Newtown, CT – October 4, 2013
There's a lot of talk about "big data" and its potentially transformative power in health care, and the Health 2.0 conference this week showed how transformative it could be – Feedback loops are the quintessential way in which all professionals get insights on the effects of their activities. By creating sequential designs of experiment, professionals can also use those same feedback loops to refine processes and interventions and move continuously towards more optimized results. Harnessing the power of "big data" into those feedback loops can provide information not only on what an individual professional is doing, but how that activity compares with that of others. As a result, an entire community can learn and improve far faster than has ever been possible until now. Several examples were shown at the conference. One was from IBM working with a health system and turning the EMR data into epidemiological timelines of CHF treatments. These timelines revealed that the most commonly adhered to treatment protocol wasn't necessarily the one yielding the best outcomes. Another was from the DocGraph project, which revealed the hidden virtual networks that exist between providers. Hidden because they're not based on legal arrangements, but rather on the referral behaviors of each provider. Those maps are essential for consumers and provider organizations as they shed light on the most likely sites for downstream care.
What this means to you – We've only begun to scratch the surface. The complex layers of relationships in health care that, together, conspire to rake in close to $3trillion have stayed opaque for a reason – easier to hide inefficiencies, conflicts of interest, and simply to maintain the status quo. These emerging projects start shedding some light on what the system could achieve, how we could better organize around clinical evidence, around poles of excellence, and in virtual networks that optimize patient care. The release earlier this year of hospital charge data by Medicare spurred RWJF to launch a simple challenge and get submissions for interactive and static visualizations of these data. In a few short months, dozens of individuals and companies found ways to turn a few spreadsheets into far more. The results are remarkable even if we all discount the validity of the underlying data. The point is that there are hundreds of individuals and companies that can deploy human ingenuity to transform these data into actionable and powerful information – feedback loops – if only given the chance. However, big data's potential to transform the industry is only possible if big data are accessible. The government now has the legal right (and, we feel, the obligation) to release the full Medicare claims files. It should do so with all haste. Employers who are guarding their claims data should similarly freely and quickly release those data to All-Payer-Claims-Databases in every state in which they incur medical claims. It's only when these data will be available that they can reveal what we have so far guessed and yet now know to be true. The layers of opacity must be peeled back, and the data transformers are there now to do just that. Let's help them reveal the truth.