Predictive malaria model

Forecasting #VisualizeNoMalaria research with PATH and Tableau

About the report

Tableau and PATH

Tableau and PATH, a global health nonprofit, launched the #VisualizeNoMalaria campaign in support of eliminating malaria in Zambia by 2021.

Since 2015, a group of Tableau Zen Masters have volunteered their time and talents to help reach that ambitious goal. By working with the Zambian government to improve data accuracy and make critical, data-informed decisions about how and where to tackle outbreaks, our partnership helps build the skills of district and facility health teams to combat the disease at community level.

In just three years, Zambia’s Southern Province has already seen reported malaria cases drop 93%, and the number of malaria-related deaths drop 97%.

The Work Continues

In spite of early success, tracking down the remaining parasites remains an incredible challenge that calls for continued innovation. That's why PATH and Allan Walker, a Tableau Zen Master, developed this sophisticated predictive model that may be yet another critical intervention.

Looking at geography and topography to analyze the flow and pooling of predicted rains, and combining that data with health data on the current prevalence of the disease, could give Ministry officials a whole new view into the spread of malaria—and how to stop it.

Constant Innovation

Plenty of research, testing, and refinement will go into the process before the model is deployed, but with so many lives on the line, we must continue exploring opportunities to provide everyone in Zambia with the chance to live a malaria-free life.

World-changing innovations rarely follow a predictable process into existence. The predictive modeling detailed in this report, born in the towns and villages of southern Zambia and with the potential to change how we fight communicable diseases all over the world, may be one more example of that.


To build a model using Tableau that automates timely, robust forecasts of malaria cases.

In an attempt to help health facility works in Zambia plan and manage resources, the model (or "capability") will:

  • Accompany and support local knowledge and fieldwork.
  • Identify, process, and comprehend critical health decision-making criteria.
  • Enable users to access granular details about health facilities on an interactive map.
Watch Video

After a period of testing and improvement, the hope is that the models could show where the asymptomatic and symptomatic carriers could be in the near future. This information would allow officials and health workers at every level to get ahead of the disease and deploy resources accordingly.

Preliminary Findings

Early results are promising, with good model fitting of prediction to recorded malaria cases. Adding further variables—like subject domain expert input, population mobility data, and drug administration data—will improve the model.

It's also important to keep in mind the cadence of the model's workflow schedule and the valuable insight from archiving and comparing outputs, which can allow for tuning and reconfiguration.

Read the report for more details and to dive into raw output yourself.


Explore the Research

This viz addresses overall risk, monitoring low incidence areas, reporting from health facilities, and more from the #VisualizeNoMalaria project.

Case Study

Read the Report

Discover the methods, results, and next steps for the model. Dive into topographic wetness, multivariate vector arrays, catchment areas, and more.


Tableau and PATH

Tableau Foundation supports PATH's work to use the power of innovation to improve health and save lives around the world.

Harnessing the power of projection

In this study, many techniques were used to better understand the areas where malaria is present. One technique included using tiled raster files to shape impacted countries and their geographies.

The Consortium of International Agricultural Research provided NASA Shuttle Radar Topographic Mission-derived digital elevation models. Here, the elevation of Ethiopia is displayed, with the goal of better understanding slope and catchment areas for identifying possible breeding grounds for mosquitos.

Read the Full Report

Social impact and Tableau, powered by analytics

Get to know the Tableau Foundation

Learn more about PATH and other partnerships