Jennifer Johnson, Published August 10 2013
Learning from millions of patients to treat one
Called “analytics,” it’s a technique taught at few other medical schools in the nation, and it promises to improve not just the treatment but also the prevention of diseases.
For rural North Dakotans, this means a patient with a certain set of symptoms could walk into a hospital and, within a short time, have them compared with the symptoms of many others around the world. By finding matches, the technique can also find the most effective treatments and perhaps warning signs of hidden health problems.
Joshua Wynne, dean of UND’s School of Medicine and Health Sciences, said analytics would bring the best capabilities of metropolitan health care to rural areas, and, by preventing diseases, help the state manage its physician shortage.
“I cannot think of a better way of trying to address (preventable illness),” he said.
Funding for the Master of Public Health program was originally justified by Wynne at the Legislature as one of two ways the state can address the national shortage of physicians, which he expected would hit harder here because of the rural and elderly population. The other way was to expand the medical school, which UND is doing now.
Ten students were enrolled in the Master of Public Health program last fall, the first year it was offered, and this fall the same number have enrolled, said Raymond Goldsteen, the program’s director.
Administrators plan to cap enrollment at 30.
While analytics is just one component of the program, it is an important one.
The national health care system is changing rapidly, and analytics contributes to the many efforts currently underway to cut costs, increase efficiency and provide universal access to care, according to Goldsteen.
“We’re saying, just as you learn biochemistry, you need to learn this, too,” he said. “University leadership believes this will make UND, a school located in rural America, at the forefront.”
“People in the MPH program will have general instruction in epidemiology and the analysis of data, and some of the students will specialize in this area,” Wynne said. “So, it isn’t necessarily that all of the students will get all of the skills, but all of the students will get some of the skills.”
Specialization would prepare students for a career as an analyst, either in the area of population health research and evaluation or health management and policy.
The data public health students will have access to includes, so far, a vast library of results from clinical trials, electronic health records, studies and other research. UND is also forming a partnership with health care companies, such as Blue Cross Blue Shield, which would allow students to look at data taken from real patients, though none will be identified.
There will be an emphasis on the study and use of information from the rural communities of North Dakota.
“You’re studying the actual experiences of large numbers of patients to see what the trends are and what you can find out by analyzing their outcomes,” Wynne said.
Every time a patient is seen by a doctor, data such as diagnosis and treatment is collected. But now, the health care industry is trying to do more with the information by standardizing medical record entries so that physicians record the same information in the same way for all patients, Goldsteen said.
“This allows comparison between patients and the population, as a whole, to determine effectiveness of care,” he said.
As health care has grown over the past 20 years, and access to more data has increased, the industry really needs individuals to take that data, translate it into information and come up with ways to turn that information into knowledge, said Ruth Krystopolski, executive vice president for research and development at Sanford Health.
“I believe the program at UND will help students do that,” she said. “I’m really excited about this next generation having access to data, and really beginning to look at how we use that information to inform our decision-making processes and how we learn to better treat patients in the future.”
In recent years, analytics has been used to identify factors that reveal if patients are at risk for kidney failure or other acute diseases while staying in intensive care units, she said.
“What they’ve been able to do is work with the data sets and look backwards and say, if a person has a temperature of X and is on drug Y and exhibits Z, they will potentially or more likely have kidney failure in the next 24 hours,” she said. “You can realign standards to more closely monitor those people and to make sure you catch kidney failures early.
“I think how we practice medicine is going to change significantly over the next decade, and use of information will allow us to give people what they need when they need it,” she said.
Although analytics hasn’t been applied to all illnesses yet, it has been targeting chronic diseases. Major predictors of preventable illness and premature death such as cigarette smoking, obesity and high blood pressure, all of which can lead to chronic diseases, make up anywhere from 30 to 50 percent of the nation’s health burden, Wynne said.
Analytics could also help policymakers find out how effective health care spending is.
North Dakota spends about 17 percent of its gross domestic product on health care, compared to 8.9 percent in Wyoming and 21 percent in Maine, Wynne said.
“The question is, is North Dakota doing better? We have to drill down and try to be able to say we’re operating at the most effective, efficient way possible, and those answers aren’t there yet, but there is data that helps us be able to tell how well we’re doing,” he said.
Krystopolski said she applauds UND’s program. There’s a lot of value in using data for information that benefits North Dakota and a lot of other places, and there is always a significant amount of alignment between clinicians and hospitals.
“I really believe that analyzing data, even around chronic disease, will be able to help us intervene earlier and better manage patients,” she said.