A Study of Spatial Pattern of Household-level Disease Transmission Variables Using DHS/MICS Data
Abstract
Household-level risk factors for infectious disease transmission have drawn increasing attention, which is promoted by the changes of climate and disease control situations in Low- and Middle-Income Countries (LMICs). In the past decades, studies have shown that not only the focuses on reducing contamination of the domestic environment through improving water storage and treatment, housing, and sanitation and hygiene conditions, but also the weather events, topography, ecology, and meteorological factors could be associated with the propagation of specific infectious diseases. In this study, the household cluster level disease transmission variables, including floors, water, sanitation, and open defecation were extracted from over 300 Demographic and Health Surveys (DHS) and Multiple Indicator Cluster Survey (MICS) from more than 100 LMICs, from 2005 to present and modeled with 21 geophysical or socioeconomic variables which are derived from high spatial resolution remote sensing or survey data sets. The Bayesian Multi-level based logistic regression was applied to develop the model to generate the predicted disease transmission variables on a quasi-global scale at 6 km spatial resolution. It is found that cropland areas, urbanization, and distance to water source show greater importance than other factors. In addition to this, the Markov Random Field (MRF) smooth algorithm was also utilized to improve the spatial correlation and continuity of the prediction maps.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2022
- Bibcode:
- 2022AGUFMGH45B0681F