Table 1. Variables used to build the modeling framework to estimate EBSMR.
| Category | Variable | Data source | Data collection |
|---|---|---|---|
| Vegetation | NDVI | Terra-MODIS 8-day composite data 2010–2013 | Extracted mean value from 1, 2, 3, 4, and 5 km surrounding circular buffer from each populated village |
| Water | NDWI | Ditto | Ditto |
| LSWI | Ditto | Ditto | |
| Geography | TWI | Digital elevation model at 30 m resolution from ASTER GDEM database [46] | Ditto |
| Temperature | P. falciparum temperature suitability index (PfTSI) | Malaria Atlas Project database [49] | Averaged to mean value for each HOD |
| Population | Population density (/km2) | Cambodia Malaria Bulletin report 2010–2013 [37–38] | Population record divided by total areas of each HOD |
| Vector control | Sufficient ownership of LLINa | Cambodia Malaria Survey 2010 [41] | Used the values reported at each provincial level |
| Treatment | Treatment failure rate by artemisinin combination therapyb | National Center for Parasitology, Entomology and Malaria Control [52] | Ditto |
aProportion of household in which distributed mosquito nets cover no more than two persons per net.
bTest positive for P. falciparum on day 28 or day 42.
EBSMR, Standardized morbidity ratio estimated by empirical Bayese method; NDVI, Normalized difference vegetation index; NDWI, Normalized difference water index; LSWI, Land surface water index; LLIN, Long-lasting insecticide-treated net; TWI, Topographical wetness index; HOD, Health operational district.