how a compound becomes a drug
Jul. 6th, 2011 09:21 am![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
pktechgirlbackupA lot of the building block research in medicine is done by universities, with government money (the NIH, NSF, sometimes the military, and smattering of smaller agencies). It's highly unlikely this research would ever get done without government money, because you can't patent a biological fact (source: Overdose). But big pharma couldn't do their work without it. All in all, I think I'm okay with how this works out. Basic bio research has all the hallmarks of things that should be government subsidized: non-rival (my use of a fact does not prevent you from using that same fact, non-excludable (meaning it's either impossible or more trouble that it's worth to keep you from using it), and undersupplied (this one has a fuzzier definition). It does lead to perverse incentives though: hormones are found naturally in the body, so you can't patent anything you discover about them.
That means there's no financial incentive to do research on the effects of bio-identical hormones, since your competitor can cite that same research when selling the product. But you *can* patent hormone-like molecules that are not identical to those found in humans, so that's what people research and sell. As someone who believes that bio-identical hormones are likely to be more effective and have fewer negative side effects, this discrepancy bugs me. But the solution is not to punish drug companies for researching their awful fake horrible hormones, since they'll just move that money to curing baldness, but to recognize that this is another undersupplied public good we should invest in. The same goes for nutritional research.
Many people complain that the government is doing all the heavy lifting and drug companies are just picking out the gems and cashing in, implying that the additional research the drug companies do is unimportant. First, if it's truly unimportant, we should tell the FDA to stop requiring it. Second, if it's so easy to do, why don't we (in the form of government and/or universities and/or private charities) do it ourselves? If we can truly do it better than the drug companies, than hurray. If not, we've wasted some money, which I don't love, but it was worth a shot. We could also patent building block research and charge companies. I have no idea if this is a good idea. The answer strikes me to be based more in fact than in principle, and I don't know enough facts. If no one else has enough facts either, and I suspect they don't, we could always try patenting a few things and see what happens. Again, it might waste money, but not a ton, relative to the potential benefits. My only concern is that we don't switch the whole system over without even investigating if the new way has the potential to work.
Another option is buying patents off of companies and releasing them into the public domain. This brings up the sticky issue of how much to pay for them. One of the nice things about prices is that they give us a signal of how valuable something is: if two drugs are equally effective but one has more side effects, how much is the better one worth? I don't trust the government to evaluate that at all. But I do more-or-less trust other companies. Not because companies are inherently smarter, but because dumb companies go out of business. So what I would like to see explored is that for any biomedical patent to immediately go up for auction at a specified point. Anyone could bid on it, including the original inventors. N% of the time, the auction would proceed as normal. 100-N% of the time, the government would step in and buy the patent for [maximum bid]*m, plus maybe compensate the would-be winners a bit. You need to pay extra because the fact that they could lose a bid will lead to slightly less rigorous research than would otherwise be done, and companies will compensate by underbidding (I also think we can help this by paying off the would-be winners, just a touch). You discourage overbidding because some percentage of the time the companies will walk away with a new patent and they don't want to overpay for it. It's not a perfect solution, in that by definition some of the patents will need to continue to be held by private companies, but it might be a pareto improvement.
I've mentioned coordination costs a whole bunch of times without explaining what I meant. The way things are now, a given drug or test might depend on intellectual property owned by several different entities. Potential profits on a given product are hard to predict. So every company has an incentive to negotiate really hard for their cut. These negotiations can be really expensive to undertake even if they lead to nothing, and thus act as a deterrent to even trying. A standardized profit-sharing structure would fix this, at the cost of fuzzing the signal as to what research was really the most valuable. According to Gridlocked, this solution is more or less coming about without any government intervention: much like open source licenses, the industry has agreed on a couple of standard patent sharing contracts, in which everyone can sign on quickly, and payment is contingent on the product actually going to market. This is a pretty textbook example of why I'm wary of government intervention even when private companies are screwing around: a government solution may be better now, but it will lass forever. A market solution, while not perfectly responsive, is more adaptable over the long run.
That means there's no financial incentive to do research on the effects of bio-identical hormones, since your competitor can cite that same research when selling the product. But you *can* patent hormone-like molecules that are not identical to those found in humans, so that's what people research and sell. As someone who believes that bio-identical hormones are likely to be more effective and have fewer negative side effects, this discrepancy bugs me. But the solution is not to punish drug companies for researching their awful fake horrible hormones, since they'll just move that money to curing baldness, but to recognize that this is another undersupplied public good we should invest in. The same goes for nutritional research.
Many people complain that the government is doing all the heavy lifting and drug companies are just picking out the gems and cashing in, implying that the additional research the drug companies do is unimportant. First, if it's truly unimportant, we should tell the FDA to stop requiring it. Second, if it's so easy to do, why don't we (in the form of government and/or universities and/or private charities) do it ourselves? If we can truly do it better than the drug companies, than hurray. If not, we've wasted some money, which I don't love, but it was worth a shot. We could also patent building block research and charge companies. I have no idea if this is a good idea. The answer strikes me to be based more in fact than in principle, and I don't know enough facts. If no one else has enough facts either, and I suspect they don't, we could always try patenting a few things and see what happens. Again, it might waste money, but not a ton, relative to the potential benefits. My only concern is that we don't switch the whole system over without even investigating if the new way has the potential to work.
Another option is buying patents off of companies and releasing them into the public domain. This brings up the sticky issue of how much to pay for them. One of the nice things about prices is that they give us a signal of how valuable something is: if two drugs are equally effective but one has more side effects, how much is the better one worth? I don't trust the government to evaluate that at all. But I do more-or-less trust other companies. Not because companies are inherently smarter, but because dumb companies go out of business. So what I would like to see explored is that for any biomedical patent to immediately go up for auction at a specified point. Anyone could bid on it, including the original inventors. N% of the time, the auction would proceed as normal. 100-N% of the time, the government would step in and buy the patent for [maximum bid]*m, plus maybe compensate the would-be winners a bit. You need to pay extra because the fact that they could lose a bid will lead to slightly less rigorous research than would otherwise be done, and companies will compensate by underbidding (I also think we can help this by paying off the would-be winners, just a touch). You discourage overbidding because some percentage of the time the companies will walk away with a new patent and they don't want to overpay for it. It's not a perfect solution, in that by definition some of the patents will need to continue to be held by private companies, but it might be a pareto improvement.
I've mentioned coordination costs a whole bunch of times without explaining what I meant. The way things are now, a given drug or test might depend on intellectual property owned by several different entities. Potential profits on a given product are hard to predict. So every company has an incentive to negotiate really hard for their cut. These negotiations can be really expensive to undertake even if they lead to nothing, and thus act as a deterrent to even trying. A standardized profit-sharing structure would fix this, at the cost of fuzzing the signal as to what research was really the most valuable. According to Gridlocked, this solution is more or less coming about without any government intervention: much like open source licenses, the industry has agreed on a couple of standard patent sharing contracts, in which everyone can sign on quickly, and payment is contingent on the product actually going to market. This is a pretty textbook example of why I'm wary of government intervention even when private companies are screwing around: a government solution may be better now, but it will lass forever. A market solution, while not perfectly responsive, is more adaptable over the long run.
