The University of Chicago Medicine is collaborating with Google to study ways to use data in electronic medical records to make discoveries that could improve the quality of health care. The work focuses on using new machine-learning techniques to create predictive models that could help prevent unplanned hospital readmissions, avoid costly complications and save lives.
“Prediction helps make patient care better. It’s a core component of prevention, and it can also make complex care safer,” Michael Howell, M.D., Chief Quality Officer at UChicago Medicine and director of the Center for Healthcare Delivery Sciences and Innovation (HDSI).
By combining machine-learning tools pioneered by researchers at Google with UChicago Medicine’s health care predictive modeling expertise, the ongoing collaboration will bring leading technology to bear on real-world health care problems. UChicago Medicine joins Stanford and the University of California, San Francisco, in the effort, collaborating with Google’s machine-learning research team to improve prediction in health care.
Howell, who is a practicing intensive care physician, has been working with predictive and risk factor models for health care for more than 15 years. Yet, he says much of the most valuable information about a patient’s previous medical care is still inaccessible through electronic medical records.
“One of the amazingly frustrating parts of that work has been knowing that we can’t use much of the data in the electronic health record, like doctors’ notes or X-rays,” he said in a press release. “Traditional tools of epidemiology and statistics simply can’t use free text or images to create predictive algorithms that could alert physicians and nurses about patients’ risks for problems. But together with Google, we can.”
Each year in the U.S., unplanned hospital readmissions cost as much as $17 billion; the CDC estimates that health care-associated infections lead to 99,000 deaths, and problems with medications cause more than 770,000 injuries and deaths.
At the same time, the amount of electronic data created by routine health care encounters has also multiplied. In 2008, only 1.5 percent of U.S. hospitals had a comprehensive electronic health record. Today, the amount of health care data on the planet is measured in exabytes and zettabytes.
Now, UChicago Medicine, along with UChicago’s Center for Research Informatics, is collaborating with Google to help make sense of these data. According to the press release, Google researchers have already shown over the past year how machine-learning can help clinicians screen for breast cancer and preventable blindness, translate research results into medical devices, and provide alerts to physicians about hospital patients at risk of deterioration. Working together with UChicago Medicine and its other partners, the company is planning further studies on the accuracy of predictive models, complete with peer-reviewed research publications.
“We believe clinical breakthroughs using machine learning will come only when the medical community and deep learning experts collaborate closely,” Google stated in their blog post. “Most of us at Google are not doctors, but everyone has been touched by an illness or injury, or even lost a loved one. We have something unique at Google that we can contribute to make care better — so we must try. We look forward to growing our healthcare partnerships in the hopes that together, we can improve the health of many millions of people worldwide.”
