Big Promise From Big Data

Oncology Business News®, October 2014, Volume 3, Issue 5

How can we slow down the growth of cancer costs and preserve-or better yet-enhance clinical outcomes?

Andrew L. Pecora, MD

Editor-in-Chief of Oncology Business Management

Chief Innovations Officer, Professor, and Vice President of Cancer Services John Theurer Cancer Center at Hackensack University Medical Center

President, Regional Cancer Care Associates, LLC

How can we slow down the growth of cancer costs and preserve—or better yet—enhance clinical outcomes? This question is vexing payers, providers and policy makers, and never forget, patients. In our current issue, it would appear that private industry, public efforts, and even, academia, are all aiming to solve the riddle of delivery of value-based care at scale to control cost and improve outcomes by using big data. There appears to be two general approaches evolving in the market place to deliver big data to the bedside. Both hold promise to promote the Goldilocks formula: not too much care and certainly not too little care…just the right amount.

The first involves what some have called the “boil the ocean” approach. Efforts from IBM (Watson), ASCO (CancerLinQ), and Flatiron Health (OncoAnalytics) and others all plan to capture large quantities of data and distill it down to actionable information, some in real-time and some in a retrospective approach to better guide decision making, and as the theory goes, change behavior through information. Capturing and distilling vast quantities of information may show patterns of care that are less effective, or equally effective but more costly. Challenges exist with this approach, however, including but not limited to, inaccurate data in the electronic health record (estimated up to 40%), lack of specificity to individual patients with co-morbidities or specific genomic mutations, and lack of timely and focused financial information.

The second approach provides point-of-service clinical and financial information in real-time, after initial sorting of patients by clinical and molecular phenotyping has occured. Taking this approach will enable a transition from fee for service to at-risk bundle type reimbursement. It also leans more on reimbursement reform to change behavior at scale and ties clinical outcomes combined with cost to ensure value.

Cancer Outcomes, Tracking, and Analysis, or COTA, (for transparency, a company I am involved with), recently announced a strategic investment from Horizon BCBSNJ and Med-Metrix Inc. The significance of this is that a dominant payer in a large market, along with a revenue cycle financial analytics company, have teamed up with practicing and academic oncologists to build a cloud-based solution for transparent value-based care delivery. This is an unlikely trio to collaborate in a private enterprise to build something, but the effect of the partnership could be profound. Challenges to this approach remain, because the requisite infrastructure for payment reform and acceptance of a common language of sorting needs to occur.

What is most encouraging is that American ingenuity appears to be well at work to offer solutions to the marketplace to avoid the need to even discuss rationing of care. We have spent over half a century uncovering the mechanisms of cancer and designing personalized approaches that truly are improving outcomes. My greatest hope is that big data does deliver on the promise and care outcomes improve while costs modulate.