On the Costs and Benefits of Health Care Regulation

The latest issue of The Economist has both an editorial and then a one-page article on America's health-care system, and the costs and benefits of the tangled maze of regulations we have. The editorial begins by noting that Milton Friedman argued that the FDA was unreformable and hence should be abolished. While I agree with that, those at the Economist do not. But their editorial does go on to make some interesting and sensible reform suggestions:

The starting point is that the FDA and its counterparts across the world need to move from a risk-obsessed, "one size fits all" approach to a more flexible system that considers the risks and benefits of new therapies. Rather than asking drugs to undergo many years of costly trials in the vain pursuit of medicines that are safe for all in all circumstances, regulators should allow speedier conditional approvals. This is especially true for the growing number of targeted therapies, such as several excellent new cancer treatments, made possible by advances in genomics--a science that identifies which genetic groups will benefit from a drug.

If initial tests and sophisticated computer modelling show promise, innovative new drugs should proceed to small human trials. Ultimately, the drug could be approved for use by a wider part of the general population deemed (through genetic testing and other screening methods) to be at relatively low risk or more likely to benefit from the therapy.

Safeguards can counterbalance this relaxation. Faster approval of new drugs in humans should be matched by more rigorous post-launch testing and surveillance. At the moment, the FDA does not have the money or authority to do this properly. Scores of post-launch studies of drugs safety have been requested by regulators, but remain undone or ignored by firms. And yet there is reason to think such post-launch surveillance, if bolstered by the use of electronic records and "data mining" techniques, could save many lives. When Vioxx, a blockbuster pain remedy made by Merck, turned out to be dangerous for some patients, private health-management organisations with excellent electronic patient records spotted the problem months before the FDA did. The second safeguard is for all drugs trials done anywhere in the world--failed or successful--to be made public and the data published online. Consumers would have the information they need to choose whether to take a drug, weighing benefit against risk. Companies would benefit from a cheaper and faster approvals process, and a lower risk of litigation.

By cutting the costs of approval, lighter regulation should help move the industry away from blockbusters towards niche products. A reformed FDA might thus speed the arrival of the long-promised age of personalised medicine. Even if it survived, Friedman would surely approve.

The article in this issue then discusses three recent studies that give real evidence that Friedman's view of the FDA is correct: namely that it does far more harm than good. And we are talking about real physical harm here folks: people's lives would be saved if the FDA didn't do the things it did. See the graph in this article that indicates the total costs and benefits for five major areas of regulation of the US health care system: medical torts, the FDA, insurance regulation, and the certification of health professionals, and health facilities. It even indicates the components of the costs that come from State vs. Federal regulations. In all five of these areas, the costs outweigh the benefits. While medical tort and insurance regulation have greater total costs, this graph indicates that the relative costs vs. benefits are worst for the FDA.

Citing a second reference (a forthcoming paper by Michael Cannon of the Cato Institute), the article notes the important problem with the FDA:

Citing the best evidence to date on the costs and benefits of FDA regulation, Mr Cannon argues that the agency "is too slow and demands too much testing", ultimately harming consumers. He points out that drugs regulators can make two broad types of errors. First, they might approve a drug too quickly, only to find out after its launch that it is dangerous or even deadly. Second, they could delay the launch of a highly innovative drug by demanding onerous or unnecessary trials and thereby deny many needy patients a new therapy.

Proper regulation requires balancing these two risks, but the pitch may be queered by bureaucratic self-interest. If the regulator allows even one drug to slip through the approval process that later proves harmful to some people some of the time, a hue and cry is sure to follow. Look no further than the recent public backlash against the FDA after several deaths were linked to Vioxx, a blockbuster pain remedy made by Merck.

And yet the second (and probably bigger) risk of leaving people untreated because of restrictions on drugs rarely gets the regulators into trouble. As Mr Cannon puts it, "no FDA official has ever been fired or faced a congressional inquiry for delaying the approval of a promising new drug, however unjustified the delay." What is more, he speculates, big drug firms may quietly acquiesce to this burdensome red tape because it acts as a barrier to entry against newcomers without the cash or lobbying power to navigate the FDA.

That is a key point. I'd love to see FDA officials grilled by members in Congress with questions like "Because you took five years to approve drug XYZ, it is estimated that 10,000 people died that would have otherwise lived. How can you defend your slow bureaucracy?" That would be precious... but I'm not holding my breath.

I hope that the FDA will be reformed for the better, though I fear it won't be or that changes will only make the situation worse.