How Should the FDA Incentivize?

The FDA often wants manufacturers to provide additional studies such as for pediatric uses or for testing of off-label uses of already approved drugs.  How should the FDA incentivize these studies?  Long-time reader Steve (who has good reason to know and thus shall otherwise remain anonymous) writes:

I was reading an article about pediatric drug testing and the BPCA, and I had an epiphany–the people at the 
table don’t have the incentives necessary to solve the problem.

…possible solutions to the problem of limited pediatric testing appear to boil down to: 1) Modify the reward (primarily through exclusivity); 2) 
Give out grants; and 3) Force studies through a government mandate. 
These solutions reflect the interests of the three groups sitting at 
the bargaining table, i.e., 1) Big pharma, 2) Academics, and 3) 
Bureaucrats. What is totally missing is the idea that incentives can be created on both the risk and reward side of the equation. … For example, if the FDA fast-
tracked NDAs with pediatric data, and guaranteed a decision in 90 days, they could, with minimal cost, cause a major shift in incentives. 

    …Any thoughts on how the situation can be improved?

The FDA significantly raises the costs of creating new drugs – there are some benefits in better safety and efficacy but I think the current system results in too much drug lag and drug loss.  I would cut back on FDA regulation considerably but I am not against more government-financed studies of safety and efficacy.  Once a drug is on the market and especially when it is off-patent, knowledge about the drug is a public good and thus often underprovided.  I would thus reduce the FDA’s control over drug choice but increase the budget for drug information e.g. through NIH financed studies like the Women’s Health Initiative which shockingly showed that then widely used homorone replacement therapy increased not decreased coronary disease.

Readers?