floods-bangladesh

Excess risk in infant mortality among populations living in flood prone areas in Bangladesh

Description

This repository includes R code to run all of the analysis for the paper:

Rerolle, F., Arnold, B. F., Benmarhnia, T. Excess risk in infant mortality among populations living in flood prone areas in Bangladesh: A cluster-matched cohort study over three decades, 1988 to 2017. PNAS 2023 120(50)e2218789120. https://doi.org/10.1073/pnas.2218789120

Should you have any questions about the files in this repository, please contact Francois Rerolle at UCSF (francois.rerolle@ucsf.edu) or see the corresponding author contact information in the manuscript.

Additional Resources

Open Science Framework

All data used in the analysis is publicly available but access must be granted by the DHS program for the 6 population surveys conducted in Bangladesh (2017, 2014, 2011, 2007, 2004, 1999). When requesting data, include GPS coordinates and spatial covariates data. The other data are deposited on OSF: https://osf.io/vrfmz/

System Requirements

All analyses were run using R software version 4.3.0 on MacOS Monterey using the RStudio IDE (https://www.rstudio.com).

sessionInfo()

R version 4.3.0 (2023-04-21)

Platform: aarch64-apple-darwin20 (64-bit)

Running under: macOS Monterey 12.6

Installation Guide

You can download and install R from CRAN: https://cran.r-project.org

You can download and install RStudio from their website: https://www.rstudio.com

All R packages required to run the analyses are sourced in the file 0-config.R.

The installation time should be < 10 minutes total on a typical desktop computer.

Instructions for Use

To reproduce all analyses in the paper, we recommend that you:

1. clone the GitHub repository

2. Create a data directory with 3 subdirectories: untouched, temp and final.

3. In the data/untouched directory copy and paste repository from OSF: https://osf.io/vrfmz/

4. In the data/untouched/dhs directory paste downloaded DHS data. You should have 6 subdirectories corresponding to each DHS survey (note these subdirectories include a download date, 05302023, which was used in this study: BD_1999-00_DHS_05302023_1740_172978, BD_2004_DHS_05302023_1740_172978, BD_2007_DHS_05302023_1740_172978, BD_2011_DHS_05302023_1739_172978, BD_2014_DHS_05302023_1739_172978 and BD_2017-18_DHS_05302023_1738_172978

5. Create child-mortality-dhs\output directory with 2 subdirectories: figures and matching

6. All of the data management and analysis scripts should run smoothly.

Running the all analyses on the above Mac configuration required ~1h.

Note that the only script that takes very long is 6b-infant-mortality-model-individiual-level.R which pertains to the sensitivity analysis estimating effects using conditional logistic regression.

We identified two additional wrinkles in our internal replication effort that users should keep in mind. In the data management workflow the script 10-extract-precipitation-data.R can be very slow to download the CHIRPS data. Additionally, the script 11-leave-one-out-cross-validation.R requires manual commenting and un-commenting blocks of code to repeat the LOO analysis, leaving out each event in turn. Unfortunately, our team simply didn't have time to code this in a more elegant way, but there are clear instructions in the script for doing that piece of the analysis.