval.
Akaike information criterion. The smaller the AIC the better the model performance.