At Mayo, big data is already improving health care. Consider the case of Javrie Burdell.

The six-year-old from Clovis, New Mexico—a cheerful lover of cartoon movies like The Nut Job—had his first seizures at two years old. Josiah and Renata Burdell, both 29, took him to a local hospital. That first night, Javrie’s respiration slowed to four breaths a minute. “Watching your child lie unconscious on a table for an hour is pretty real,” Renata says.

Puzzled doctors sent the family to Lubbock, 100 miles away, then 220 miles in the other direction to Albuquerque. Seizures continued, diagnoses multiplied, and his parents say Javrie’s development regressed. The Burdells went to four hospitals before, fed up, they Googled the top pediatric neurology departments in 2010, then wrote to Mayo.

The tests continued at Mayo. A spinal tap, an MRI, and tests for genes linked to known disorders were all negative.

In September 2012 the hospital opened its Center for Individualized Medicine, an interdisciplinary effort to use genomics to identify diseases that have stumped the world’s top hospitals, and Javrie became one of the first patients in its Disease Odyssey program. The idea of Disease Odyssey: to sequence patients’ exomes—a subset of the human genome that includes all the body’s important instructions for building proteins—and use resulting data to comb for clues.

“Is this big data, or just lots of data?” Mayo CEO John Noseworthy said at a conference last year. “This really is big data. This looks for associations we didn’t predict or anticipate and lets us really change the story.”

After years without a diagnosis, some 37 percent of patients get one within about three months, says Gianrico Farrugia, a gastroenterologist who directs the center.

Javrie’s journey took longer. Even after he came to Mayo, it took years for members of the hospital’s team of 369 data scientists to invent the technology that would diagnose him as one of 10 children on earth with a mutation so rare the syndrome it causes has no name. To a six-year-old, it was a simple blood draw, taken last July. Javrie’s parents also supplied DNA samples. Behind the scenes, much more was happening, Farrugia says. In cases like these, samples are sent out for sequencing and analysis. Data is compared with medical records, published literature, and normal genomic patterns to try to forge a diagnosis. A team of as many as 20 people, from geneticists and pathologists to ethicists, review each pending case every Wednesday.

On January 3, the Burdells were told that their results were ready. After an 18-hour drive, they learned that Javrie has a mutation on the PACS-1 gene. That accounts for his balance problems and subtle facial abnormalities—a slightly bulbous nose and widely spaced teeth that many people don’t initially notice.

Big data gave the Burdells the comfort of understanding the cause of their son’s challenges, but it did not provide a cure. “It’s really kind of a bittersweet deal, but our life’s better,” Renata Burdell says. Still juggling symptoms and medications, she and her husband hope the next wave of data unlocks therapies for a little boy who still knows only 160 words.

Most times, of course, big data in health care focuses on simpler problems that affect more people—crunching data from many sources to spot similarities between cases, identifying treatments that work best and cost least. At Mayo, that means efforts like the three-year-old Center for the Science of Health Care Delivery, which studies innovations in organizing care. The center’s projects include a series of departmental clinical engineering labs, each managed by a clinician in that specialty and a data scientist. The first lab was built for emergency medicine.

The ER at Mayo’s Saint Marys campus is now taking advantage of an ongoing renovation to embed RFID equipment in the walls, the better to study patient flow. In a year, Mayo hopes to have better data on how it keeps track of patients, how long it takes to get them seen by a doctor, and how long they stay in the ER, says Tom Hellmich, a pediatric ER doctor and one of the managers of the lab studying the emergency room. Of special note: a project to understand and solve the problem of psychiatric ER patients who need to quickly find spots in programs that can help them. Data from that effort should be useful to both hospital administrators and legislators interested in mental-health reform, says Kalyan Pasupathy, co-director of the Emergency Department Lab.

Of all Mayo’s data initiatives, the effort that may have the most long-term impact is the partnership last year with United Health Group’s data-analysis unit, called Optum Labs.

The idea is to study claims records from 109 million patients, contributed by Optum, and 30 million medical records, including five million from Mayo. Research under way is investigating such topics as how compounds’ performance in clinical drug trials compares with their effectiveness in large patient populations once approved; how medical practice varies in different locations; and how to attack problems like excessive hospital admissions.

As changes in policies and insurance shift more cost to patients and providers, there are financial as well as medical reasons to pursue this kind of research.

One study of joint-replacement patients under 65 found that weight and diabetes played previously unknown roles in knee and hip problems, says Nilay Shah, associate professor of health-care policy and research at Mayo. Early intervention could mean fewer operations.

“Is this big data, or just lots of data?” Mayo CEO John Noseworthy said at a conference last year. “This really is big data. This looks for associations we didn’t predict or anticipate and lets us really change the story.”