HCR -- A Case Study in Comparative Effectiveness Research

By John Ballard

Comparative effectiveness research documents what works and what doesn't.  And thanks to the new law, there is money (or not) on the results. Joe Paduda points to a textbook example of how this process works. 

Merrill Goozner's piece on the FDA's decision to pull a stent from the market after it was shown "2 1/2 times (14.7%) more people either died or had a repeat stroke after receiving the stent than those who received drugs and counseling (5.8%)." shows how science can - and should - be brought to what is all-too-often the "art" of medicine.

The stent was approved based on a rather limited study by manufacturer Stryker, but fortunately only approved for use if it that use was as part of an evaluative study.

That study was stopped early due to the higher death/repeat stroke rate; unfortunately it appears that use of the stent may have played a role in patients dying and/or having more strokes.

The good news is the stent is, or soon will be, off the market. The bad news - outside of that delivered to the families of those who died possibly as a result of the stent - is that this is actually "news".

The reason this device was pulled from the market is because it was only approved on a limited basis by the FDA, who could pull that approval relatively easily. For devices and drugs and treatments already approved by, or not subject to approval by, the FDA (or any other regulatory authority) it is much more difficult to get them off the market. And it's impossible for Medicare to factor effectiveness into payment.

Goozner's report was equally straightforward.

The Gateway-Wingspan system, manufactured by Stryker, was approved by the Food and Drug Administration in 2005 based on �a small, less rigorous study,� according to this morning�s Washington Post. But use has been limited by a decision by the Centers for Medicare and Medicaid Services to deny reimbursement unless patients receiving the device were enrolled in a clinical trial. This �coverage with evidence development� policy began during former CMS director Mark McClellan�s tenure during the Bush administration.

Walter Koroshetz, chief of NIHDK, told the Post, �This is an example of how to do it right.� I partially agree. The government did the right thing, given existing laws and policies. But Stryker should have done the larger trial in the first place. The potential profits from the device, if it was useful, would have far exceeded the $20 million spent by the government to prove it wasn�t.

The company should be asked to reimburse the taxpayers. And this episode shows that the FDA�s device approval process should be made more stringent to require the same kind of trials for devices that it now requires of drugs � the trials that prove with statistical significance that new products do in fact improve public health.

The reader may not be aware that new drugs and devices are required to be tested for safety and effectiveness, but there is no requirement that they be compared with what may already be on the market. For research purposes the control group may be one made up of patients with no treatment at all. Those expensive new drugs you see advertised with budgets that rival that of automobiles and cosmetics are often only marginally better than the competition, or even other products newly available in generic form.