CHICAGO – Navistar’s International Truck and Engine Corp. uncovered a surprising fact two years ago after analyzing employee medical claims.
The truckmaker’s workers were nearly twice as likely as the public to suffer from diabetes, a chronic condition that can lead to serious health problems. One reason is average age – a relatively high 48 years. Add to that some were overweight.
Employees with diabetes accounted for a hefty 24 percent of the company’s health-care spending. Left alone, the trend would only get worse. Armed with these findings, the company’s medical staff developed a voluntary testing and prevention program.
Big companies have seen per-worker health-care costs rise by more than 80 percent since 2000, according to benefits consultant Hewitt Associates. Struggling to get ahead of this curve, firms such as International Truck are turning to an emerging science called predictive modeling that promises to help them control spending by intervening before workers get seriously ill.
They’re analyzing claims data and information collected from employees about their lifestyles for clues to why some workers rack up average medical costs one year only to go off the charts the next. Using artificial intelligence, they’re searching for combinations of factors that are likely to lead to sharply higher medical claims.
At Pitney Bowes, which like International Truck has an older work force with a higher incidence of chronic disease, the company found it wasn’t only a diagnosis that predicted cost spikes, but whether employees filled all their prescriptions.
So the company cut the cost of co-payments on drugs for diabetics and asthmatics to ensure more workers would follow through with treatment.
The change reduced the average annual cost of care for diabetics and asthmatics by 6 percent and 15 percent, respectively. Even pharmacy bills declined because workers suffered fewer complications.
“It resulted in savings for us and improved quality of life for our employees,” said chief medical director Dr. Jack Mahoney.
But ethicists say the trend toward digging deeper into medical histories and lifestyles raises troubling issues. Employees may not feel free to refuse to participate in prevention programs, or they may fear losing their jobs if they turn out to be potentially high cost.
Employers say privacy is not an issue because they look only at aggregate data. They rely on outside firms to work with individuals under strict confidentiality agreements.
Even so, some find the trend worrisome.
“We’ve known for some time that someone who smokes or is morbidly obese is high risk,” said Mark Rothstein, director of the Institute for Bioethics, Health Policy and Law at the University of Louisville School of Medicine. “What we now see is an attempt to go beyond common-sense prevention strategies to a much more multidimensional inquiry into health status. The predictions get a little bit iffier and the intrusiveness gets ratcheted up.”
At Humana Inc.’s Center for Health Metrics in Louisville, Ky., computers produce clinical profiles that look like weather maps, turning from blue to red-orange when health risks increase.
“We try to find people at those tipping points, before they peak,” said the center’s director, Carol McCall.
International Truck’s chief medical officer, Dr. William Bunn, said better-integrated data allow health-care providers to intervene earlier.
“We’d notice you’ve been to the doctor, had some foot care and are on a high dose of insulin, so we’d call and offer coaching,” Bunn said. “We’re going to intervene because we see a pattern of medication and clinical care that suggests we might be able to help you out.
“We were doing that back in the 1990s, but the information wasn’t that good. We didn’t see health-care providers merge data with pharmacy providers.”
Advances in technology and medical science make it possible to predict with growing accuracy which employees are at high risk for serious illness in the next two years, proponents of the science say. One example: Doctors have identified a syndrome of traits from waist size to triglyceride levels that, in combination, have been found to greatly increase risk of diabetes, heart disease or stroke.
“Predicting the population is relatively easy,” said Dee W. Edington, director of the University of Michigan’s Health Management Research Center. “We can predict for individuals. That’s where the power comes.”
Edington is a pioneer in the use of so-called health risk assessments that are used to sort workers into low-, medium- and high-risk categories.
These questionnaires typically ask about smoking, seat belt use, diet and exercise, for instance. But some also request medical information and delve into mental well-being.
(END OPTIONAL TRIM)
Fernando Lopez, 40, an electrical engineer at Xerox Corp.’s research group in Palo Alto, Calif., decided to participate after his company offered discounted insurance premiums.
“At the beginning I was hesitating how much I wanted to tell them,” he said. “I gave them a chance, and in my case it’s worked out.”
He talks monthly by telephone with a coach from Harris HealthTrends Inc. who helps him stay on track toward his goals: eat better and exercise more. The healthy father of two young children has shed 25 pounds and runs 21/2 miles every other day.
“I feel more energetic,” Lopez said. “Sometimes you need a little push to do something that you know is a good idea.”
Thirty-five percent of large employers offered health risk assessments to employees enrolled in their largest health plan last year, up from 27 percent the previous year, according to Mercer Human Resource Consulting.
(EDITORS: BEGIN OPTIONAL TRIM)
Edington has tracked more than 2 million assessments over 15 years for clients such as JPMorgan Chase’s Bank One, one of the first users. Employees’ names are stripped out and the self-reported data is correlated with medical, pharmacy, short-term disability and workers’ compensation claims.
His research suggests that an average 25 percent of employers’ health-care and pharmacy costs are related to what he calls “excess risk,” which are factors that can be influenced by changes in lifestyle or medical care.
“The science is here now. Employers don’t have many options left and employees, if we don’t do something, all the costs are coming their way,” Edington said, referring to consumer-driven health plans that shift the burden of managing costs onto workers.
He foresees employees using their health risk scores to make spending decisions. His institute’s assessments predict with 83 percent accuracy which employees will end up among the 20 percent most-costly claimants, he said.
“If I’m predicted to be high cost in the next two years, I want a low-deductible plan and vice versa,” he said. “There are all kinds of ways the data can be used by individual employees if we can get all the confidentiality issues settled.”
(END OPTIONAL TRIM)
But getting employees to participate is a challenge.
When Xerox introduced health risk assessments in late 2003, 61 percent of workers filled them out. But when the company added a requirement that employees agree to be contacted at least once by telephone, participation dropped to 47 percent despite the inducement of a $200 break on their health-care premiums, Xerox benefits director Larry Becker said.
International Truck, which has been using assessments for at least a decade, gets 90 percent participation at its headquarters in Warrenville, Ill., but only about 30 percent participation at unionized plants, Bunn said.
Scott Insurance, a brokerage in Lynchburg, Va., with 180 employees, got 100 percent participation last fall when it initiated an assessment and intervention program offered by Haelan Group, said Scott’s health risk management director Dina Van Cleave. The incentive was hefty: Scott picked up the total cost of participants’ health insurance.
(EDITORS: STORY CAN END HERE)
Lewis Maltby, president of the New Jersey-based National Workrights Institute, considers incentives a form of privacy tax.
“You can say you’re offering a discount to people who fill out questionnaires,” he said. “You can also say that employers are going to charge employees more for health insurance if they won’t divulge their health and medical history.”
Yet another ethical decision is how to allocate spending using predictive models that provide more precise calculations for return on investment.
Edington offers the example of a 50-year-old smoker who is overweight and diabetic, compared with a 30-year smoker with no other health risks. The older smoker would be at high risk for serious illness within two years, he said, but the younger smoker would not be.
“We know you can spend up to $500 on the 50-year-old and still get a real return in terms of health-care costs and productivity,” but not on the younger worker, he said.
Nonetheless, Edington and others advocate spending more on helping healthy employees stay well.
“It’s a lot cheaper and easier to invest in the low-cost healthy group than not do anything and down the line have to pay for the higher-cost group,” said Dr. Wayne Burton, wellness and productivity executive at JPMorgan Chase’s Bank One.
—
(c) 2005, Chicago Tribune.
Visit the Chicago Tribune on the Internet at http://www.chicagotribune.com/
Distributed by Knight Ridder/Tribune Information Services.
—–
GRAPHIC (from KRT Graphics, 202-383-6064): DATA
AP-NY-06-07-05 0559EDT
Comments are no longer available on this story