Skip to contents

DSIR

Data Science Infrastructure for Global Health

From Global Health Data to Clear Insights

DSIR is an open-source R package that simplifies access to global health data, supports consistent data processing, and provides reusable tools for analysis and visualization.

CRAN version Total CRAN downloads R CMD check status Lifecycle: maturing

Explore Features Get Started

Technical Overview GitHub

Open source · Available on CRAN · MIT licence

A consistent R workflow
  1. AccessWHO GHO · WHO GHE · UN SDG
  2. PrepareClean tables, country codes and data context
  3. AnalyseHealth indicators, mortality and progress
  4. PresentConsistent charts and analytical tables

Reusable tools for each stage of your analysis.

The everyday challenge

Why DSIR?

Health data are spread across platforms, with different table structures and geographical identifiers. Analysts repeatedly retrieve, clean and format data, while keeping their work reproducible and their reports consistent.

DSIR brings commonly used global health data sources and analytical utilities into a consistent R workflow, reducing repetitive data preparation and supporting reproducible analyses.

For technical teams preparing country comparisons, monitoring indicators and producing public health reports.

What you can do

Core capabilities

Access Global Health Data

Retrieve observations from WHO Global Health Observatory (GHO), WHO Global Health Estimates (GHE), and UN Sustainable Development Goals (SDG) services. Select countries, years and population breakdowns.

Explore WHO data access

Harmonize Data

Work with a consistent cleaned table structure, WHO Member State metadata and country identifiers. Keep additional dimensions and source information when needed.

See data preparation tools

Analyse Health Indicators

Use life tables, direct age standardization and measures of annual progress to support mortality analysis and indicator monitoring.

Explore analytical methods

Visualize and Present Results

Apply reusable ggplot2 themes and flextable formatting to create consistent charts and analytical tables. GitHub version 0.14.0 also provides checked indicator plots and experimental maps.

View visualization examples

A common table structure does not make indicator definitions or estimation methods comparable. Check source methods, population breakdowns and units before comparing results.

From tools to tasks

Practical applications

For WHO technical teams and public health researchers working with international health data.

  • Monitor health indicators

    Explore observations across countries and years, keeping population groups and data sources visible.

  • Explore mortality and disease burden

    Examine WHO estimates by cause, age, sex, country and year.

  • Assess progress towards SDG targets

    Retrieve health-related SDG series and analyse changes over time.

  • Prepare reproducible reports

    Reuse data preparation steps, charts and table styles across analytical outputs.

These are potential uses of the package, not claims of official adoption by WHO.

Ongoing development

Development roadmap

GitHub version 0.14.0 introduces experimental indicator maps with dsir_map(), using supplied sf polygons or explicitly downloaded WHO ADM0 boundaries. Data objects retain acquisition and cleaning metadata.

Maps require the optional sf package. WHO boundaries are downloaded on demand and cached locally; missing or invalid geometries are not repaired automatically.

The person behind the project

Developed and maintained by Shanlong Ding

Technical Officer (Data Analyst)
WHO Regional Office for the Western Pacific

View GitHub profile

DSIR is a personally maintained open-source project. It is not an official WHO software product and is not endorsed or published by WHO.

Start with the package

Getting started

CRAN provides version 0.13.0, published on 30 October 2026. Install version 0.14.0 from GitHub for the experimental data objects, metadata, plots and maps. 

# CRAN: 0.13.0
install.packages("DSIR")

# Or GitHub: 0.14.0
# install.packages("remotes")
remotes::install_github("shanlong-who/DSIR")

# Optional dependency for dsir_map()
# install.packages("sf")
library(DSIR)