Doctors Wielding Data

Helping patients and cutting costs

If you told most people when they walked into a hospital that the care they were going to get would be as strictly regimented as production in a factory, they'd probably turn around and walk out the door. Americans tend to think quality medicine means customized medicine. For care to be good, it has to be tailored just for them.

Brent C. James says people's ideas about health care are all wrong. As vice-president for medical research at nonprofit Intermountain Health Care in Salt Lake City, he's using data and networks to standardize Intermountain's 21 hospitals and 90 clinics. The key: Pore over data on different ways of treating diseases, find the approach that works best, and get every doctor to use it. It's a strategy that's gaining currency in medicine.

Standardization, James says, will mean fewer mistakes, more consistent application of the latest research, and lower costs. His belief that a data-driven redesign of care can clean hundreds of billions of dollars in waste out of the $1.9 trillion U.S. system is grounded in Intermountain's results. Its inpatient hospital costs are 27% below the national average, and its operating margins, at 4%, are near the top of the hospital industry. The trade publication Modern Healthcare has tapped Intermountain five times as the best-run of the nation's 582 integrated health systems. Intermountain CEO Bill Nelson says: "We may be the only hospital in the country that can demonstrate the business case for using best practices."

"Systems Envy"

Certainly, Intermountain is one of America's most tech-savvy health-care organizations. It has poured $106 million into info tech since 2000, helping it make the American Hospital Assn.'s list of the 100 most-wired hospitals and health systems for six straight years. And it's a pioneer in adopting cutting-edge systems, from digital patient records in the 1970s to automated pharmacies beginning in the early '90s. "When I look at Intermountain, I get systems envy," laughs Ivan Seidenberg, chief executive of telecom giant Verizon Communications (VZ ) and a member of a federal commission studying health-care data systems.

The real secret, though, is how Intermountain uses the data from those systems to continually reevaluate and redesign care. Teams of doctors use the electronic medical records of every Intermountain patient to periodically review the approaches the institution takes in treating diseases such as diabetes or heart failure. James's trick: Get doctors involved in the process of coming up with approved ways of treating them, so they won't rebel when asked to use a single therapy. "He's not saying doctors are less important; he's saying that now they are important in new ways," says Harvard Business School assistant professor Richard Bohmer, who co-authored an HBS case study on Intermountain.

The system's two dozen or so care-process models have made big changes in the administration of drugs and treatment of diseases, including pneumonia and diabetes. A careful study of how Intermountain's LDS Hospital was doling out medication led to changes that slashed so-called adverse drug events to 230 a year, from 570. Since the industry rule of thumb is that 3% to 4% of adverse drug events are fatal, that could save about a dozen lives a year. Changes in how oxygen is administered to critical care patients reduced fatalities by preventing pneumonia, and saved millions of dollars a year. "Brent is showing that when you spend more, you don't necessarily get more," says Megan McAndrew, editor of the Dartmouth Atlas of Health Care at Dartmouth Medical School.

Still, Intermountain's approach can be controversial. After data was used to detect a pattern of complications among newborns whose birth was induced before their mothers were 39 weeks pregnant, Intermountain adopted a protocol preventing most early inductions unless medically necessary. Intermountain saves about $500,000 a year and, more important, keeps about 5% of the babies who would arrive early from a stay in neonatal intensive care. But early induction is a popular choice. About 30% of pregnant women nationwide use it.

Still, a big question for health systems eager to mimic this success is whether James's approach will work outside of Intermountain's walls. Plenty of hospitals are trying. But McAndrew argues an important reason for Intermountain's low costs is that the Utah market hasn't been glutted with excess hospital beds, as many other markets have. So it doesn't have the same pressure to fill beds that doctors in other markets may. And unlike most hospitals, Intermountain can standardize care more easily because it employs many of the doctors who practice in its hospitals.

Even James concedes that, up to a point. But there's little doubt that innovations like Intermountain's raise the quality of care. Even if major cost cuts prove elusive, experts believe tech-based reengineering of care is an important way policymakers and private companies can at least better contain health inflation. With insurance prices rising twice as fast as inflation, James's ideas may be coming soon to a hospital near you.

By Timothy J. Mullaney

    Before it's here, it's on the Bloomberg Terminal.