Panel
Big Data for Better Cures: Where to Invest?
Milken InstituteCecilia Arradaza, Marty Cohn, Joel Dudley, Magali Haas, Risa Sack, Andy von Eschenbach, Harlan Weissman, Garrett, Mary Dwight
- Accelerate medical research to save time and lives by treating big data volume, velocity, variety, veracity, value, and variability as fuel, though variety presents challenges due to the necessity of integrating all available data rather than relying on single sources.
- Implement $250 million investments in genomics and informatics at the Icahn School of Medicine at Mount Sinai and approximately $800 million in big data, digital health, and healthcare IT over the last year to build learning health systems and predictive models.
- Project that within 10 years technology will enable targeted drug therapies for tumors, predictive healthcare via mobile devices, and localized services such as "doc in a box" or mobile units, requiring clinical validation of existing technologies.
- Address a looming socioeconomic crisis where brain disorders affect one in five people and represent a two trillion dollar cost in the U.S., with Orion BioNetworks developing models similar to weather prediction to understand disease causes.
- Utilize data to reduce waste and control costs by targeting 75% of healthcare spend on chronic diseases, prioritizing prevention over hospital admissions to avoid expensive complications and infections.
- Shift from fee-for-service to pay-for-value models, currently representing less than 5% to 10% of the market, driven by incentives to keep patients healthy and reimbursed monitoring technologies for conditions like diabetes.
- Redefine disease as a massive continuum rather than discrete buckets, with Mount Sinai undertaking projects to create a normal average human model at DNA, immune, metabolite, and brain levels on a minute-by-minute basis.
- Transform healthcare paradigms through consumer-directed care and employer incentives, as self-insured entities like GE seek data to optimize benefits and patients become more cost-incentivized due to shifting costs.
- Leverage vast data exhaust from monitors and patient-powered networks, where 30,000 patients at Mount Sinai's biobank have consented to broad data reuse, though adherence data sharing faces reluctance due to privacy concerns.
- Anticipate regulatory changes including relaxed rules on diagnostics and improved piloting processes from payers, while pharmacy benefits managers use big data to monitor compliance and correct trends.
- Apply Bayesian network inference and clinical decision-making approaches to unify social, emotional, environmental, and genetic data, aiming to determine point-of-care needs rather than focusing solely on mechanisms or cures.
- Recognize risks such as the hype cycle where reality sets in requiring significant work, inconsistent data conflicts predicted to reach 80% by 2015, and historical difficulties in hospital-centric product sales and insurer willingness to pay for wellness.