Where care is far, and who can least afford the distance.
Sehatlas maps two things for every region of Saudi Arabia — how easily people can reach healthcare, and how vulnerable they are if they can't — and shows exactly where those two problems overlap.
Built on open data
All Phase 1 inputs are publicly available. No permissions required.
What sets this apart
Four decisions that shape how the platform is built.
Access, not counts
A facility count per region says nothing about how far anyone actually travels. The access score is a two-step floating catchment area (2SFCA) measure — how many facilities a population can reach within 30 or 60 minutes, discounted by how many other people compete for the same facilities.
A vulnerability index that doesn't exist yet
No published small-area socioeconomic vulnerability index exists for Saudi Arabia. This one combines census socioeconomic data, satellite environmental exposure, and chronic disease burden into a single, standardised measure.
Two axes, not one score
Access and vulnerability are kept deliberately separate. Merging them into one number would hide which problem a planner is looking at — distance, need, or both — and those require different interventions.
Open and reproducible by design
Every input is freely available data. Every method is published. Anyone — a ministry planner, an academic reviewer, another researcher — can check the work or reproduce it end to end.
By the numbers
What Phase 1 is built to work across, per the MOH Statistical Yearbook 2024 and the 2022 GASTAT census.
administrative regions
hospitals
primary healthcare centres
million people
independent axes
priority quadrants
The platform
Two tiers, deliberately kept separate — different audiences, different risk profiles.
From data to decision
The workflow a planner follows once both axes are computed.
Measure
Compute the access score and the vulnerability index independently for every region, validated against official facility statistics.
Classify
Plot both scores against each other. The quadrant a region falls into — not either score alone — tells a planner what kind of problem they're looking at.
Act
Prioritise the high-vulnerability, low-access quadrant first. In later phases, a siting optimiser recommends where a new facility would help most.
Where we are
This is a pre-launch student project. Here is exactly what exists today.
Built
- Methodology for both axes and the quadrant classification.
- The facility-validation approach comparing OpenStreetMap coverage against MOH statistics.
- This demonstration interface, running entirely on labelled, synthetic data.
Running
- Assembly of the open data inputs — WorldPop, OpenStreetMap, GASTAT, MOH — for the first end-to-end Phase 1 pipeline run.
Pending
- The first full Phase 1 pipeline run on real data.
- IRB exemption confirmation for secondary analysis of public data.
- A first manuscript draft for external review.
This platform is complementary to national health infrastructure such as the Sehhaty and NPHIES ecosystem. It does not replicate national data platforms; it provides an open, auditable analytical layer built on public data.