Skip to Main Navigation Skip to Content Skip to Footer Navigation
Give Now

Make a Difference

Visualizing Risk

Regular screening is often promoted as a way to detect cancer before it becomes deadly. But cancer tests aren’t always accurate. For example, the number of women who receive positive mammogram results is far higher than the number who actually have breast cancer. Those inaccurate results can have troubling consequences.

“Patients are not well informed about the limitations of the tests they are given,” says Remco Chang, an assistant professor of computer science at Tufts. “Knowing the difference between a positive test result and true health risk is very important for patients as they decide what to do next.”

False positive test results take an emotional, physical, and financial toll, particularly when healthy people receive unnecessary treatment. To tackle this problem, Chang has joined forces with two Tufts colleagues: Paul Han, M.D., an assistant professor of medicine, and Holly Taylor, a professor of psychology. The interdisciplinary team hopes to develop a better way for physicians and patients to consider the difference between a positive test result and the patient’s true health risk.

“Our team is uniquely strong,” says Chang, who specializes in data visualization, a field that presents data in graphical forms to help people analyze it. “Paul teaches risk communication to medical students. Holly helps us look at the psychology of medical decisions.”

So far, the team has tested a variety of visualization tools, charts, and diagrams to see how well patients understand them. A typical chart might use “icon arrays,” a set of different-colored stick figures, to compare the number of patients who have a disease to the number who test positive for the disease and how much or little the two groups overlap.

Such graphics work well for people with strong spatial abilities, the team has found. But people with weak spatial abilities can look at these tools and not gain any greater insight into their actual health risk.

“Some people are good with spatial reasoning, some are not,” says Chang. He sees the team’s next challenge as coming up with a set of tools that will help all patients understand their actual health risk, no matter how skilled they are at reading charts. “One of my goals is to come up with better designs that will help patients make more informed decisions about surgery and other potentially dangerous treatments.”