Health Fellow Spotlight: Neale Batra, Applied Epi

Within public health, field epidemiologists use analytical software to help control epidemics, but in a recent global survey 91% reported being constrained by the cost or capabilities of their current software. A resounding 96% said that training in the free and open-source software “R” should be a high priority, but in most countries there is insufficient R training capacity for public health. Transitioning public health analytics to R would dramatically improve epidemic response by making analyses more efficient and reproducible, enabling local problem-solving, and democratizing innovation. My big idea to support this critical transition is to create an ecosystem of public health-specific training resources for R and distribute them through a global training campaign led by frontline public health practitioners.

R is a free and open-source (community-built) analytical software that can be used to handle data, conduct statistical analyses, create visual graphics and dashboards, and produce informative, semi-automated reports. Since R is free and offers more advanced capabilities than comparable commercial software, its adoption is a natural choice. The primary barriers are the lack of relevant learning materials and that for many public health practitioners, R is their first exposure to coding. Yet these challenges are also R’s strengths - R’s organic, community-driven toolkit rapidly evolves to address new challenges like COVID-19, and because analysis is written in code it can be easily repeated, corrected, and shared with others. Widespread adoption of R would improve collaboration between agencies in emergencies, boost reproducible science, and result in more competitively-skilled epidemiologists in the Global South. 

At the beginning of the COVID-19 pandemic, my colleagues and I found ourselves fielding hundreds of requests from public health responders who wanted to use R but lacked training materials relevant to their daily outbreak control challenges. I assembled a team of 50 practitioners from around the world to quickly address this problem. We worked in our off-hours to create the “Epi R Handbook” - a free and comprehensive R reference manual, which in just one year has been used by over 180,000 people in nearly every country. Together, we wrote over 50 chapters (1200 pages) covering everything from basic data management, visualizations and report generation, to advanced statistics and version control. Organizations such as the US CDC, Doctors without Borders, and WHO have adopted our Handbook, and 70 volunteer translators are making it accessible in 11 languages. Since its release, we have formalized into the nonprofit organization Applied Epi and now offer many other services, such as live R training courses, online tutorials, case study repository, a community discussion forum, and an R Help Desk service for public health.

In a recent emergency, I worked alongside a Haitian epidemiologist who had used our resources to transition his outbreak analytics to R. His routine of data management and report generation to guide interventions was reduced from 30 hours per week to just 30 minutes, with fewer errors – giving him time to focus on big-picture improvements to the health system. His experience is not uncommon.

Another conversation that left an impression with me was with a leader of a large health research entity in eastern Africa. He said that our push to make data analytics more accessible will put power in the hands of local researchers and practitioners in the Global South.

Containment of epidemics before they become pandemics rests upon the local capacity to detect outbreaks early and to respond with agility and coordination. Epidemics can start anywhere in the world, and our organization, Applied Epi, focuses on strengthening this first line of epidemiological defense. These frontline practitioners often work under different priorities and pressures than epidemiologists based in academia, and their needs have traditionally been deprioritized.

Equipping this group with R improves the quality of epidemic intelligence, enables unprecedented analytic versatility, and facilitates collaboration between agencies because R is free. As an open-source software, R also upends the historical “top-down” dynamic because local practitioners can drive the creation of new R tools. 

To be prepared for infectious disease threats, the world needs more capable, connected, and supported local epidemiologists. While our first objective is to establish R as public health’s free standard analytics tool, this will lead to strengthening of the scientific methods underlying outbreak response. To foster mentorship and resource sharing we are building community spaces for our user-base of 180,000 people. To provide timely and direct assistance we are piloting an applied epidemiology Help Desk. Ultimately, our vision is to increase access to practical, high-quality training in field epidemiology more broadly. One of our longer-term projects is building an Applied Epi diploma certification that will be accessible worldwide - producing practitioners that have technical skills, rigorous methodological training, and practical experience.