A Better Way to do Risk/Benefit? (Part 2)

As we discussed yesterday, FDA is taking a small step towards a potentially revolutionary goal: developing a more formal, systematic approach to making judgments about the risk/benefit ratio of prescription drugs (and other regulated products). And we highlighted all the reasons why many in industry and FDA think this is a terrific idea.

Be careful what you wish for.

We don’t think it is very likely that any quantitative model can succeed in drawing a bright line between drugs that are “unsafe” and drugs that are “safe enough." But if it does, such a tool would almost certainly be applied just as readily to determining whether a drug is or is not “worth it.”

Why? Because any conceivable approach is almost certain to rely on economic analyses, using concepts like quality-adjusted life-years or other metrics that attempt to assign common values to different types of health effects.

That at least was FDA’s conclusion from its initial work on modeling earlier this decade. And it is hard to imagine any other way to compare benefits like allergy relief against risks like heart attacks.

Neither FDA nor industry is eager to see economic analysis embedded too deeply in the regulatory process. We, at least, haven’t heard anyone in industry ready to argue that FDA should approve drugs based on a cost-effectiveness standard—because that is in effect what we are talking about here.

Of course, the US may be moving in that direction anyway. Certainly both Presidential candidates want to see more emphasis on the value of medical interventions—and both share the traditional emphasis on pharmaceuticals as key focal point.

So in fact there may be a case to make that industry would be better served by having this discussion in the context of the regulatory process it already relies on. It would certainly be preferable to having it in the context of a Breakthrough Drug Pricing Board of the type proposed during the Clinton health care reform era.

But there is a bigger issue: we just don’t think even the most robust, scientific risk/benefit analysis model will make regulatory decisions fundamentally easier or more predictable. When it comes to issues involving health, and especially the risk of death, the public consistently refuses to set aside emotion in favor of arguments based on even the soundest economic principles.

Consider the response to a proposal by the National Cancer Institute more than a decade ago to stop recommending routine mammograms for women aged 40-49.

NCI’s position at the time reflected a simple statement of the evidence – there was no demonstration that mammography led to any improved outcomes in women that young. But the agency came under intense pressure to reverse its position and thus restore the presumption that routine mammography would be covered in that age group.

Similar reactions have greeted analyses questioning the cost-effectiveness of interventions to protect the blood supply from HIV or other unknown pathogens, efforts to address the risk of “mad cow” disease in livestock, or even the assignment of benefits to victims of the World Trade Center attack in 2001 based on projected lifetime earnings.

In other words, we are a long way away from a world in which the public will accept regulatory decisions based on purely “scientific” measurements of risk and benefit.

On the other hand, as industry is all too well aware, there is no shortage of public support for arguments that drugs are overused or overpriced.

So, while a quantitative methodology to balance risks and benefits certainly has great promise for industry, it also has its perils. And it may be that both FDA and industry have to think carefully about the risks and benefits of going too far down the path of developing the perfect risk/benefit tool.