Skip to content

Determining the most important predictors of diarrhoea in children under five in South and Southeast Asia by exploring the spatiotemporal association between diarrhoeal incidence and various behavioural, socio-demographic, and environmental factors.

License

Notifications You must be signed in to change notification settings

cjabradshaw/childDiarr

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

70 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Predictors of diarrhoea in children under five years old

DOI

Determine the most important predictors of diarrhoea in children under five in South and Southeast Asia (Pakistan, India, Nepal, Bangladesh, Myanmar, Cambodia, Philippines, Timor-Leste) by exploring the spatiotemporal association between diarrhoeal incidence and various behavioural, socio-demographic, and environmental factors.

Dr Syeda Hira Fatima
Global Ecology | Partuyarta Ngadluku Wardli Kuu, Flinders University, Adelaide, Australia
September 2024
e-mail

and

Prof Corey J. A. Bradshaw
Global Ecology | Partuyarta Ngadluku Wardli Kuu, Flinders University, Adelaide, Australia
September 2024
e-mail

Project collaborators: Dr Melinda Judge, Prof Peter Le Souëf, Dr Lewis Weeda, Naomi Henry

Focal manuscript

Fatima, SH, MA Judge, PN Le Souëf, CJA Bradshaw. Impact of climate change on diarrhoea risk in low- and middle-income countries. In review.

  • DHSDiarrProcessing_1.R: R code to load and merge DHS survey and GPS data and select the appropriate list of variables.
  • DHSDiarrProcessing_2.R: R code to preprocess, recode, and create new variables where necessary.
  • DHSDiarrProcessing_3.R: R code for imputation of variables with missing data.
  • DHSDiarrProcessing_4.R: R code for processing of raster data.
  • DHSDiarrProcessing_5.R: R code for processing of variables at the cluster level and standardization.
  • DHSDiarrAnalysis.R: R code to reproduce the resampled boosted regression tree analysis for determining the relationships between probability of diarrhoea, and socio-economic, maternal, child, climate data (full dataset).
  • DHSDiarrAnalysisIndaOnly.R: R code to reproduce the resampled boosted regression tree analysis for determining the relationships between probability of diarrhoea, and socio-economic, maternal, child, climate data (India only).

Data

  • DHSclusterLevelDiarrData.csv.zst: Demographic and Health Surveys data summarised by cluster with central parameter (mean, proportion, etc.) and variance per cluster. Overlaid (cluster-level) climate data derived from WorldClim bioclimatic variables (mean annual temperature, temperature annual range, total annual precipitation, precipitation seasonality, and precipitation of the driest quarter). Unzip .csv data file prior to analysis. The file is a high-compression .zst of the .csv base file; use the following command in Terminal to decompress: zstd -d 'DHSclusterLevelDiarrData.csv.zst'. Due to licencing constraints, we are not permitted to post the summarised (cluster-level) data here.

Required R libraries

  • cowplot, dismo, dplyr, foreign, gbm, GGally, ggplot2, ggpubr, gridExtra, haven, leaflet, mice, raster, reshape2, sf, sp, spatstat.random, tidyr, truncnorm, usdm


Flinders University   Global Ecology Lab     UWA     The Kids Research Institute     Future Child Health