Over at CEPR, Dean Baker has a proposal for publicly funded drug trials. It' a fairly obvious idea when you think about it. A tremendous proportion of clinical trials are currently carried out by producers of the drug or treatment in question. Their incentives, then, are pretty obvious. Make the drugs seem like they work, so they can start selling them. And, sure enough, we've had many a study evaluating trials by funder, and it turns out that those financed by for-profit organizations return more positive results than those funded by non-profits. Additionally, we've seen no end of attempts to conceal negative results or harmful side effects. The poster child for this is Vioxx, but it's pretty common behavior as pharmaceutical companies bury bad data and rerun trials hoping for a better result. Dean proposes an easy fix to this: Fund trials publicly. If you're worried about the stench of government, have the Feds pay private contractors to run the tests. The benefits would be enormous. First and foremost, the conflicts of interest would largely evaporate, as would the capacity to hide bad results. Moreover, the data would be freed from the proprietary grip of the drug corporations -- we'd be able to effectively evaluate pharmaceutical effectiveness, and researchers would be able to use the results to further related work. That could lead to scientific advances on the one hand, and massive cost savings on the other, as we began to learn what was and was not worth covering. The political problems with this are, of course, significant, as drug companies like being able to hoodwink us into paying for bad drugs. In a more rational market, where insurers actually wanted to incentivize cost effective care, they'd be fighting on behalf of this change, as they need better information in order to figure out how to do that. But this is not a very rational market.
