Mount Sinai and the Atlantic Host Working Summit on Big Data in New York City

Vast amounts of data from genomic sequencing and electronic health records (EHRs) have the potential to radically improve the health of individual patients, but first, institutions must learn how to manage the data, and adopt uniform standards in usability that will enable them to share vital information.

This discussion was center stage at a Working Summit on Big Data hosted by The Atlantic in partnership with the Mount Sinai Health System on October 23 in New York City, at which 24 policy makers, entrepreneurs, and health care leaders shared their thoughts and experiences in harnessing petabytes of data for use in improving human health.

"Big data analytics influence how we deal with individual patients. We now sequence the genetics of a single tumor, model that tumor in an animal model, test drugs on that animal model and take those findings back to the patient. We can't do that without a tremendous amount of data," said Dennis S. Charney, MD, Anne and Joel Ehrenkranz Dean, Icahn School of Medicine at Mount Sinai, and President for Academic Affairs at the Mount Sinai Health System. Speaking at a roundtable discussion moderated by Corby Kummer, Senior Editor of The Atlantic Dr. Charney added, "We're responsible for the health of millions of people. We have to determine what works and what doesn't in large numbers of patients with a lot of different diseases, and that has required us to make a significant investment in a high-performance computing system."

Few academic research institutions have on-site computing capabilities comparable to Mount Sinai's, which allow researchers to integrate and analyze different dimensions of biological data from large populations of people, and perform predictive modeling. At the center of this infrastructure is Minerva, a powerful supercomputer named after the Roman goddess of wisdom and medicine that has helped Mount Sinai scientists publish more than 25 research papers in the last year-and-a-half. Minerva is used for numerous initiatives across the institution. Examples include storing information from Mount Sinai's biobank, or BioME, a collection of DNA and plasma samples from more than 27,500 individuals, and then analyzing them for specific diseases, and sequencing thousands of individuals with autism spectrum disorder for the multinational Autism Sequencing Consortium.

Under the leadership of Eric Schadt, PhD, Director of the Icahn Institute of Genomics and Multiscale Biology, Mount Sinai has recruited top data scientists from Stanford University, IBM Corp., and other leading institutions to form a new epicenter for personalized medicine that will lead to better treatments for patients. Through the use of sophisticated clustering algorithms, Dr. Schadt, also Chair of Genetics and Genomic Sciences, and the Jean C. and James W. Crystal Professor of Genomics, says Minerva analyzes and categorizes patients so they can receive personalized treatments. "Once you've grouped the patients, then you can see how they were diagnosed, how they were treated, and whether the treatment worked," he said.

Many of the summit's participants acknowledged the steep learning curve, significant financial investment, and challenges involved in sharing data with other institutions.

"Not all data is good data, and not all data is standardized. Sometimes the problem is how well you asked the patient a question, or how well the answer was coded," said David Artz, MD, Chief Medical Information Officer at Memorial Sloan-Kettering Cancer Center. "Sometimes we can't get apples to apples comparisons across cancer centers."

Jim Peters, Founder and Chief Executive Officer of MedMining, a Geisinger Health System unit which licenses HIPAA-compliant, de-identified EHR data extracts to promote health research, said it is important to consider "smart data," which is often not reflected in health records. "One percent of the patient's experience is their encounter with the physician," he said. "What if during the other 99 percent of the time they're making poor choices? We need to gather more data from the patient. The more we can encapsulate behavioral preferences, the closer we will get to having smart data."

Kenneth L. Davis, MD, Chief Executive Officer and President of the Mount Sinai Health System, said Mount Sinai understands the importance of managing the health information of large populations of patients, and has created incentives for physicians to record thorough patient information.

"We're on the cutting edge of getting our physicians to use the data we generate," Dr. Davis said. "In between the clinicians and the data are case managers who make sure the physicians act on that data, and a significant percent of a physician's salary is tied to using the data effectively to control readmissions, inappropriate utilization, and all the other things that drive health care costs. When a patient misses an appointment or doesn't fill a prescription we know it right away and our people are on the phone making sure that mistake doesn't become a major health event. If we don't align the payers, the providers, and the patients with the same incentives the system doesn't work."

Mount Sinai is also using Minerva to change patient care at a population level. In recent study of Medicare patients enrolled in Mount Sinai's Accountable Care Organization, Dr. Davis said Minerva helped researchers discover that the use of comprehensive EHRs cut the 30-day readmission rate by 44 percent, and emergency department visits by 53 percent.

Creating an open platform for data sharing was another important topic of discussion at the summit. Today, most health information is collected and siloed by those who have access to it. The event's participants agreed that incentives were needed to encourage sharing across institutions and with patients themselves.

"The days of hoarding data will end. The winners will be those who understand how to leverage the data," Dr. Schadt said.

"Bravo to open source information," said Deborah Estrin, PhD, Professor of Computer Science at Cornell NYC Tech, and Professor of Public Health at Weill Cornell Medical College in New York. "We will have to push that from a policy position."

Some summit participants, including entrepreneur Vinod Khosla, Founder of Khosla Ventures, said an open flow of information would also empower more patients to take control of their health.

Dr. Charney explained that Mount Sinai's innovative new tool called CLIPMERGE (Clinical Implementation of Personalized Medicine through Electronic Health Records and Genomics) provides physicians with guidance at the point-of-care by showing them the patient's genetic profile and sending alerts to avoid prescribing certain medications.

The "holy grail," said Dr. Schadt, will be when patients and their physicians have a meaningful discussion about the information they have, and use it to impact their decision making.