Visualise an end to malaria
Born as an idea in Seattle during a coffee shop conversation, the initiative named ‘Visualize No Malaria’ has evolved into one of the world’s biggest partially statefunded health IT programmes running in Zambia, Africa. "Our goal was and is to actually eradicate malaria across the entire African country," says Neal Myrick, Director of Social Impact at Tableau and Head of the Tableau Foundation, looking back at the programme launch in 2015 when the disease had already been reduced from a raging fire to a burning ember. "The last step required us to test, track and treat the 80% of infected, but asymptomatic people by means of data and data visualisation," Myrick comments.
To achieve its targets, goal analytics and visual technologies have been set up to pipe real-time reporting of new cases into visualisations that allow ministry officials to deploy health resources to the right places at the right time. Neither PATH nor Tableau would have stemmed this huge task alone they believe. By involving Zambia’s Ministry of Health, a front-runner in malaria control, to a large extent and also resourcing help from the National Malaria Control Centre in Zambia, the initiative went beyond data collecting and sharing for analysis, and even installed a full feedback loop to flow back data results to Zambia’s frontline health workers effectively. "Over a period of two years we could enable users to receive performance management inputs, weekly updates on the number of reported cases in their area, text messages if their reports are complete and even insights where they might expect to find flaring up break outs of malaria," Jeff Bernson, Director at PATH for Malaria Monitoring and Evaluation, tells Insights.
Working with the Ministry, sourcing public health data
However, obstacles had to be overcome when starting off with the ambitious objective: "One big challenge was collaborative design, namely building visualisations with the Ministry instead of for them," Myrick says. Two dozen Tableau power users plus a team from PATH built prototypes of dashboards assumed to be most effective. The group even flew to Zambia for a series of workshops with the Ministry: "We tweaked everything from analytical frameworks over mapping strategies to dashboard layout," Myrick continues.
This eye for detail paid off, not least because of the data sourcing: "All of our sources come from public health data that the country has been managing and developing for over a decade. Especially the diligently run National Malaria Indicator Survey serves as main epidemiological source for data," Bernson says.
Dashboards for quality control and incidence surveillance
With Tableau Software including Tableau Desktop and Server vital success tools have entered the history of PATH data management. They allow anyone to connect to data, build visualisation and interactive exploration tools, as well as internal sharing of confidential health data. "Basically this is the same setup used by hospital systems around the world," Myrick explains, referring to the cutting-edge security features.
In order to map and visualise malaria in Zambia, the platform has been customised. Bernson, who is technical lead in the project, sums up the critical tools in service: "On our dashboard around the quality of reporting we can look at all facilities of a certain district and immediately see their performance in reporting, if reports are on time, complete or contain errors." Health facilities in Southern and Western provinces of Zambia report weekly to help the NGO get a clear picture.
Beyond this quality control socalled low incidence dashboards are in use to visualise the incidence level in a certain district as well as the total number of cases that occur. "They allow us to identify areas close to elimination, suggesting us to intervene in the particular regions in order to drive and maintain incidence at level zero," Bernson states. On the other side, if there are hotspots the NGO can immediately react by deploying diagnostics and planning resources at early stage.
Major use cases for this malaria surveillance on the ground are workers on a district health department level: "We have the District Health Information Officers being in charge of all data and diseases and the Malaria Focal Point, a person held accountable for timely and high-quality reporting from the facilities," Bernson sums up. The customised dashboards ease their jobs in terms of reporting malaria cases and spotting potential challenges. Apparently, responses have been inspiring: "Even facility workers now demand for tailored dashboards as they want to know their impact by submitting data," he says.
Broad-ranging IT partners
Another decisive moment for the initiative’s success has been a broad partnership with the tech sector. Along with Tableau, PATH collaborates with IT service providers such as Mapbox, Alteryx and Twilio. "They brought in the technical expertise and resources required to be successful," Bernson says. For Twilio, a cloud-based messaging company, this meant developing automated alerts based on district-based dashboards informing local health districts via text message to fill in missing cases. Teaming up with data processing specialist Alteryx and map provider Mapbox, PATH and Tableau managed to get predictive modelling about the likelihood of contagion.
Health data forecasting as second stream of work
Data forecasting is the other potential treasure buried inside visual data modelling. Although functionalities are built into the programme, the NGO and its partner are still in the process of field testing the predictability features. "In 2015 we started to combine our data sources with big data like projected weather data and detailed topographical models to project the likelihood for places to see an instance of malaria," Myrick says. In doing so, Zambia’s Ministry officials get the unique opportunity to be proactive in distributing medicines, bed nets, insecticides and improve drainage systems in these areas. Bernson goes on to advise that: "This exciting work is rolled out this year. By pressure tests over time we make sure these models really reflect reality."
Results have been great. Zambia’s Southern Province has seen a 93% reduction in reported cases, and Myrick goes on to say: "They recognised that visual analytics can even overcome the education limitations of community health workers and that professional managers at district level could flourish with better, more reliable data.
"If this is the straightforward prescription for eliminating some of the world’s deadliest diseases, we really have no excuse not to do it."
A chance for major health concerns?
Could what has proven to be effective in Zambia possibly be a solution for other public health scenarios too? Indeed. According to PATH and Tableau more and more Ministries of Health need to adopt these types of easy-to-use tools to catch contagious diseases long before they have a chance to become cross-border outbreaks or, worse, pandemics. Vietnam, for example, has been working with PATH on all kinds of infectious diseases from Dengue to Zika and Influenza. Here again Tableau Software was linked as an easy way to mash up a lot of data coming from many different sources.
All efforts made by PATH, Tableau and Zambia’s Ministry of Health demonstrate that keeping the information moving faster than a disease can be one of the key benefits of smart and visual data management for healthcare providers and organisations likewise. Myrick concludes: "It puts people in a position to make lifesaving decisions with it."
View Insights 5.4 eBook - Visualise an end to malaria