4 ms·

"A clinical trial for patients with metastatic prostate cancer lasts only three years compared to an 18-year-long trial for those suffering from a milder, local

"A clinical trial for patients with metastatic prostate cancer lasts only three years compared to an 18-year-long trial for those suffering from a milder, localised prostate cancer. Since a typical patent is in force for 20 years, firms only have two years of effective patent length left to commercialise a new drug against localised prostate cancer. The data paint a bleak picture. The economists find that pharmaceutical companies conduct 30 times more clinical trials for recurrent cancer drugs than for preventive drugs"

maxxxxx 12y ago
How do you conduct a trial for a preventive drug? Just wondering.
- hga 12y ago
  Your control group during the trial gets placebos, and you compare how many contracted the thing you're trying to prevent in two groups. In some ways simpler than the control group getting the best current treatment, in which case you have to compare the efficacy in both groups.
- arjie 12y ago
  Presumably you would have to do a longitudinal study (could be an RCT or cohort or something else). If I think about it, it seems like studying how long a vaccine is effective for is a similar problem.
- richardwigley 12y ago
  (While we wait for expert views) I would say it was no different from any other medical trial (you would have two populations, one getting a treatment and one not getting it, and you would observe them and see if there was a statistically difference) - With a preventative drug you would be looking for people getting cancer - this is a rare event (and so need more people under observation) and would take many years before you had a marked difference in the two populations. Compare this with late stage cancer trial where you are monitoring people dying - this is going to, unfortunately for the people under observation, give you a much quicker 'outcome'.
- TheLoneWolfling 12y ago
  Give a bunch of people the drug and a bunch of people a placebo, wait a long time (decade+), see if there is a statistically-significant decrease in the thing that it is trying to prevent.
  - maxxxxx 12y ago
    It seems quite difficult to set this up. Who wants to take an experimental drug for decades? And if it's a rare disease you need a lot of people.
- chris_va 12y ago
  Vaccine trials are essentially this. You give some percentage of the population the vaccine, and you compare their survival rates to the rest of the population.
danielweber 12y ago
30 times more clinical trials for recurrent cancer drugs than for preventive drugs How do the economists know that it's for economic reasons? Maybe it's a lot easier to theorize and conduct trials for one class of drugs versus the other.
- carbocation 12y ago
  On a logistical level, it's much easier to start a 3-year trial than a 30-year trial. Tracking people for 30 years is more challenging. Giving an experimental drug to healthy people for 30 years is challenging and ethically more delicate. But perhaps from the view of the scientists, motivating yourself to perform a trial that won't yield results for 30 years is quite an effort.

maxxxxx 12y ago

How do you conduct a trial for a preventive drug? Just wondering.

hga 12y ago
Your control group during the trial gets placebos, and you compare how many contracted the thing you're trying to prevent in two groups. In some ways simpler than the control group getting the best current treatment, in which case you have to compare the efficacy in both groups.
arjie 12y ago
Presumably you would have to do a longitudinal study (could be an RCT or cohort or something else). If I think about it, it seems like studying how long a vaccine is effective for is a similar problem.
richardwigley 12y ago
(While we wait for expert views) I would say it was no different from any other medical trial (you would have two populations, one getting a treatment and one not getting it, and you would observe them and see if there was a statistically difference) - With a preventative drug you would be looking for people getting cancer - this is a rare event (and so need more people under observation) and would take many years before you had a marked difference in the two populations. Compare this with late stage cancer trial where you are monitoring people dying - this is going to, unfortunately for the people under observation, give you a much quicker 'outcome'.
TheLoneWolfling 12y ago
Give a bunch of people the drug and a bunch of people a placebo, wait a long time (decade+), see if there is a statistically-significant decrease in the thing that it is trying to prevent.
- maxxxxx 12y ago
  It seems quite difficult to set this up. Who wants to take an experimental drug for decades? And if it's a rare disease you need a lot of people.
chris_va 12y ago
Vaccine trials are essentially this. You give some percentage of the population the vaccine, and you compare their survival rates to the rest of the population.

danielweber 12y ago

30 times more clinical trials for recurrent cancer drugs than for preventive drugs How do the economists know that it's for economic reasons? Maybe it's a lot easier to theorize and conduct trials for one class of drugs versus the other.

carbocation 12y ago
On a logistical level, it's much easier to start a 3-year trial than a 30-year trial. Tracking people for 30 years is more challenging. Giving an experimental drug to healthy people for 30 years is challenging and ethically more delicate. But perhaps from the view of the scientists, motivating yourself to perform a trial that won't yield results for 30 years is quite an effort.