Both the New York Times and the Wall Street Journal carried stories today on Medicare's expansion of the number of drug-listing compendia that can now be used to justify reimbursement for the off-label use of anti-cancer drugs. This expansion, which GoozNews covered last summer (see posts here, here, and here), will sharply increase Medicare spending on anti-cancer drugs of questionable medical value.

It has also provided drug companies with an alternative system for getting reimbursed that won't require their going to the Food and Drug Administration to prove that the regimens listed in the compendia actually benefit patients.

The effects of this new system were understated in the articles. The Times estimated that the higher spending by Medicare will come on top of the $2.4 billion the senior citizen health care program spent on cancer drugs in 2007. But according to this 2006 testimony by Center for Medicare and Medicaid Services official Herb Kuhn, Medicare spent about $10 billion on "Part B" drugs in 2005 (these are drugs administered in physicians' offices, which includes most cancer drugs), and about half of that went to oncologists.

The Times also buried near the end of its story (and the Journal ignored) revelations about the most significant conflict-of-interest affecting how drugs get off-label indications listed in the compendia. It involves the oldest and traditionally most trusted of the compendia -- the American Society of Health-System Pharmacists' drug information formulary or AHFS-DI.

In late 2007, the cash-strapped trade group for in-hospital pharmacists hired a misnamed non-profit called the Foundation for Evidence-Based Medicine to serve as a go-between for drug companies that wanted to get off-label uses of their drugs into the compendium. If companies gave the Foundation $50,000, it in turn would provide AHFS-DI with the resources to conduct a review.

AHFS-DI had come under fire from oncologists and drug companies for being slow to approve off-label uses. Indeed, that was the reason offered by CMS for wanting to expand the number of allowable compendia in the first place.

Gerald K. McEvoy, who runs AHFS-DI, told the Times that the foundation "insulates the guide’s staff from industry pressure" and that "fewer than one-third of the reviews under the new arrangement have resulted in a positive recommendation in the compendium."

That statement is disingenuous. You don't need to be "recommended" to get reimbursement for an off-label use. Medicare's regional contractors, who make payment decisions in the absence of national coverage decisions, have discretion to pay for drugs even when the evidence is equivocal. When presented with payment requests from community oncologists, most do. And virtually all the most recent listings in AHFS-DI under the new system (see the most recent listings here) received at least an equivocal rating.

Moreover, McEvoy's statement ignores the most pernicious aspect of the system. A pay-to-play listing system -- just like the Prescription Drug User Fee Act at the FDA -- allows the drug industry to determine what off-label uses will get evaluated. Think about the poor academic researcher who conducts and publishes a clinical trial that shows an innovative use for a generic chemotherapy drug that has been around since the 1950s. Who will pay the $50,000 to AHFS-DI so that use gets listed in the compendium?

Beyond the conflicts of interest now driving both oncologists decisions about and Medicare's payment for anti-cancer drugs, the more serious long-term issue in this government-encouraged expansion of off-label use revolves around Medicare's refusal to collect outcomes data. Again, the issue was addressed in a single easy-to-miss paragraph in the Times story that is worth highlighting:

“We have very little faith that those indications that make it into the compendia are safe, let alone effective,” said Dr. Allan M. Korn, the chief medical officer for the Blue Cross and Blue Shield Association, who added that Medicare should cover off-label drugs only if the results of their use are carefully tracked afterward. There is no such requirement in the new Medicare guidelines.

In the stimulus bill now moving through Congress, both houses have set aside more than $20 billion to encourage physicians to adopt electronic medical record keeping. It includes hefty tax incentives for physician offices and hospitals to adopt the new systems. Doesn't it make sense to also require that physicians who get those incentives, including oncologists, collect outcomes data from the off-label use of drugs and report that information, properly protected for patient privacy, to paying agencies? How else are we going to learn if all the "hope" that taxpayers financed for dying cancer patients actually turned out to have helped?

The one thing we can be sure of is that the less-than-rigorous clinical trials that lead to compendia listings, often evaluated by compendia committees that have close ties to the very companies that make and sell the drugs, do not amount to evidence-based medicine, no matter how many foundations with that name the compendia-writing organizations set up.