Posted by: AtlasMD

October 1, 2013

The Future of Big Data in Medicine Looks Bright

MIT Technology Review published an exciting article. Jeff Hammerbacher, a 30-year-old reputed as Facebook’s first data scientist, is working with The Mount Sinai Medical Center. He once said, “The best minds of my generation are thinking about how to make people click ads. That sucks.” But those days are over. He wants to engineer data to help doctors better predict the likelihood of illness and complications.

Hammerbacher will apply his insight acquired while working to target online advertisements. These data-crunching techniques will be used to help doctors and researchers suck in medical information and spit out predictions. The hospital believes this data, if analyzed properly, could cut the cost of health care.

Will Big Data Promote Results-Driven Medicine?
Mount Sinai knows the healthcare system is shifting dramatically. And they are candid about the fact that because hospitals bill by the procedure, revenue is proportional to how sick patients are. However, there is healthcare reform in Washington that wants to push hospitals towards something called “accountable care.” In the model, the hospitals will be compensated by keeping patients healthy.

As always, implementing this proactive incentive within our behemoth government might backfire, but we’re not going to complain about the idea that hospitals make money by keeping their patients healthy.

Even before hiring Hammerbacher, Mount Sinai is using a computer model where factors like disease, past hospital visits, even race, are used to predict which patients stand the highest chance of returning to the hospital. The model, built using hospital claims data, helps caregivers identify chronically ill people. That way, staff can reach out to them to get extra attentive care, like follow-up calls and more. In a pilot study, the program cut readmissions by half; now the risk score is being used throughout the hospital. And that was before the hottest name in big data was on board.

We learned about this development from Marcie Stoshak-Chavez, a practicing Emergency Medicine doctor in Chicago, IL. She was excited about the announcement and posted her own blog about it. She wonders, Will a social media data scientist break the code of predictive analytics in medicine and improve health and wellness? She believes if great minds combine with great machines, perhaps it CAN happen. And we agree with her there.


But wait, will this hospital’s data help other doctors?
In the long run, maybe. But for right now, the data analysis will only empower doctors at Mount Sinai to predict events for current patients. However, Hammerbacher’s new computing facility is designed to supercharge patient insights. It will do this by running a version of Hadoop, a software that spreads data across many computers and is popular in industries like e-commerce. The fact remains–patient data are slim, and not very dynamic. According to MIT Technology Review, “Records get added to infrequently—not at all if a patient visits another hospital.” That’s a limitation, Hammerbacher says. Yet he hopes “big-data technology will be used to search for connections between, say, hospital infections and the DNA of microbes present in an ICU…”

Hammerbacher himself admits for now his data collection really only benefits and relies on patients who enter Mount Sinai. That said, what he’ll be capable of comparing will trump anything currently on the market. So before we all pour ourselves a glass and say data fixed healthcare, let’s realize we are a long way off from social medicine as connected as our public personas, the ones that live on Facebook.