Can we use routine data for decision-making? A time trend comparison between survey and routine data at regional and national level in Mali

*Denotes equal authorship.

All analysis code for Can we use routine data for decision-making? A time trend comparison between survey and routine data at regional and national level in Mali have been made available for replication or extension.

GetData.R saves the raw DHS and HMIS data, GetTables.R calculates survey-weighted indicators and produces the tables in the paper, and GetFigures.R reproduces figures from the paper. Users can run GetFigures.R without accessing the raw data.

We cannot directly provide the HMIS routine data which was provided to us by the Ministry of Health, nor can we provide the DHS household data, which require access permission. However, we encourage all who are interested in obtaining DHS data to register for accounts with dhsprogram.com.

We used these DHS files in this analysis:

MLIR41FL.DTA

MLIR52FL.DTA

MLIR6HFL.DTA

MLBR41FL.DTA

MLBR52FL.DTA

MLBR6HFL.DTA

GetData: loads raw data and saves as .RData file (runtime = 7mins)

GetTables: loads saved .RData and produces results for tables 3-6 (runtime = 4mins)

GetFigures: loads output from GetTables and makes figures 1-4 (runtime = 2mins)

We recommend maintaining the naming convention of the DHS, and routine data, files. If you put these files in the “Data” subfolder, and access the .R scripts by first opening the MaliTrends.Rproj file, you won’t have to edit data file names or directories.

Scripts were written in R 3.5.1. R packages required are listed at the top of each script. Users can install the most recent versions of these packages by running this code:

install.packages(c(' readstata13', 'haven', 'readr', 'survey', 'dplyr', 'data.table', 'ggplot2', 'tidyverse', 'gridExtra', 'grid', 'cowplot','stringr'), dependencies = T)