Why Medicare Wants More Drug Studies

Large, long-term trials can better show which treatments are most effective, but hurdles include high costs and drugmakers' resistance

By Amy Tsao

The medical profession has a costly, little-known secret: Many treatments don't work all that well for many -- even most -- people. And doctors often don't know what are the best therapies for any given disease, especially when faced with competing classes of drugs for, say, high blood pressure. As a result, many people aren't getting the best possible treatment -- and the resulting costs to the health-care system are estimated to be in the billions.

The solution is conceptually simple -- but incredibly expensive: Fill the gap by doing large clinical trials that follow patients' progress for as long as possible.


  Drugmakers usually won't pay for such studies because of the chance that results unfavorable to their product may emerge. Government agencies have tried to step into the breach. The National Institutes of Health paid for a $125 million comparison of drugs for hypertension, with the surprising finding that older, cheap diuretics were as effective as newer, more expensive medicines.

"We need to be asking the kinds of questions that the industry is not likely to," says Larry Friedman, acting deputy director for the National Heart, Lung & Blood Institute.

But this scratches only the surface of what's needed. Mark McClellan, the new head of the federal Centers of Medicare & Medicaid Services (CMS), has been trying to find broader ways of collecting the kind of evidence needed for long-term evaluations of drug efficiency. He says his aim is to "develop better evidence so doctors and patients can get more out of what [Americans] spend on medical coverage."

CMS makes the reimbursement decisions for the massive Medicare and Medicaid insurance programs. So if the agency even hints at a change in the kind or amount of drug-effectiveness evidence it wants in order to start or maintain the treatment's coverage, the pharmaceutical and medical-device industries will pay close attention. And CMS seems serious. "We're exploring applying this approach to virtually every coverage decision we have under review," says Sean Tunis, the agency's chief clinical officer.


  It's a controversial idea that could have long-term repercussions on sales of particular drugs. Recognizing the added risk that gathering more evidence can pose, drug outfits and medical-device makers are already fighting it.

Furthermore, such trials wouldn't come cheap, and it's not clear who'll fund them. "This is something everyone in the health-care system is interested in," says Walter Moore, vice-president for government affairs at Genentech (DNA ), "but no one can figure out a way to pay for it."

CMS will surely have to be creative in how trials are paid for. Most likely it will share the costs of gathering better evidence with industry associations and government agencies. Instead of costly studies like the NIH's hypertension-drug comparison, it will favor setting up cheaper methods of collecting information, such as patient registries, which cost roughly $1 million to $2 million a year to maintain. Although these data aren't well controlled, they can be illuminating, especially when they involve a large swath of patients.


  So far, CMS has been easing companies into the possibility of conducting more trials. On Sept. 28, CMS said it plans to expand coverage of implantable cardiac defibrillators, which cost as much as $30,000 each, to a broad range of patients with heartbeat abnormalities. As part of the coverage expansion, CMS will work with manufacturers like Medtronic (MDT ), Guidant (GDT ), and St. Jude Medical (SJM ) to develop a registry to track patients' experience with the devices. Over time, the database should make clearer who benefits most from the defibrillators.

In mid-September, the agency also said it would reimburse hospitals for the use of positron emission tomography (PET) scanning in diagnosing Alzheimer's disease. As part of that decision, the agency said it will support a clinical trial to evaluate whether the technology actually helps patient care. CMS will pay the $1,500 cost of each scan, while an industry association is likely to pay for the data-collection costs.

Also on the line are the new cancer drugs Erbitux and Avastin. Studies showed they worked in a minority of colon-cancer patients. Now, there's a clamor to use them for other cancers, which raises huge questions: Do they work, and are they worth the monthly cost of as much as $17,000? Should Medicare pay?


  To find out, CMS will work with the National Cancer Institute (NCI) to develop better evidence about drug efficiency, says McClellan. CMS would likely pay for treatment costs, while NCI will probably pay for the cost of analysis and data gathering. "CMS wants to know what [uses] are appropriate for Avastin before off-label use jumps," says Moore of Genentech, which makes Avastin. "That's an appropriate response."

Still, the gaps in evidence are many and wide. Under McClellan, CMS will be active in pursuing answers, but it's taking on an enormous task.

Tsao is a reporter for BusinessWeek Online in New York

Edited by Thane Peterson

Before it's here, it's on the Bloomberg Terminal.