The author ignores two obvious, extremely high leverage changes that would solve the problems in his model.
First, on the revenue side, extend patents on pharmaceuticals as long as possible. His model assumes a very narrow window of payback; quadrupling or quintupling that window would make many more drugs feasible to investigate.
Second, to use the author’s word, “de-risk” all trials. The best way to do this would be to end the FDA. But an intermediate step would be to eliminate the requirement to show efficacy, only retaining the requirement to show safety. Safety is orders of magnitude cheaper to prove.
Trials could be conducted much more cost effectively with these two changes. There would be more time to form patient and investigator networks, the structure of trials would be less complex, many more drugs could be tested, and many more drugs would be profitable.
Extending patents wouldn't work that well in his model - as you can see from the graphs it is already discounted crazily (I'd argue for a 6% discount rate, not 8, but it hadly changes the calculus). Besides market exclusivity can often be extended by adding indications and pediatric studies, so they aren't as short as first stated. Plus on some drugs you then get other patents - see inhalers where you have patents on the devices as well as the drugs.
Equally asking to prove safety only... you still have to do the studies, and need a lot of exposure so why not capture outcomes too? I'm not convinced it would be a lot cheaper to omit the outcomes - you still have to recruit the patients, and monitor them. Take HIV, if you don't test blood, you don't know what is happening to patients blood counts, since you have the blood, why not also check for efficacy in viral supression?
Regulatory is also based on benefit:risk, without evidence of benefit, why tolerate any risk? Also how many clinicians would give new drugs with only theoretical efficacy?Hopefully none. All the drugs that fail trials we also think should work, and spend money to prove it! Not to mention the market for pharmaceutical is worldwide, the EMA, Health Canada, Swissmedic and others would throw something out with no evidence of benefit, greatly limiting your market, even in the US.
Are the FDA perfect? No. Are they excellent? Yes. Some of my research touches up against what they do, and my respect for them grows with every interaction. They're also continually improving. The fundamental problem is that we're looking at a 'better than the Beatles' problem
https://en.wikipedia.org/wiki/Eroom%27s_law
What might work? More collaboration between companies. If we can enrol one placebo arm, but test 4 active products against it, we cut the costs (and improve the ethics by giving fewer patients placebo). There are a few other minor tweaks too, but it is something we all want to improve, and are trying to do so!
Safety is the largest component the FDA analyses for, not exactly "orders of magnitude" less than efficacy.
Increasing the patent length, however, would increase the payback. Not as much as one might think, however, because the time-value-of-money analysis means that money far in the future is worth less, so there is diminishing returns at work.
First, on the revenue side, extend patents on pharmaceuticals as long as possible. His model assumes a very narrow window of payback; quadrupling or quintupling that window would make many more drugs feasible to investigate.
Second, to use the author’s word, “de-risk” all trials. The best way to do this would be to end the FDA. But an intermediate step would be to eliminate the requirement to show efficacy, only retaining the requirement to show safety. Safety is orders of magnitude cheaper to prove.
Trials could be conducted much more cost effectively with these two changes. There would be more time to form patient and investigator networks, the structure of trials would be less complex, many more drugs could be tested, and many more drugs would be profitable.