Disease management: continuous health-care improvement – includes related article on drug companies

Disease management: continuous health-care improvement – includes related article on drug companies – Cover Story

Ken Terry

Even as their health-care costs level off or decline, large employers are seeking further savings through quality improvement. Citing estimates that 30 to 40 percent of medical services may be inappropriate, some companies are pressing providers for more preventive care and better treatment of chronic diseases.

HMOs are responding with a variety of initiatives to help patients manage their ailments and stay out of the hospital. The attempts of drug companies and pharmaceutical benefit managers to participate in this process have attracted attention under the banner of “disease management.” But the concept of improving quality along disease-specific lines didn’t originate with the pharmaceutical vendors; rather, it represents the continuing evolution of managed care, from reviewing components of utilization to developing a broader perspective of what happens to patients as they move through the health-care system.

Advocates of disease management hold that a component-based approach to cost containment is bound to fail because it doesn’t look at the natural progression of a given disease. “If you try to manage the pharmacy, you get higher emergency-room bills; if you try to manage ER access, the patients come in sicker and have to be admitted to the hospital more often,” says Robert Nesse, M.D., vice chairman of the family practice division at the Rochester, Minn.-based Mayo Clinic.

Disease management, in contrast, applies principles of continuous quality improvement to the whole spectrum of care for a particular condition, including outpatient, inpatient, and ancillary services. By persuading physicians to follow practice guidelines, measuring the results, and feeding those results back to the doctors, health plans, and medical groups hope to reduce variations in care and produce better outcomes.

As logical and promising as that approach sounds, several factors hinder its development. For one, there’s the matter of cost. Proponents have little hard evidence to show that disease management saves money, and in some circumstances, it may even cost more. Also, few health plans have sufficiently sophisticated information systems to provide all the data that successful disease management requires, and major upgrades of computer hardware and software are big-ticket investments. Finally, many physicians remain skeptical about the appropriateness and effectiveness of clinical guidelines.


Most disease-management programs focus on such chronic problems as asthma, diabetes, hypertension, and low-back pain, along with preventive measures such as mammography and smoking cessation counseling. Primary care in these areas affects 80 percent of health-care costs, says Richard Bartsh, M.D., senior vice president and chief medical officer for Moline, Ill.-based John Deere Health Care. That makes such conditions a particularly attractive target for employers bent on scaling back health expenditures. But disease management’s ability to deliver big savings seems to be an article more of faith than of verifiable fact.

Consider John Deere Health Care, which covers two-thirds of Deere & Co.’s 106,000 U.S. health-care enrollees through its own IPA- and staff-model HMOs. John Deere is currently instituting one of the nation’s most advanced disease-management systems. But despite the company’s multimillion-dollar investment, and the obvious hope that health-care costs will decline as a result, there’s no specific target for savings. “We want to make sure we’re providing high-quality, efficient care,” says Bartsh. “Then we’ll see what that costs. Since there’s a lot of inefficiency in the system, we think we’ll save money over time.”

The Business Health Care Action Group, a consortium of 22 large companies with 250,000 covered lives in the Minneapolis area, has a similar outlook. “In the long run, you can reduce costs by improving health,” asserts Steven Wetzell, BHCAG’s executive director.

Dwight McNeill, manager of managed-care benefits for the Stamford, Conn.-based GTE Corp., also stresses the need for quality improvement, and he does have figures showing it saves money, at least in the short term. Using Health Plan Employer Data and Information Set (HEDIS) 2.0 performance criteria compiled for 90 of the 139 health plans that serve GTE, he ranked the plans and found that those in the top quartile had lower premiums and lower price increases last year than those ranked further down. “Quality plans cost less,” declares McNeill. “When we convert an employee and his family from indemnity to HMO, we save $1,000 a year in overall costs. Within the HMO environment, when we migrate people from average-quality to high-quality plans, we save another $500 a year.”

In the medical groups and HMOs that have tried it, disease management has scored some minor victories. For example, a Mayo Clinic guideline on urinary tract infections saved $60,000 during a six-month trial at one Mayo center that has 35,000 patient visits a year. While patient outcomes haven’t been affected, the center’s physicians have saved money by prescribing antibiotics for three days instead of 10, avoiding routine urine cultures, and having nurses triage patients over the phone.

Likewise, an asthma-management program reduced Harvard Community Health Plan’s (HCHP) hospital admission rate for pediatric asthmatics by 25 percent, and for adults by 10 percent. But HCHP has no figures on cost savings, says Larry Gottlieb, M.D., the HMO’s associate medical director of clinical quality management. While he thinks the program has saved money, he notes that improving care has also increased expenses for nurses and outreach, among other things.

That’s one of the problems with disease management: Providing better treatment for more patients with a particular condition may mean spending more, not less. For example, a recent study found that primary-care physicians, who treat 60 percent of depressed patients, often prescribe the wrong medications for depressives and don’t know how to counsel them. If the doctors followed expert guidelines, their patients’ functional status would improve, but the average cost of treating them would rise 35 percent. And that doesn’t take into account all the undiagnosed depressives who might be treated if they were sought out.

Steven Wetzell admits that, in the short term, disease management may increase costs. “But the bottom line on health care is to keep people as healthy as you can, not to avoid cost,” he says. “There are also productivity issues here: Having depressives running around the worksite isn’t a money-saver.”


Most HMOs aren’t waiting for proof that disease management saves money, however. According to a recent study of 62 managed-care organizations by the San Francisco-based Zitter Group, more than half were considering, developing, or implementing disease-management systems. Among the leading HMOs that have disease-specific quality programs are Minneapolis’ HealthPartners (which has an exclusive contract to provide care for employees of the companies in the Business Health Care Action Group), Seattle’s Group Health Cooperative of Puget Sound, the Lovelace Health System of Albuquerque, N.M., and the Brookline, Mass.-based Harvard Community Health Plan. Kaiser Permanente’s northern California division is also planning a major effort in this domain.

Much of the push for quality improvement is being driven by competitive pressures among health plans. The emergence of HEDIS and other quality measures used by employers has spurred providers’ attempts to better their performance in benchmark clinical areas. HMOs are targeting some of these conditions, such as asthma and diabetes, with disease-management programs.

These chronic conditions also happen to be among the most expensive to treat, and they typically involve large practice variations. The possibility of reducing variations–and, presumably, cutting costs in the bargain–gives HMOs another good reason to focus on such diseases. Group Health Cooperative, for example, has devised a clinical “road map” for diabetes because it accounts for 13 percent of the plan’s costs while afflicting just 3 percent of its members. Better care and patient self-management, it’s hoped, can reduce expensive complications.

The same factors are motivating many providers to improve treatment of asthma, which in 1990 cost the nation $3.6 billion in medical outlays and $2.6 billion in indirect costs, such as time lost from work.


Disease management is spreading fastest in large staff-, group-, and network-model HMOs, although some IPAs are also embracing it. The more vertically integrated an organization is, the better its capability to measure and change how patients are treated in various phases of a disease. Also, staff- and group-model plans are more likely than IPAs to have some kind of clinical information system.

But most such systems, say providers, can’t track care in the detail that disease management requires. Partly as a result, the number of plans with true disease-management systems is “still fairly sparse,” says Eran Broshy, a vice president of the Boston Consulting Group, which did an influential study on disease management for the pharmaceutical industry. “Everybody is talking about it,” he says, “but fairly few providers have converted their systems so they can do it.”

Most health plans and groups are nowhere near computerizing patient charts. If their doctors have access to electronic records at all, the information is usually derived from claims or “encounter” data. An HMO’s encounter system typically includes the diagnosis and treatment codes used in billing systems. Replete with errors, these codes can be used to identify only a few of the patients with specific conditions. Only by combining these data with abstracted chart information can providers begin to implement disease management.

Even fairly advanced systems have severe limitations. For instance, Larry Gottlieb notes that Harvard Community Health Plan’s electronic medical record doesn’t distinguish type I from type II diabetics and sometimes erroneously ascribes conditions to patients. (HCHP is building a new system to replace its present one, parts of which date to 1969.)

Most observers agree that effective disease management will require state-of-the-art electronic medical records with specially designed software. Some plans and groups–including Group Health Cooperative of Puget Sound, the Mayo Clinic, Park Nicollet in Minneapolis, and John Deere Health Care–are moving in this direction.

But developing the software and providing the hardware to thousands of doctors is a very expensive proposition. Not surprisingly, BHCAG had to back off its original demand that its providers create automated clinical systems right away. “They couldn’t justify the capital investment,” notes Steven Wetzell. “So for now, we’ve gotten them to agree to build the data base. When the hardware and the network are available, we won’t have to go back to paper records and enter them all.”


Hampered not only by the lack of good clinical data, disease management also has to overcome physicians’ suspicion of clinical guidelines. Moreover, it takes time to implement disease-management strategies. John Deere Health Care, for instance, has developed 11 programs over the past three years but has implemented only two, for asthma and hypertension.

Dwight McNeill of GTE thinks providers are exaggerating the obstacles to quality improvement. “The cynical view is that employers have caused a huge migration into managed care and that most of these plans have such large surpluses that they don’t know what to do with them,” he says. Citing one HMO that racked up large profits and promptly bought a couple of other plans, he asks, “Did that help marketplace competition? Don’t whine to me about not having a satisfactory infrastructure. You’ve got the capital to build one. It just hasn’t been a priority.”

Along with 13 other big companies, GTE has been working with 16 health plans on the Outcomes Management System (OMS) Project of the Managed Health Care Association. According to a recently released OMS study of 6,600 adult asthma patients, the use of appropriate asthma medications and peak-flow meters varied significantly across plans, patient compliance with doctors’ orders was poor, and half of the patients didn’t know how to manage their asthma. Given such findings, it’s not surprising that 40 percent of the asthmatics had missed at least one day of work because of illness or injury during the month prior to the survey.

“This project provides the first compelling evidence that outcomes data can be collected on a routine basis and at relatively low cost,” declares McNeill. He regards such disease-specific functional status surveys as a much better tool for comparing health plans than the HEDIS “process data” that many HMOs are supplying to large employers.

“I don’t care as much about whether people have received diagnostic tests or what a plan’s coronary artery bypass rate is,” he says. “That doesn’t tell me or the consumer anything. The best vehicle for combining the interests of providers and purchasers is to look at condition-specific outcomes.”

But is patient-reported functional status a reliable indicator of health-care quality? While there’s little proof to date, Michael Huber, executive director of the Minneapolis-based Health Outcomes Institute, which is running the OMS project, says, “We’re feeling more positive about it all the time.”

Functional status surveys alone cannot be used to assess HMO performance, contends Janet Corrigan, vice president for planning and development at the National Committee for Quality Assurance, which created HEDIS. That’s because many factors outside the plan’s control influence those outcomes, she says. But measurements of functional status or patient satisfaction, in conjunction with HEDIS, could indicate how well a plan is performing, she concludes.


While outcomes research is still nascent, quality experts and physician leaders agree that much can be done right now to improve the care provided by managed-care plans.

For instance, the Mayo Clinic, which has helped John Deere Health Care develop its disease-management strategies, analyzed the care received by JDHC patients with chronic obstructive pulmonary disease (COPD) in 1990-1991. It found that the largest number of their doctor visits had been to the ER, that no tests had been done on almost 40 percent of them, and that fewer than 30 percent were receiving the best medication for COPD.

How to improve on such unfavorable results? Set specific, measurable goals, says Robert Nesse of Mayo. Aim to increase referrals of COPD patients to smoking-cessation counselors, prescriptions for ipratropium, and follow-up spirometry tests. Try to increase the number of these patients’ visits to primary-care doctors and to decrease ER visits.

Donald Berwick, M.D., president of Boston’s Institute for Health Care Improvement and a leading quality expert, agrees that providers can make beneficial process changes without outcomes data. “If I want to shorten the waiting times for a visit to my radiology suite for an ambulatory X-ray, I could and should do that,” he says. “I don’t know whether it will produce better outcomes for lung cancer or pneumonia, but that doesn’t matter. My aim is to shorten waiting times and reduce costs.”

On the other hand, Berwick observes, it’s simplistic for payers to assume that they can force change simply by showing providers where their outcomes fall short. “For example, if a company’s work force is racially mixed, that employer could demand that its providers reduce the discrepancy in infant mortality between families of white and black employees. That’s a measurable outcome. But to accomplish this goal, clinicians would need tremendous amounts of information about tokolytic agents, prenatal counseling, rates of herpes infection, and all sorts of things the customer wouldn’t dream of asking for.

“It’s like the difference between a golf score and a golf game. The golf score is what the purchaser wants, and the golf game is what the provider has to work on. This idea that, ‘Oh well, we’ll just get the results measured and then they’ll know what to do’ is where the silliness comes in. The purchaser can afford to view improvement as a black box; the provider cannot,” says Berwick.


Members of the the Business Health Care Action Group in Minneapolis have gone far beyond demanding better outcomes: They’re playing an integral role in the process of change. Benefit managers in member companies are working with provider committees that are designing clinical guidelines, and BHCAG is committed to upgrading patient education and employee wellness programs.

“If providers and purchasers don’t jointly manage consumers’ expectations and educate them about their role in a changing health-care system, then we’ll be fighting an uphill battle,” says Steven Wetzell of BHCAG.

According to Tom Pavey, benefits director for American Express Financial Advisers, a BHCAG member, “We don’t expect employers to be making clinical decisions. However, many guidelines may have a workplace component, and we have to break out of thinking that health care can be delivered only in the doctor’s office.”

Some HMO executives also believe that it makes sense to tie clinical quality improvement to health promotion on the job. “Every aspect of life should reinforce the same set of messages,” says Gary Sennett, M.D., a medical staff director at Group Health Cooperative. “Since people spend a lot of time at work, we’re trying to capitalize on that. And that means a pretty active partnership with employers.”

But not all companies feel they have to help their providers do disease management. For instance, although GTE has a health promotion program, Dwight McNeill wants GTE health plans to know that the company considers it primarily their responsibility to educate patients and reach out to the covered population.

Is disease management the wave of the future? Much depends on whether cost savings materialize and can be documented. Integrated delivery systems are just starting to emerge in many parts of the country, and most providers can only dream of electronic medical records. On the positive side, many pharmaceutical companies are interested in disease management, and they could become a powerful catalyst in the development of this approach. (See page 68.)

But ultimately, says Robert Nesse, employers will decide the fate of disease management. “If they choose to reward high-value care and information regarding patient outcomes and accountability, it will become a big deal and grow dramatically. If they decide to award their contracts to someone who gives them a 20 percent discount, it won’t grow. But I’m convinced that the long-term interests of the patient lie in intelligent decisions regarding the best care and not just the cheapest you can find.”

Will Drug Companies Play a Role?

As managed care spreads through the land, establishing formularies and cost-conscious purchasing policies, major pharmaceutical companies are gearing up for disease management. While their ultimate goal is to sell more of their products, some companies say they want to build risk-sharing partnerships with health plans and medical groups. Observers believe these efforts will succeed only if the pharmaceutical firms really try to improve clinical quality.

“The drug companies have to prove that they’re not just pushing more of their stuff and that they’re truly interested in managing the case,” says Fred Spong, M.D., a San Diego-based consultant for Milliman & Robertson. “The proper approach may involve more or less medication, and it may not involve their drug.”

Right now, only a few pharmaceutical companies are actually working with health-care providers. Glaxo, for instance, is participating in an asthma program at Norwell, Mass.-based Pilgrim Health Care, and Eli Lilly and Boehringer Mannheim are co-funding physician training in the diabetes program of Seattle’s Group Health Cooperative. But, according to a recent study by Market Measures Inc., a Livingston, N.J.-based research firm, an increasing number of managed-care organizations are responding to the drug firms’ overtures. “For the first time, there’s a sense that pharmaceutical companies are serious about disease management,” says Richard Zucker, vice president of Market Measures.

What can the drug companies do for managed-care organizations? To begin with, they can supply expertise that many of the smaller health plans, in particular, are lacking. For instance, Integrated Disease Management, a subsidiary of Eli Lilly, is offering help in analyzing care processes, organizing data bases, building clinical information systems, and improving patient-education programs. It can also give customers access to case managers or subspecialists, such as diabetologists.

“The optimal use of pharmaceuticals is a very small part of disease management,” says Robert Browne, M.D., IDM’s vice president and chief scientific officer. However, he doesn’t deny that Lilly hopes IDM’s efforts will have a positive effect on its core business. One benefit, he notes, could come from building better relationships with health plans and medical groups. Also, he points out, “There are cases within certain therapeutic areas in which pharmaceuticals are being underused. In those areas, disease management may result in a net increase in the use of pharmaceuticals, although Lilly may or may not make those drugs.”

IDM also hopes to gain access to the rich data base of PCS, a pharmaceutical benefits manager (PBM) that Lilly recently acquired. “The ability of PCS to identify patients who are having problems will be very valuable,” Browne says. “Patterns of drug use tell you a lot about patients, including whether they’re doing well or not and what diseases they have.”

Eran Broshy, vice president of the Boston Consulting Group, which has a number of clients in the disease-management field, says that pharmaceutical firms can also offer health plans their expertise in information technology. “Those pharmaceutical companies that have tied up with PBMs will leverage those players’ capabilities–including information processing–to create relationships with providers,” he says.

Moreover, he points out, drug companies have already organized their marketing and much of their research and development according to disease states. “Pharmaceutical salespeople could be used in some cases to educate physicians and patients in how to adhere to a set of guidelines,” he points out.

One drug company-owned PBM moving in that direction is Montvale, N.J.-based Merck/Medco. Speaking at a Zitter Group conference in San Francisco, Medco Vice President Randy Frankel said the company is using its data base on prescriptions and patient compliance to give feedback to physicians. Medco wants to “train patients to take care of themselves,” while keeping their physicians informed of the results of patient surveys, he said. The company has launched a pilot program in diabetes and is also looking into other chronic conditions.

PBMs like Medco have been trying to feed back data on patient outcomes and compliance to doctors for some time, says Richard Osborne, a principal in Cambridge, Mass.-based EDS Management Consulting. But he doesn’t believe they have any proof that their efforts have paid off.

If PBMs become involved in disease management, what will prevent them from either favoring the cheapest drug or one made by the pharmaceutical company that owns the PBM? Eran Broshy replies that it’s in a PBM’s interest to use its parent’s drugs only if it makes clinical sense. “They can’t put inappropriate medicine in place of a competitor’s product,” he says. If they did, says Dee Kemnitz, vice president of employee benefits for the Minneapolis-based Carlson Cos., they’d lose their credibility, because “they wouldn’t get the outcomes.”

Other employers and providers believe that the best way to guarantee the honesty of pharmaceutical companies and PBMs is to make them share risk. “As our care systems begin to focus on keeping people healthy, the drug companies havE; to change their thinking and move from detailing to disease management,” declares Steven Wetzell, executive director of Minneapolis’ Business Health Care Action Group. “The healthier they keep a population, the more money they make.”

Various risk-sharing angles have been proposed. For instance, says Robert Nesse, M.D., vice chairman of the Mayo Clinic’s family practice division, a PBM might propose to help reduce a group’s $1 million annual bill for pediatric asthma by working with it on patient education and tracking patient compliance. In that case, the PBM might receive 10 percent of the group’s savings–or eat some of its loss.

Robert Browne says this is exactly how IDM works. “As we help our customers optimize their clinical processes, the net result is to save money, and we share in those savings,” he says. “They might be direct, easy-to-measure bottom line savings, or a program that has perceived value and will cut costs in the long run. If what we do increases the cost, the agreement may say we’re responsible for part of that.”

COPYRIGHT 1995 A Thomson Healthcare Company

COPYRIGHT 2004 Gale Group