IBM embarks on multi-year research effort to better understand actions needed to improve human health

IBM today launched a multi-year research effort to connect and analyze enormous collections of data from a wide variety of sources in order to enable individuals, governments and businesses to better understand which actions to take to improve human health. The project will initially focus on solving the issue of childhood obesity.

The IBM Research project will combine and analyze massive data sources that have never before been integrated to simulate the cause-and-effect relationships between agriculture, transportation, city planning, eating and exercise habits, socio-economic status, family life, and more. Predicting real-world reactions that influence human health, the project aims to provide fact-based recommendations of actions to take and ones to avoid.

"Our ability to advance the health of our population is currently limited to maintaining healthy life choices and working within a health care delivery system because it's been impossible to understand and to quantify precisely how each factor in our environment plays a role," said Martin Sepulveda, M.D., IBM Fellow & VP, Integrated Health Services, IBM. "We hope the results of this project will help individuals, governments and businesses actually understand exactly how the actions they take affect health – and then work together to make better decisions that make it easy to be healthy."

In the United States, chronic diseases such as diabetes, heart disease and obesity account for 70 percent of all deaths and more than $1.5 trillion of healthcare spending annually. Factors far beyond the traditional healthcare system – including finance, urban planning, individual behavior, disease transmission, clinical research, media and many others – influence human health. Understanding these interconnected factors is critical to developing effective programs that enhance health and well-being.

Today, only simple cause–and-effect connections can be made because of the way in which information is collected, stored and accessed. For example, the connection between obesity and processed foods or lack of exercise is widely acknowledged, but in many cases it is a guessing game as to whether a bigger impact would be made through incentivizing a new health food retailer to move into town or expanding the bus routes in an area with a high concentration of dual-income families. These are the complex interactions that make up a 360 degree view of health and the kind of advanced simulations IBM researchers are working to develop, with the goal that they can help predict which programs would be most successful before they are implemented.

"Managing health, be it for a single patient or an entire population, is an overwhelmingly complex challenge," said Gary An, Assistant Professor of Trauma and Critical Care at Northwestern University Feinberg School of Medicine. "Despite the critical influence of cultural, socio-economic and environmental factors, the doctor-patient relationship remains the mainstay of "delivering" healthcare: all these complex issues need to meld into a single thread of conversation as I talk to my patient. Therefore, any initiative – like the one IBM is launching – that can help bring together these disparate and often potentially contradictory forces and aid me in tailoring how I can help my patient improve his or her health, is both greatly needed, and greatly welcomed."

The IBM Research project could help pinpoint incentives governments and businesses might offer or what types of investments might be needed and how to prioritize them. For example, to foster an environment that helps people make healthier choices, the project could simulate different models to incent retailers to locate grocery stores near key transportation hubs so people without cars can get to them. In another scenario, it could provide more understanding around impact of food labels and how they might affect marketing practices, buying habits or school lunch quality.

"In many cases, the data and models exist. They just need to be put together in a consumable way that shows the wider connections and potential actions that can enhance individual and community health," said Paul Maglio, research scientist, IBM Research – Almaden. "This is a huge challenge from both a social and technological perspective, but we believe our expertise in service science, computational modeling, math and large-scale analytics can help answer these important questions."

IBM intends to partner with public policy and food experts, medical clinicians, economists, simulation experts, industry leaders, universities and others in this collaborative endeavor.

Last week IBM gathered many of the leading thinkers from these areas at the 10th annual Almaden Institute in San Jose, California to discuss the fundamental issues of this grand challenge and to make connections to establish ecosystem-wide, fact-based, and scientifically sound decision making about health.