It isn't clear to me that bypassing conventional standards of rigor or as yet to be invented standards with low likelihood of succeeding designs (biomarkers, observational study designs) is inherently going to increase number of effective medicines out there. When might it still make sense? It would if it turned out that all the expected sources of biases in absence of placebos, with poorly validated biomarkers, with all the confounding and selection biases of observational data turned out to be negligent in the vast majority of drug candidates evaluated! There is a spectrum of rigor here in that you need to tradeoff some bias for trustworthy validation that occurs sooner rather than later.
For example, the FDA has adopted pioneering methods from causal inference to demand that drug developers use certain methodologies to maximize the "valid causal bits" of information to extract from real world data to actually make it satisfy some rigorous of real world evidence. This is clearly not the standard of real world evidence being demanded for veterinary / animal drugs.
Regarding Airbus, another interesting example of successful European industrial policy is GSM. How did the Europeans dominate mobile technology in the 1990s and early 2000s
TIL what a 5 over 1 means. I love everything WIP posts and look forward to them landing in my inbox. You can feel the desire to satiate a curious mind in every post. Thank you and happy holidays!
I love this newsletter, but fyi the link to the article on the Hanseatic League doesn't work. I missed it and would love to read it!
Thank you - fixed, and you can read the article here: https://worksinprogress.co/issue/the-rise-and-fall-of-the-hanseatic-league/
Thanks a lot!
Lots of good things here! Look forward to reading:
Chinese towers and American blocks
The merits of unified ownership
How to redraw a city
Liberté, égalité, radioactivité
How to spot a monopoly
Re: FDA on animal drugs.
It isn't clear to me that bypassing conventional standards of rigor or as yet to be invented standards with low likelihood of succeeding designs (biomarkers, observational study designs) is inherently going to increase number of effective medicines out there. When might it still make sense? It would if it turned out that all the expected sources of biases in absence of placebos, with poorly validated biomarkers, with all the confounding and selection biases of observational data turned out to be negligent in the vast majority of drug candidates evaluated! There is a spectrum of rigor here in that you need to tradeoff some bias for trustworthy validation that occurs sooner rather than later.
For example, the FDA has adopted pioneering methods from causal inference to demand that drug developers use certain methodologies to maximize the "valid causal bits" of information to extract from real world data to actually make it satisfy some rigorous of real world evidence. This is clearly not the standard of real world evidence being demanded for veterinary / animal drugs.
Regarding Airbus, another interesting example of successful European industrial policy is GSM. How did the Europeans dominate mobile technology in the 1990s and early 2000s
TIL what a 5 over 1 means. I love everything WIP posts and look forward to them landing in my inbox. You can feel the desire to satiate a curious mind in every post. Thank you and happy holidays!
I wish I could get an epub of these 10-11 articles. Some of them are
delicious!
The print issue of WiP is wonderful, as well.
That’s great to hear. I think the best is yet to come!