Big Data's Promise, Potential and Pitfalls

Medicine has been considered a mix of art and intuition, but more and more it’s going to be informed by data, said U.S. News & World Report Executive Editor Margaret Mannix at a session during the U.S. News Hospital of Tomorrow forum Tuesday. The session addressed big data’s promise, potential and pitfalls.

Jon Bandes, administrative director of business intelligence at Memorial Healthcare System, discussed the changes his system has made and is making, thanks to what it’s learned from data. His position, for example, was created in response to the growing need to collect and analyze data in order to control costs and to provide better services.

[WATCH LIVE: The 2014 Hospital of Tomorrow Conference]

One challenge, Bandes said, is a knowledge gap between health care employees who want to use the new systems and their ability to do so. “What we’re trying to do is remove the silos to better educate our people and to move forward,” he said.

Niloo Sobhani, the corporate director of information systems at New York Presbyterian Hospital, talked about some of steps her hospital has taken to work with big data, including creating a “big data” team tasked with learning about the topic and testing various initiatives “without the pressure of a promise or a deadline.”

The team is now trying out various projects, such as one tool that creates a visual story of a patient’s medical history – in one page. The idea is to make it faster and easier for clinicians to learn what they need to learn – but no more – about new patients. “The challenge was, too [often], physicians are being inundated with information during patient care episodes,” Sobhani said.

The third speaker, John D. Halamka, chief information officer at Beth Israel Deaconess Medical Center, addressed an overarching problem with big data: “Turning data into knowledge and wisdom,” he said.

[READ: There’s a Giant Job Ahead for Big Data]

One way his center has worked to address this challenge is by employing “care traffic controllers,” or nonclinical staff who look at patient data to determine ways to reduce re-hospitalizations, readmissions and address other issues. If a patient with COPD, for example, comes to the hospital every August, these employees might be the ones to notice the trend – and suggest buying the patient an air conditioner in response. “Maybe you can reduce sickness through a social rather than medical intervention,” Halamka said. “These are some of the things big data can do.”

Lastly, Sachin Agrawal, director of provider solutions at IMS Health, offered audience members “lessons of the road” by detailing three main questions medical systems should ask before embarking on a project with big data: First, consider what problem you are trying to solve; then, ask what the real need is, he said. While many of Agrawal’s clients think they want all the data possible in real time, “that’s not necessarily the answer,” he said. Finally, and importantly, evaluate how ready the organization is to adopt big data, he said.