Executive Perspective

Using “Big Data” to Predict – and Improve – Your Health

| Jun 26 2014
Top

The Centers for Disease Control and Prevention report that more than 29 million Americans have diabetes. Before a person has type II diabetes, they very likely could have metabolic syndrome.

Metabolic syndrome is a grouping of five risk factors – large waist size, high blood pressure, high triglycerides, low HDL (‘good’) cholesterol and high blood sugar – that may occur together. An adult with three or more of these risk factors is five times as likely to develop type II diabetes as a person who does not have metabolic syndrome. They are also twice as likely to develop heart disease. And metabolic syndrome is even more common than diabetes, affecting one in four adults in the U.S.

These numbers are startling, but there is good news.

Many people with metabolic syndrome or pre-diabetes can avoid type II diabetes or heart disease by becoming more active, starting a healthier diet and losing weight. But how does somebody know if they are at risk for metabolic syndrome or diabetes?

Aetna’s Innovation Labs teamed up with GNS Healthcare to use “big data” to predict the one-year risk of developing metabolic syndrome. The groups used computer models to study information from 37,000 members of one of Aetna’s employer customers. The data came from:

  • Medical claims
  • Pharmacy claims
  • Demographics (age, gender)
  • Lab tests
  • Biometric screening results (blood pressure, cholesterol)

The models were able to predict the risk for metabolic syndrome among groups and individuals – even down to the specific risk factor. The models could also create personalized exercise, weight management, and care management programs for individuals. The study also found that:

  • Reducing waist circumference and blood glucose had the largest health benefits and the biggest reduction in medical costs.
  • Two other actions helped people change their risk factors – having regular doctor visits and using their prescription medicines appropriately. Having a routine, scheduled outpatient visit reduced the probability of metabolic syndrome in nearly 90 percent of people.

Results from the study are published in the June issue of the American Journal of Managed Care.

After the study, two more Aetna employer customers started using the metabolic syndrome reporting and prediction capability. The models helped identify people at risk for metabolic syndrome. These people then received specific suggestions for how to reduce their risk. Because of the success of this program, we expect the metabolic syndrome predictor will be more broadly available to Aetna employer customers in 2015.

Using data to help predict when people might have health problems means we can try to help them avoid those problems. This is another great example of how technology and information can create “population health” tools that help empower people to lead healthier lives, reduce health care costs and improve the health care system.

Adam Scott, Managing Director at Aetna Innovation Labs, co-authored this article.

Aetna Innovation Labs

Aetna Innovation Labs is a source for unique ideas and programs that provide market-leading capabilities to improve health care quality and reduce costs for our customers and members. Through this organization, Aetna can:

  • Test specific initiatives such as those related to disease prediction and intervention;
  • Rapidly determine success rates and impact across populations of members; and
  • Quickly expand programs that show promise.

The information in this story is based on a pilot program. It is not currently available to all Aetna members. However, it may be expanded to more individuals in the future.