The IFTF Blog
PMWC Preview: Colin Hill on Big Data Analytics
Health Horizons researchers will be attending the 2013 Personalized Medicine World Conference here in Silicon Valley, January 28-29. In the run-up to the conference, we’ll be talking with some of the scheduled speakers to preview their talks at the event. (For more information on the conference itself, please visit the PMWC 2013 conference website.)
We had the privilege of speaking with Colin Hill, CEO and Co-Founder of healthcare analytics company GNS Healthcare, about how big data analytics will be instrumental in advancing personalized medicine. As a leading thinker in this field, Hill writes a Forbes blog entitled Healthcare 2020.
Since the founding of GNS Healthcare in 2000, Hill has witnessed an explosion in the amount of healthcare data available. Humans, however, are not equipped to make sense of all that electronic data. “Without computer driven and artificial intelligence approaches, all that data is just data,” Hill said. Advances in machine learning and cloud-based supercomputing are speeding up the rate at which this data can be used to generate new insights. GNS Healthcare is working on putting complex data sets into its machine learning platform “to create predictive models of what interventions work for which patients, and replace the standard of care for the hypothetical average patient with data-driven, personalized treatment algortihms.”
In a recent article co-authored by Hill, he explains the difference between the old way of understanding what works: a clinical trial showing that a treatment is 80% effective; and a new understanding: that same treatment is actually 100% effective for 80% of patients. This is far from semantics - especially if you are of the 20% for whom that treatment will not work. As the price of genetic sequencing continues to drop, having detailed molecular data will be standard practice. Hill thinks that, in cancer for example, eventually each treatment could be tailored to the genetic specifics of a patient’s disease. However, we will not arrive at this kind of personalized medicine through the double blind trials of the past. Instead, predictive computer models will work with the data from millions of daily health interactions and be able to pinpoint the best option out of thousands of possibilities. Hill attributes more than $500 billion in wasteful spending each year to treatments that don’t work or sometimes even worsen a condition. Eliminating this waste would improve patient outcomes while simultaneously reducing costs.
In September 2012, GNS Healthcare launched an initiative with Aetna that aims to treat and prevent metabolic syndrome (the disorder that often leads to heart disease, stroke, and diabetes, and according to the American Heart Association, affects 47 million Americans). Using Aetna’s claims data, GNS Healthcare’s supercomputing Reverse Engineering and Forward Simulation platform (REFS™) identifies those at the highest risk for the disorder, and helps pinpoint the factors for an individual which, if reduced, lead to the most effective reduction in risk of developing metabolic syndrome. This initiative was made possible because one of Aetna’s large national employer groups had data from several years of full biometric screenings for all approximately 100,000 of their employees. The success of personalized, early interventions like this helps support the need for full integration of electronic health records – both to improve population health and allow healthcare providers to meet the imperative to lower costs. GNS Healthcare also recently paired with Dana-Farber Cancer Institute and Mount Sinai School of Medicine to use this platform to create a computer model of multiple myeloma, the second most common blood cancer in the U.S., that will help researchers discover novel therapies.
As we start the new year, Hill says, “2013 and 2014 will be very big years in the transformation of healthcare into a much more data-driven enterprise that will have dramatic effects in terms of health outcomes for patients, and reduction of costs.” He sees this as “synergistic with Accountable Care [Organizations],” which incentivize lowering costs while improving outcomes – something that can only be achieved if healthcare providers know what works for whom. The U.S. healthcare system, which is now notorious for costing too much and delivering too little, could lead the way in advancing data-driven medicine. Within 10 years, Hill thinks that U.S. health outcomes could move into the top 5 from their current rank of 37.
And of course, improving outcomes is the end goal. “At the end of the day, we will have failed if we haven’t changed the way that someone gets treated,” he said.
Join us at the Personalized Medicine World Conference for more from Colin Hill and many others using big data analytics to fundamentally transform the practice of medicine!