An FSI-AIBR Disease Intelligence Platform

Global Malaria Watch

Supporting evidence-informed malaria surveillance, epidemiological analysis and operational decision-making across malaria-endemic regions.

The platform consolidates published malaria burden, intervention coverage, diagnostic, treatment and financing indicators for 46 sub-Saharan African countries, 20002024, into a single auditable time series — with every aggregation rule documented and every figure traceable to a source country-year observation.

Current position · 2024

Estimated cases, 2024

46 countries in scope

270.7M
Estimated deaths, 2024

WHO modelled estimate

594k
Incidence per 1 000 at risk

Re-derived, not averaged

210.5
Indicator series

2000–2024, country-year

15

Release

Research Preview v0.1

Last updated

2026-08-04

Coverage

46 countries · 2000–2024

Provenance

WHO GHO · World Bank WDI

Why this exists

Malaria data is abundant. Integrated malaria intelligence is not.

Sub-Saharan Africa carries approximately 95% of the global malaria burden. The analytical constraint is rarely the absence of data — it is fragmentation, inconsistent aggregation and the distance between a published estimate and an operational decision.

  1. 01

    Data exists, but not in one place

    Burden estimates sit in WHO Global Health Observatory releases, denominators in World Bank series, coverage in household surveys, environment in satellite archives. Assembling a single comparable view is a multi-week analytical exercise repeated independently by every ministry, donor and research group.

  2. 02

    Aggregation is where most dashboards fail

    Rates per 1 000 or per 100 000 are frequently averaged across countries rather than re-derived from summed numerators and populations. Coverage percentages are combined unweighted. The resulting regional figures are not wrong by a rounding error — they are structurally misleading.

  3. 03

    Missing data is silently manufactured

    Country-years without an observation are commonly zero-filled or interpolated, which understates burden and inflates apparent progress. This platform preserves nulls and displays the number of reporting countries behind every aggregate.

  4. 04

    Surveillance is separated from decisions

    Epidemiological outputs are rarely framed for the operational question in front of a decision-maker. Global Malaria Watch presents the same audited dataset through executive, programme and clinical lenses, and is being extended toward forecasting and scenario simulation.

Intended users

One audited dataset, four decision lenses

The same underlying observations are framed for the question in front of each user, rather than reformatted into a generic analytics view.

UserPrimary analytical useModule
Ministries of health & state executivesNational and sub-regional burden position, peer comparison, concentration of cases and briefing-ready figures for cabinet and parliamentary reporting.Executive lens
Programme managers & implementing NGOsIntervention coverage read against burden trajectory — ITN access and use, IPTp3, IRS protection and domestic financing on a common baseline.Programme lens
Clinical & facility leadershipDiagnostic mix, test-confirmation rate and ACT treatment ratios to interpret case-management performance in epidemiological context.Clinical lens
Researchers, academics & analystsDocumented indicator definitions, transparent aggregation rules, preserved missing values and full CSV export from every table.All modules

Product positioning

The dashboard is the first module of a disease intelligence ecosystem.

This release establishes the verified historical surveillance layer. Subsequent phases extend the same architecture to environmental and climate covariates, explainable machine-learning risk forecasting, and a digital twin for scenario simulation and intervention planning. Nothing is displayed on the platform before its data source is licensed, documented and reproducible.

Review the four-phase roadmap
  1. Phase 1

    Historical surveillance

    WHO burden, coverage, diagnostic, treatment and financing series with World Bank denominators.

    Current release
  2. Phase 2

    Climate intelligence

    Rainfall, temperature, vegetation and environmental monitoring as transmission covariates.

    Planned
  3. Phase 3

    Predictive intelligence

    Machine learning and explainable AI for validated short-horizon risk forecasting.

    Planned
  4. Phase 4

    Digital twin

    Scenario simulation, decision support and intervention planning at national scale.

    Planned

Begin with the evidence

Scrub 25 years of burden and coverage, isolate any country for an intelligence briefing, and export every table as CSV for independent analysis.

Open surveillance dashboard