Cost-effectiveness research pinpoints best values for limited health care dollars—and the results may surprise you
An interview with Harvard School of Public Health’s Milton Weinstein offers some revealing insights into how the U.S. health care system could save money by focusing on the cost per year of healthy life that each medical intervention provides. Not all new technology is too costly, he says—nor is every prevention strategy a money saver.
Weinstein, an expert on cost-effectiveness in medicine, is the Henry J. Kaiser Professor of Health Policy and Management at the Harvard School of Public Health (HSPH) and professor of medicine at Harvard Medical School. He spoke with Review guest editor Madeline Drexler.
Annual pap smears or dialysis?
Q: Why did so many people equate cost containment in health care, and assessing the costs and benefits of medical technology, with “death panels”?
A: Because we don’t like to have government—we don’t like to have anybody—make decisions for us. We don’t mind using markets to ration things. If the price of a bottle of wine is too high, then we’ll buy a different bottle of wine. But if a big sign says the U.S. Department of Agriculture has determined that you can’t have prime rib because it’s too expensive, people don’t like it.
Q: How can studies by you and others in cost-effectiveness research help answer the question of how we might pay for universal health care coverage?
A: Cost-effectiveness looks at technologies and drugs and treatments through an economic lens. How much do they cost? What do they cost compared to alternatives? And not only what do they cost, but is it worth the cost?
For example, we developed a concept called the quality-adjusted life year, or QALY. It reflects how many years of high-quality life a patient gains with a particular intervention. Another number that we use to measure value is the cost-effectiveness ratio. Basically, it tells us the “price” of buying more healthy years with a new treatment compared with the standard treatment, and whether it’s a good value.
Q: On that scale, what dollar amount is considered a good value—or cost-effective?
A: The World Health Organization [WHO] has a rule of thumb: Three times per-person income per quality-adjusted life year gained is a cost-effective intervention. In this country, per-person income is about $40,000, so an intervention that costs less than $120,000 per quality-adjusted life year would be considered cost-effective according to the WHO rule. David Cutler, the Harvard economist, has suggested $100,000 as a reasonable value.
Here are some examples. If a doctor prescribes a beta-blocker for a high-risk patient after a heart attack, it costs about $5,000 to buy that person one quality-adjusted life year. If a doctor gives a patient with HIV combination antiretroviral therapy, it costs $20,000 to buy one quality-adjusted life year. Dialysis for end-stage kidney failure costs $50,000 to $60,000 per quality-adjusted life year, which is still a good value in this country.
Q: Are today’s new, expensive treatments usually bad values?
A: Not necessarily. Some expensive breakthroughs not only bring better health outcomes, but are well worth the money. One surprising example is the implantable cardioverter defibrillator, which uses electrical shocks to restore normal heart rhythm. Its cost-effectiveness ratio compares favorably to dialysis for end-stage renal disease—which we accept as being worth the money.
Another example is a new class of drugs for breast cancer, called aromatase inhibitors. A colleague of mine was at a clinical meeting where a well-known cancer specialist said these drugs will never catch on, because they’re too expensive—costing more than twice as much as the standard treatment. Well, it turns out that the cost-effectiveness ratio was on the order of $20,000 per quality-adjusted life year. It’s an expensive drug, but the benefits are dramatic, mostly in longevity.
Q: What are examples of routine interventions that are poor investments?
A: The annual Pap smear. The cost-effectiveness of screening every year compared to screening every two years is almost a million dollars per quality-adjusted life year. It’s not because it costs a million dollars to do a Pap smear every year. It’s because the gain in per-person life expectancy is on the order of hours to days. By doing a Pap smear every year on every woman, you only catch a few treatable cervical lesions that you would have missed if you did it every other year, but the extra cost of doing this for every woman is much higher. That’s not to say it’s not worth doing Pap smears. Doing a Pap smear once every four years is extremely cost-effective. Doing it every three years instead of every four is still cost-effective. Every two years instead of every three years starts to get less cost-effective than the implantable cardioverter defibrillators I was talking about. And screening every year instead of every two costs about $800,000 per life year gained compared to every two years.
That’s why the standard of care is gradually moving toward less frequent screening. If you get three consecutive normal Pap smears, it’s OK to start doing them less often. If vaccination against the virus that causes cervical cancer—human papillomavirus—catches on, then guidelines may well shift toward even less frequent screening.
Q: How much money could be saved if we thoroughly analyzed the cost-effectiveness of medical care?
A: There are wide variations in how often doctors order tests, prescribe medicines, do surgeries—not just in different parts of the country, but in hospitals that are right next to each other. One place may do many, many times more procedures of a particular kind than the place next door.
And if you look across regions of the country or across hospitals or states, you often see negative relationships between expenditures and outcomes: areas or states or hospitals that spend more do worse by their patients.
One interpretation is that if we could make the high-spending/poor-performance hospitals or regions or service areas more like the lower-spending/better-outcome ones, we could save money and improve health at the same time. Some people take that to mean there’s waste in the system. But the evidence says that we already may have cut most of the waste.
Q: So what’s a better explanation for these gaps in spending and performance?
A: The low-cost areas are doing things that the high-cost areas aren’t. In other words, the low-cost areas are using more cost-effective services: counseling to quit cigarette smoking, colonoscopies, giving beta-blockers to patients after heart attacks. These are well-established interventions that are effective and also are cost-effective. But they’re underutilized.
Q: What about the high-spending systems? What are some of their overused practices that are not cost-effective?
A: Intensive care unit treatment for patients with several fatal conditions, extra diagnostic tests such as MRI, CT scans, and PET scans. They’re expensive, and for many patients who don’t have clear indications of a disease, you get teeny-tiny gains. Sometimes you’re talking about cost-effectiveness ratios of millions of dollars per quality-adjusted life year. Many of the same tests are cost-effective for the right patients, but very cost-ineffective for the wrong patients.
If you do more of those expensive things that have marginal value and less of the cost-effective things that have proven value, then you get places that spend more and get worse outcomes.
Q: How do other nations handle this problem?
A: Most countries of the developed world use cost-effectiveness analysis to form policy around their national health insurance plans. We don’t have a national insurance plan, but we do have Medicare, which is national health insurance for people over 65. Yet Medicare doesn’t look at cost.
Q: What’s at stake if we don’t have a national discussion about the costs of medical technology?
A: Costs will keep going up. People will keep demanding costly new procedures. More and more people will have inadequate care. From a public policy viewpoint, we could end up with more disparities in this country than we already have—which is the worst in the developed world.
Q: In other words, rationing?
A: Yes. The biggest way we ration is by cutting people out of care. When 15 percent of people in this country have no health insurance, that’s rationing.
Q: If you were America’s medical technology assessment czar, with an unlimited budget and staff, what would you do to make this a rational, transparent system?
A: Within a market-based system, we can create incentives to use more cost-effective medical care. On the patient side, we currently have tiered co-payments for pharmacy purchases. They could be linked to cost-effectiveness. Patients could be required to pay up to a set amount per year, based on their income, for medicines that are not cost-effective. For an antihypertensive medicine that’s cost-effective, you waive the co-pay. You can also reimburse physicians based on cost-effectiveness.
If I were the czar and I had the ear of the president, I would urge him or her to have fireside chats. I’d say: Let’s talk turkey. Let’s be candid about how much of our health care dollar is going to interventions that offer benefits on the order of only days or hours of improved health. Some of these interventions cost a lot.
No president has talked about this, ever. They dance around it. They talk about cost savings and prevention and waste.
Q: Why can’t they talk about it?
A: People don’t want to think about it. They think they can have their cake and eat it too.
It’s amazing how uninformed people are. “I want the best available medical care regardless of cost”—90 percent of people agree with that. “I think that health care is too expensive”—90 percent of people agree with that. “I think health care should be available for everyone”—90 percent of people agree withthat. You can’t have it all.
The price of health
How do economists calculate value for money when it comes to delivering health care?
One way is to measure health improvement in terms of the “quality-adjusted life year,” or QALY. This number reflects how many years of life are gained as a result of an intervention, on average, per patient, per episode—and weights the extra years of life by how patients subjectively describe the quality of those years.
Another number used to measure value—the cost-effectiveness ratio—is the net dollar increase in the cost of health care compared to the standard treatment, divided by the net gain in health. Effectiveness and cost are always comparative, because one treatment or procedure is always compared to another.
Cost-effectiveness calculations yield a number on a continuous scale, ranging from a very low number of dollars to gain a year of life to a very high number of dollars to gain a year of life. An intervention that costs $100,000–120,000 or less per quality-adjusted life year is considered cost-effective.
Most medical treatments lack evidence that they are effective
More than half of the medical treatments delivered today lack clear evidence that they work, according to the Institute of Medicine (IOM). To remedy the situation, the U.S. Congress, in the American Recovery and Reinvestment Act of 2009, set aside $1.1 billion to jump-start research on which interventions are and are not worthwhile.
In June 2009, the IOM, part of the National Academy of Sciences, issued a report that lists 100 areas where popular medical interventions need to be rigorously compared, head-to-head. Top candidates for comparison are treatments for:
• Atrial fibrillation (the most common form of abnormal heart rhythm)—comparing surgery, catheter ablation, and drug therapy.
• Managing prostate cancer that has not spread beyond the prostate gland—comparing watch-and-wait, removal of the gland, and radiation therapy. Such studies would compare survival, recurrence, side effects, quality of life, and costs.
• Low-back pain.
• Reducing infant mortality and preterm births among African American women—comparing prenatal care, nutrition counseling, smoking cessation, and substance abuse treatment.
• Preventing falls in older adults—comparing exercise and balance training versus clinical treatments.
A tale of two cities
In Texas, medical care is cheaper—and patients fare better—in El Paso than in McAllen. What difference does 800 miles make?
In 2006, per capita Medicare expenditures in McAllen, Texas, hovered around $15,000 per enrollee. In El Paso, 800 miles away, the figure was half as much. What’s behind the discrepancy? “Compared with patients in El Paso and nationwide, patients in McAllen got more of pretty much everything—more diagnostic testing, more hospital treatment, more surgery, more home care,” writes Atul Gawande, associate professor in the Department of Health Policy and Management, in the June 1, 2009, issue of The New Yorker. “The primary cause of McAllen’s extreme costs was, very simply, the across-the-board overuse of medicine.”
In “The Cost Conundrum,” which was quickly touted as required reading in the Obama White House, Gawande describes McAllen as “the most expensive town in the most expensive country for health care in the world.” But his story isn’t just about irrationally lavish medical treatment. McAllen’s five largest hospitals also perform more poorly, on average, than El Paso’s.
This confirms a large body of research from Dartmouth Medical School, suggesting that patients in high-cost areas often get more expensive treatments of marginal value but less of what actually made them better. One study, for example, found that patients in high-cost areas were less likely to receive modestly priced preventive services, such as flu and pneumonia vaccines, faced longer waits at doctor and emergency room visits, and were less likely to have a primary-care physician. According to Gawande, “They got more of the stuff that costs more, but not more of what they needed.”
Gawande’s prescription for change? Emulate models such as the Mayo Clinic, which is among the highest-quality, lowest-cost health care systems in the nation. The clinic pools all the money doctors and the hospital system receive and pays everyone a salary, so that physicians aren’t tempted to pad their own incomes by ordering unnecessary procedures. It also carefully coordinates patient care, with a sprawling team of medical personnel working in sync with one another.
Gawande calls not only for comparative effectiveness research on specific treatments, but also for studies of what makes the best health care systems successful.
“I’m fascinated by the positive deviants of the world—the El Pasos that outdo the McAllens. They have learned something. And in fact, there are numerous communities across the country with lower-cost and higher-quality results,” he observed recently.
“We need local medical leadership to acknowledge that we as clinicians are slowly bankrupting the country—and that we have the ability and responsibility to work on our costly problems of overtreatment, undertreatment, and mistreatment.” Otherwise, expenses will continue to skyrocket and quality of care will remain uneven. As Gawande writes in “The Cost Conundrum,” “[W]e are witnessing a battle for the soul of American medicine.”
Madeline Drexler is guest editor of this issue of the Review
Photo: Christopher Thomas/Getty Images