Jeffrey Yasskin’s blog

2/4/2005

Reforming the FDA

Filed under: Health — Jeffrey Yasskin @ 4:24 pm

The Food and Drug Administration regulates drugs wrong. They have a hugely long and expensive testing process which terminates in a simple yes-or-no decision about whether that particular drug can be sold. This delays the use of many life-saving drugs and drives up their prices when they are finally released. There is a better way.

We need a way to summarize three pieces of information. Drug tests reveal how safe a drug is to take and how effective it is. Each test has a margin of error which indicates how reliable it is. The margin of error, risk of harmful side-effects, and chance that the drug actually works should be summarized in a single number each, which needs to be prominently displayed on every advertisement, piece of promotional material, prescription, bottle, and package of the drug. The goal is to make drug safety comprehensible to ordinary consumers and non-research doctors. In order to help people understand the numbers, the FDA should publish its opinion of safe ranges for these numbers. It’s all right if the safe ranges change over time, but calculation of the numbers themselves should not.

Once this is done, the FDA should stop restricting what drugs are sold and focus its efforts on assigning these three numbers. This has three advantages. First, it decreases government regulation and allows the market to decide how much risk it’s willing to take and how much it’s willing to pay for less risk. Second, drugs start making money sooner, which makes them cheaper to develop, which gets more drugs on the market, which makes them cheaper and improves everyone’s health. Third, the desperately ill can choose for themselves to buy a new miracle drug sooner even though it hasn’t been fully tested.

As a drug is on the market longer, more research will become available, which sometimes will conclude that the drug is more dangerous than originally thought. This is not the drug company’s fault. Liability laws need to be reformed to make it very clear that a drug company is not liable for information it didn’t have. As long as the company promptly gives any new research to the FDA, who then updates the drug’s safety ratings, the company should be immune from lawsuits. Of course, if the company hides damaging research, it should be vulnerable to extremely harsh civil and criminal penalties.

I don’t know the best way to summarize all of the research about a given drug into three numbers. Maybe we need four, or can make do with two. I think that a mechanical formula should be used so that there are no human decision-makers who can be bought. But statistics is such a complicated field that finding this formula may be impossible. If you have any ideas, I’d be interested in seeing them in the comments.

Powered by WordPress