Increasing access to care using big data

Artificial intelligence (AI) and data science have the potential to revolutionize global health. But what exactly is AI and what are the obstacles to a more widespread integration of big data in global health? Duke’s Global Health Institute (DGHI) organized a Think Global Webinar Wednesday 17 Februarye to delve into these questions and more.

The webinar panelists were Andy Tatem (Ph.D.), Joao Vissoci (Ph.D.) and Eric Laber (Ph.D.), moderated by Liz Turner (Ph.D.) Director of Design and analysis of DGHI research. Tatem is Professor of Spatial Demography and Epidemiology at the University of South Hampton and Director of WorldPop. Vissoci is Assistant Professor of Surgery and Global Health at Duke University. Wrasse is Professor of Statistical Sciences and Bioinformatics at Duke.

Panelist moderator, Lisa Turner

Tatem, Vissoci, and Laber all use data science to solve problems in global health. Much of Tatem’s work uses geospatial data sets to help inform global health decisions like the distribution of vaccines in a certain geographic area. Vissoci, who works with the GEMINI Duke’s lab (Global Emergency Medicine Innovation and Implementation Research) tries to harness secondary data from health systems to understand issues of access and delivery of care, as well as the delivery of care. Laber is interested in improving decision-making processes in healthcare spaces, trying to help healthcare professionals synthesize very complex data through AI.

All of their work is vital to modern biomedicine and healthcare, but, said Turner, “AI means a lot of different things to a lot of different people. Laber defined AI in healthcare simply as the use of data to improve healthcare. “From a data science perspective,” Vissoci said, “[it is] synthesize data… an automated way of giving us information. This returned information is trends and digestible understandings derived from very large and very complex data sets. Tatem said AI has already “revolutionized what we can do” and said it is “powerful if run the right way.”

A screenshot from

We are often drawn into a sci-fi version of AI, Laber said, but in reality it is not a dystopian future, but a set of tools that maximize what can be derived from the data.

However, as Tatem stated, “[AI] is not a magic scenario, press a ‘button where you get automatic results. Much of the work of researchers like Tatem, Vissoci, and Laber involves ‘aligning’ work with data producers, understanding data quality, integrating datasets, cleaning data, and other processes. “Backgrounds”.

This comes with many caveats.

“Bias is a huge problem,” Laber said. Vissoci reinforced this, stating that models built from AI and data science will represent the sources of data they can access, bias included. “We need to do better work to get better data,” Vissoci said.

In addition, there needs to be more direct listening and communication with the “end-users from the start” of projects, Tatem stressed. By stepping back and listening, tools created through AI and data science can be better received with real adoption and less skepticism or mistrust. Vissoci said that “direct engagement with people on the ground” turns data into meaningful information.

Better structures for winding privacy concerns also need to be developed. “A major overhaul is still needed,” Laber said. This includes things like better consent processes for patients to understand how their data is being used, although Tatem said it gets “very complex” when integrating data.

Nonetheless, the future looks bright and every panelist is confident that the benefits will outweigh the challenges ahead in bringing big data to global health. A cool example given by Vissoci of an ongoing project deals with the influence of environmental change through deforestation in the Brazilian Amazon on the impacts of indigenous populations. Through working with “heavy multidimensional data”, Vissoci and his team have also been able to optimize the barely distributed Covid vaccine resources “for use in areas where they can have the most impact”.

Laber envisions a world with small or no clinical trials if “randomization and experimentation” is integrated directly into health systems. Tatem noted that he has seen extreme growth in the field over the past 10 to 15 years, which appears to be only accelerating.

Much of this work is about making better decisions about resource allocation, as Turner said at the start of the panel. In an age of reassessment of equity and access, AI and data science could be used to bring both into the realm of global health.

Message from Cydney Livingston

Sean N. Ayres