The Visualize No Malaria (VNM) initiative, a partnership that began in 2015 between the Zambia Ministry of Health (MOH), PATH, and eight tech companies, has introduced a new set of data visualization tools to support Zambia's efforts to eliminate malaria.
The tools are a set of dashboards that help health managers within the MOH to view the malaria situation in real time and make decisions on how to address it based on data. While previous iterations of the dashboards were an improvement from paper-based processes, they could be be incomplete. To create more precise and user-friendly dashboards, the VNM team workshopped prototype visualizations and dashboards with end-users in Zambia, which led to meaningful changes in the content and organization of data.
Before and after. On the left, a DHIS2 dashboard shows a single month of data reporting rates, on the right, an updated interactive VNM dashboard showing 4 data quality attributes by facility and over time, plus aggregated by district.
Making data more accessible
During the development process, users gravitated toward dashboards that help strengthen data quality and decrease the effort it takes to understand data reporting issues. This co-development approach provided a model for creating and testing user-friendly dashboards that make health data more accessible to users at all levels of the health system hierarchy.
The primary data source for the dashboards is malaria surveillance data reported via mobile phones in DHIS2, a widely-used open-source health management information system, from nearly 500 health facilities and more than 3,000 community health workers. Other data sources include timing and location of malaria-specific interventions like mass drug administration (MDA), indoor residual spraying (IRS) data from mSpray, bender distribution, and household level reactive case detection information.
Data sources include DHIS2 (covering ~500 facilities & 3,000+ CHWs), geospatial data, and timing and location of interventions including MDA, IRS data from mSpray, bednet distribution, and household-level reactive case detection.
Engaging data collectors
But data is only reliable if it's complete. To ensure accuracy and prompt submissions of missing reports, weekly email and text alerts are pushed through the VNM platform to district and health facility staff. Since 2017, more than 12,000 text messages to 239 different facilities have been sent, leading to a 30 percent decrease in missing reports in some locations. With increased confidence in the quality of data, the dashboards have enabled leaders to make informed decisions about where to target interventions and deploy resources, including where to implement IRS and MDA campaigns, and which facilities to include in data quality audits.
Frontline health workers collect data that is fed into the national health system and turned into easy-to-use data visualizations that allow public health officials to view data in real time and make informed decisions about where to deploy resources.
Expanding the dashboards
Via in-person and virtual trainings the VNM initiative aims to cultivate a sustainable community of end-users who are empowered to collect, combine, and analyze their data and to generate meaningful insights from it.
Based on successes since implementation in 2015, the Zambia MOH plans to scale up Visualize No Malaria nationally. Similar dashboards have also been introduced in Senegal to fit country-specific needs, and initial steps have been taken to develop cross-border malaria surveillance dashboards .
