In almost every industry, aggregating data on a large scale and running predictive analytics have the power to improve our lives. With healthcare, this power is magnified because conclusions drawn from analytics can directly affect patient health and well-being.
Unfortunately, discussions of so-called big data applications often are filled will vendor hype and sales hyperbole. It’s a shame because there are many practical lessons and examples to illustrate the value of predictive analytics.
One of these use cases came forth during a CXOTalk discussion with the one of the foremost healthcare CIOs in the world, Dr. John Halamka. He described a personal situation demonstrating how data and analytics can overcome certain limitations of traditional health care.
My wife was diagnosed with stage IIIA breast cancer in December 2011. The genomics of her tumor were HER-2 negative, estrogen positive, progesterone positive and at the time she was a 50-year-old Asian female.
Now, wouldn’t it be interesting to say:
Of the last 10,000 Asian females, with a tumor like this, how were they treated, what was the outcome. Did they get sicker, did they get well, what were the side-effects?Ensure she gets the medicine that seems to provide the best outcome for 10,000 people like her.
That is what we call a learning health care system. [In] today’s healthcare system, it takes, on average, 20 years for an innovation from one hospital to diffuse throughout the country.
We were able to take all the data at all the Harvard hospitals and do the query, to find the medication that would be most effective for her. She is totally cured, and everything is fine.
That’s sort of “big data,” although I’m not sure what big data is – we have three petabytes, so it’s not that big. We were able to treat her optimally using that kind of analytic across multiple institutional data sources.
CXOTalk brings together prominent executives, authors, and analysts to discuss leadership, technology, and innovation. Join me and Vala Afshar every Friday for a new episode of CXOTalk.