Since finishing my research on Big Data, I have been wondering about the potential of mixing data–where the sum of the parts is greater than the individual pieces. I was delighted to meet like-minded individuals bringing creativity, zest and enthusiasm to the discussion, while attending my first StrataRx.
Here are six “flavors” of data sharing as explicated at StrataRx:
1. Stephen Friend and John Wilbanks of Sage Bionetworks
My imagination was sparked listening to Stephen Friend and John Wilbanks of Sage Bionetworks, a not- for- profit research organization exploring open models in the practice of biomedical science and medical research community.
Stephen’s StrataRx talk, entitled “Dreaming of IPO’s While Patients Die” was an emotionally powered description of illusions about our medical system, with hopes that citizens and patients can be activated to improve outcomes. Stephen argues that physics has made the transition to a data-intensive science- with equipment, IT interoperability, open information systems, and places to post the information- and so can healthcare.
On a similar theme, in John’s TED talk, he describes the unintended consequences of informed consent and suggests that we build a commons – a public good built from private goods.
In John’s StrataRx talk, a discussion of policy entitled, “Choose Your Monopolies Wisely” he discusses the concept of ”openness as a monopoly”, and asserts that it is a worthwhile long term investment, assuming that we build a standards-based privacy platform for informed consent that is honest, portable and reusable.
At Sage Bionetworks, Stephen Friend and John Wilbanks have created Synapse a collaborative compute space that allows scientists to share and analyze data together and create a rapid learning environment.
They are proud of a recent project on breast cancer, that put a cohort of 2,000 patients into the public domain, allowing anyone to build a classifier. Demonstrating the power of crowdsourced competitions, it took only six weeks for a single individual to beat the existing classifiers.
2. Fred Trotter shares a Medicare Data Set
Fred Trotter-@fredtrotter – a well-known HIT Tweeter, and a long time advocate for open source and e- patients, had us all in suspense about his mystery data set. Wearing a new Walking Gallery jacket, with a painting of the world upside down, his StrataRx talk, entitled “Buxton’s Ghost: The Ethos of Healthcare Data Science” released a Medicare Referral dataset, which shows referral patterns for medicare procedures.
Calling it a “doctor social graph,” he asked developers to create an algorithm that “patients find useful and doctors find fair.”
Although this may sound simple, it took him about two years to get the data, spending a tremendous amount of time engineering his request so that it fully complied with a 1979 court injunction that barred the release of data about physician compensation from Medicare. The court injunction means that you cannot get data that can be used to calculate a doctor’s payments from Medicare.
He is planning to make the data open source with a crowdfunding project on Medstartr.
3. Marcia Kean shares cancer data with Project Data Sphere
Marcia Kean, a long-time savvy healthcare communication expert and founder of Feinstein Kean, brought yet another approach to data sharing – Project Data Sphere. Her StrataRx session entitled “Slaying the Dragon,” describes the efforts of the CEO Roundtable on Cancer’s Life Sciences Consortium to bring data sharing to cancer.
The first project, led by pharma company Sanofi, will contribute de-identified clinical trial data into a freely accessible repository. It is hoped that other pharmaceutical companies will follow suit with their own clinical trial data.
Future projects will include development of standards that are lacking for data exchange; collection of longitudinal healthcare data from patients through their membership in disease advocacy organizations; and campaigns to change the protocols of government research grants to weave in data-sharing as a requisite of funding.
See the next post to learn about 3 more data sharing flavors coming into the mix.