More on the decline effect in drug research

There are a number of good articles on the ongoing problems with developing new drugs, but this one leaped out at me. In particular, check out the logarithmic decline in new drugs brought into use per billion $ of research and development funding:

There are some facts not shown in the graph that need to be pointed out. First, the number of new drugs (NME’s in the graph) that have appeared each year has risen from an average of around 15 in the 1940’s to around 25 now. At an average of 20 per year, we have well over 1,000 new drugs that did not exist in the 1940’s. But the effectiveness of research in dollar terms has fallen by a factor of around 100 or so, after inflation, meaning that we are spending more than 100 times as much each year, to achieve not quite double the output of new drugs today.

Very sobering stuff. I need to get back to work on my book!

One Comment

  • Admin’s Note: This was a post that came in while I was busy changing jobs and apartments, and I didn’t get a chance to ask the author about it. Part of it seems to have been lost, and the remainder has been applied to a non-climate-related post. So I’m a little unsure what to do with it. I am approving it for the moment while I figure out where it should go and where the missing portion might have gone.

    * * * * *
    , your arguments make sense. Unfortunately they come retahr late.Here is the way I see it, basically. You can trust science or not. There are multiple reasons to believe that there are huge risks associated with our current trajectory. These risks are confirmed by, essentially, all the worlds leading scientific bodies.If you choose not to believe this consensus, then you can take on the expense of replicating the work or even improving upon it. Given the time frames involved this would have been much better to undertake a decade ago or more. The underlying information has not drastically changed since then.The ideal would be a completely independent verification.You would need to put scientifically adept people on the case who were essentially independent from the current climatological community and with whom you had some sort of established network of trust. Then you would need to expend comparable or greater resources on their efforts than go into existing work. (The dominant cost of the typically quoted amount, 2 billion, goes to obervation satellites. You’d have to decide whether to replicate those. You would be looking at relatively modest expenditure on the order of a billion dollars. If you don’t trust the instruments, much more, up to the cost of replicating much of NASA’s unmanned research programs. The latter seems excessive; in any case it would enforce a delay in excess of a decade.Then you have to decide whether to do this as a clean room implementation. Here there are problems either way. If you don’t expose the participants to the existing literature, there is no guarantee that they would catch up to the current state of knowledge, which emerged over seventy years. In any case few would agree to such conditions for a long time.Since you can’t close the literature, your participants’ objectivity would be tainted by existing expectations. This latter case is perhaps the best you can do, in practice.Consider this, though: the purpose of IPCC was to provide a sober review of the literature for the policy sector, and the result is that IPCC is perceived as every bit as tainted as the community it reviews. In the end, you would just get more climatologists. No matter where they started, once they were sufficiently educated their jaws would drop to the floor, one by one, and they’d cuss and basically see how idiotically we are behaving. Perhaps more realistically, you can demand more formal record keeping and more effort at communication with the larger scientific community and the public. These tasks are difficult and hitherto unfunded. While I very much wish more could be expended on these things, again you are looking at a significant spin-up time.Anything one could do to enhance credibility and checking will have time and expense associated. Since the arguments that serious implementation efforts must start by 2010 to avoid huge costs have been known for about 15 years it seems retahr late. Accordingly, it is irrational not to begin putting the policies in place now. Maybe whatever improvements you might wish to fund might be happening in parallel. These expenditures will remain tiny compared to the energy sector and should. If you want better qulaity of results, stop cutting funding and squeezing the work as punishment for getting the wrong results.The model I use every day suffers badly from a cut that essentially fired the documentation and support team just as the code was finalized. I waste weeks guessing how to run this or that version. These are weeks I don’t have to set up better software or build better tests.

Leave a Reply

Your email is never shared.Required fields are marked *