Why evidence needs to drive approval

Yesterday I focused on the problems that evidence-based medicine has getting past politicians.

Now let's talk about why that's important.

Our text comes from Dr. Joshua Fenton at UC-Davis (right). He's out with a paper showing that, once any therapy is approved for reimbursement its use skyrockets, whether or not the therapy is useful.

Fenton looked specifically at computer aided detection (CAD) systems aimed at helping interpret images for breast cancer screens. Only 1 in 20 cases used it when Medicare first approved paying for its use, but two years later it was being used in nearly 25% of cases.

CAD, it turned out, had only marginal utility when compared with biopsies, and the extra mammograms generated by CAD in 2003 alone came to $19.5 million, Fenton writes.

This happened despite the fact that "systemic reviews point to uncertainty regarding whether CAD has a clinically important impact on key breast cancer outcomes."

Fenton's paper, in the Archives of Internal Medicine, points to a crucial distinction between American and European health systems.

In Europe payment approval is granted only when a treatment is deemed cost-effective. In the U.S. the market is supposed to decide what to  buy.

But while that sounds great in theory, in practice an American approval for payment can have the same impact as a European one. What regulators think is a yellow light gets treated as a green one, because once the decision is made to pay the cash spigot opens.

Doctors will use what they can get paid for using. Once the decision to pay is made the horse is out of the barn, and it can't be put back in unless the treatment in question is shown to be dangerous.

Maybe, if we want cost-effective medicine, the evidence needs to be a little stronger before the cash is allowed to flow. Individuals may pay for newly-approved but unproven therapies. Why should I be made to?

This post was originally published on Smartplanet.com