Maybe we should just give up on privacy for medical records and require that all prescriptions, tests, symptoms, etc. are thrown into a large database where they can be analyzed.
Then approval is given for X number of patients, when hypothesis is defined. And further approval follows automatically as long as on-the-fly analysis of all health records supports the hypothesis.
Okay, I'm naive, and we integrating all health records is certainly non-trivial. My point is every prescription is an experiment that could be used as further evidence.
Any time you care about the results of a set of actions, you should measure those results. If we care about the results of medical care, we should measure those results.
It's irresponsible not to. I would venture to guess that millions of people are dying early because we don't have this as a mandatory, built in part of the process.
But let's put this in terms of not just science but economics. A working market depends on having information about the normalized value of a particular proposition. To understand that value, you need to know the proposition's cost and its benefit. Insurance and other systems remove us from information about cost. Throwing away results data obscures the benefit. Is it any shock then that prices spiral out of control?
Could you point to where this number comes from more precisely? I'm incredibly curious, given that it is an order of magnitude different from what I've seen. For example, there's the Tufts study:
which arrives at $2.6B. And if you are interested in keeping up with some of these studies from someone in the industry, I highly recommend Derek Lowe's blog -- he's a medicinal chemist. For example, http://blogs.sciencemag.org/pipeline/archives/2017/09/12/it-...
The tufts study is the definitive source I usually hear as well. But most of the $2.6B is cost of failed candidates. The actual cash costs of developing a successful drug are probably more like $100-200M minimum from start of lead optimization to FDA approval, based on the tufts data and also anecdotally if you track venture funding raised from companies that get drugs approved
In some indications however one phase 3 study can be $300M+
Why shouldn't we consider the cost of failed candidates? That's like saying lottery tickets are a great investment, if we only look at the winning ticket.
I completely agree you have to include cost of failed candidates and did not mean to suggest I did not
I mentioned the cash costs for developing an approved drug as an additional data point to that illustrate GPs $35M is too low even excluding cost of failure
The actual cost of developing a successful drug includes failed candidates. If you knew what chemicals would be effective medicines from the start, you wouldn't need to do research.
> If you knew what chemicals would be effective medicines from the start, you wouldn't need to do research.
I mean that's not really accurate, even if you know a drug will be effective you're still legally required to do the research. There are plenty of drugs we know will be effective in advance, but they still need to be tested. Similarly, there are lots of drugs that we know will have maybe only a 5% likely chance of being effective, but it's worth spending money to test them anyway because of the potential profits. So including the costs of failed drug trials doesn't really make any sense.
If you knew it was effective what's the point of the research? The research is what's used to demonstrate efficacy. The FDA requires research because it cannot be known if a drug is safe and effective without research. They don't require you submit all the research that found a drug didn't work. They only require it of the one that did. It doesn't change the fact that you have any number of chemicals that could do any number of things, and you do research to demonstrate which ones have the desirable properties. Not that no one has to constructively invent a process to manufacture a drug, but what a drug patent typically grants is exclusive manufacture for proof of medical utility.
I don't know the details of the nexium approval but for novel drugs you don't know it's effective until you test it
The biggest driver of the high cost of drug development are failed late stage (phase 2 and 3) studies. Things look effective in the lab, then look effective in animals, then look potentially effective in exploratory human studies, then when you test rigorously for effectiveness in humans, they often fail. And these late stage human studies are the most expensive
For generic drugs where you "know" it is safe and effective, there's an expedited pathway with fewer / smaller studies. And for slight tweaks to approved drugs there is another process that's more burdensome than the generics approval process but not as bad as a full new chemical entity path
Because doing controlled trials is how you come about that knowledge. Otherwise you're just guessing. You can guess something is safe, but until you try it for real in a systematic way, that's all it is, a guess. That's not knowledge.
The upper bound isn't really relevant though, because of course any company can spend infinitely much on anything, and will if they're trying to maximize the amount of time they can sell the drug while it's under patent. The only thing that matters is how much the least expensive drug approvals cost.
I'm probably taking you too literally but companies can't spend infinite amounts because they don't have infinite money
The upper bound is actually very relevant. Those $300M studies are cardiovascular outcomes studies, which are required for many diabetes / cardiovascular drugs, and those expensive studies have basically stifled innovation in those areas because the cost of developing those drugs is too high compared to the expected profits
Not sure I understand the statement that the only thing that matters is how much the least expensive drug approvals cost?
I'm sure there are other folks on here who know more on the subject, but the last time I looked into data from that Tufts group, it smelled fishy to me: closed source analysis of closed source data.
It seemed to hardly count as verifiable academic research. The raw data was seemingly made available privately to the researchers by way of their close relationship with the companies they're studying, the raw data is not made available by reproduction of their analysis, and there's no way for others to acquire similar data for comparison.
They may be honest, and the data and analysis may be accurate, and it may not be trying to paint a picture that favors its benefactors, but I did not conclude any of those things.
The oft-cited $2.6B cost to develop a drug from the tufts studies pertains to r&d costs up to FDA approval. You need FDA approval to market a drug and the tufts studies don't include any pre-approval market research costs from what I can tell
However a big part of the $2.6B is the money invested into drugs that don't end up getting approved, ie cost of failure. The actual cash cost of getting a drug approved is probably an order of magnitude less
Once approved most drugs do 90%+ gross margins so are quite profitable
On the flip side, most drugs don't actually do anything. Or have side effects that are worse than the thing they're trying to cure.
sure, i'm on board with lowering the cost, but not at the expense of lowering the efficacy.
I could see a good argument for balancing clinical trials vs toxicity, commonality and threat. Perhaps a drug that's only useful for .0001% of the population that would die otherwise, and isn't toxic, well maybe we can lower the bar there. but something that half the population takes, and overdoses will kill you, that should require really high standard. this approach is tough with the current system of off-label prescriptions though.
I would say that a high bar for toxicity and side effects is good. Leave it there.
But there can be a more graded approach for effectiveness. There is little reason to completely forbid the commercialization of a drug because it's effectiveness isn't certain to a very low p-value.
But, as I understand, they aren't independent considerations; toxicity, side effects, and effectiveness (and the nature of the condition they treat) are weighed together, not evaluated independently. The acceptable toxicity and side effects of a drug that's near 100% effective at treating quickly-fatal condition with no other available treatment is not the same as for something with only moderate effectiveness at treating headache pain.
Second, I'm not so convinced by a science writers rather magical, in that we can't see his variables (just trust me I got it all right!), napkin math from 2013, and the complaints of a drug industry power player.
> We should be finding better ways to do clinical trials to help companies lower the cost.
I'm sure the rich drug companies can find ways to do this. Like cut ridiculous compensation packages when apparently your operating costs are skyrocketing?
I'm not inclined to "help out" companies doing their share to dig themselves into a financial shitshow.
They have enough power, including, apparently, an SC decision that let's them expropriate private property on vague promises of jobs, then providing no jobs:
And let's not forget the freedom they have to tweak a generic (with no change to the effect of the drug) and earn a fresh patent claim.
I keep hearing pundits go on about how government can't seem to control it's costs, and is ridiculously inefficient -- we best just shut it down. Then in the next breath, they proclaim giant corps are having a hard time managing theirs, but obviously WE MUST HELP THEM.
https://www.forbes.com/sites/matthewherper/2013/08/11/how-th...
We should be finding better ways to do clinical trials to help companies lower the cost.
I'm glad there was a blockchain event in Miami recently about this. It's one of the most under discussed issues in tech/healthcare today.