Mount Sinai Medical Center is experimenting with innovative data-crunching techniques that could one day help doctors to make personalized predictions for their patients, Courtney Humphries writes in the MIT Technology Review.
Mount Sinai—a 1,406-bed hospital and medical school that treats half a million patients annually—is beginning to operate like an information business, Humphries reports. For instance, it has assembled a biobank with 26,735 patient DNA and plasma samples, as well as just having completed the installation of a new $120 million electronic health records system.
Now, the hospital's Icahn Institute for Genomics and Multiscale Biology is building a new computer cluster, adding to an existing supercomputer that runs from a basement of a nearby building. Through computing, it hopes to alter the way physicians provide care by proactively identifying risk and trends in patients through data.
- What is Crimson—and how can it help my hospital? From billing systems to electronic medical records, there's no shortage of useful data—but where do you start? This video explains how Crimson Clinical Advantage is helping over 1,000 health care organizations turn the chaos of big data into big outcomes.
Investing in 'big data'
To design the new computer cluster, Mount Sinai has recruited 150 experts in the last year, including Jeff Hammerbacher, a 30-year-old known for being Facebook's first data scientist.
Hammerbacher says he plans to apply the same techniques used at Facebook to target online advertisements to develop a research engine that uses medical information to generate predictions that curb health costs. "We're going out on a limb—we're saying this can deliver value to the hospital," says Hammerbacher.
Mount Sinai hopes "big data" will allow it to succeed in the transition to new, value-based payment models, according to Eric Schadt, who runs the Icahn Institute. He says the prospect of new economic incentives for hospitals that can cut costs and provider high-quality care has shifted Mount Sinai's position to, "'Hey, use all your resources and data to better assess the population you are treating.'"
How Mount Sinai hopes to the use the 'big data'
Mount Sinai has already begun experiments with computer models that use factors like disease, previous hospital visits, and race to predict which patients are likely to be readmitted. The model relies on hospital claims data and identifies which patients should be targeted for follow-up calls and extra care. In a pilot study, the program cut readmissions in half.
Hammerbacher 's new computing facility is designed to "supercharge" these types of insights. It will operate a version of Hadoop, software that spreads data across many computers to generate large amounts of rapidly changing information. Among other things, Hammerbacher hopes the technology will be able to:
- Uncover connections between hospital infections and DNA microbes present in an ICU; or
- Track data from patients wearing at-home monitors.
Hammerbacher will work with Joel Dudley, director of biomedical informatics at Mount Sinai's medical school. Dudley predicts that in the future, every patient will be represented by a "large dossier of data." Before they are treated, or even diagnosed, they will be compared to "every patient that's ever walked in the door at Mount Sinai," he says, adding, "[Then] you can say quantitatively what's the risk for this person based on all other patients we've seen" (Humphries, MIT Technology Review, 9/26).
Next in the Daily Briefing
Let your kids lose—they might ultimately win