The term (in the US) wasn't originally that long. It was decided individually but not to exceed 14 years. That's long, but much more reasonable in a time when it could take months for just some basic roundtrip long-distance correspondence.

I don't think you can just state that in good faith while ignoring how much harder/more expensive R&D is these days.

How much harder/more expensive is R&D these days?

Is it harder by an amount that exceeds the amount by which the readily addressable market has expanded for most things?

I wouldn't be surprised if it takes 5 years (or more) to both get a patent and litigate it though any appeal. FWIW, in the current U.S. system, my understanding is the time it takes to get the patent comes out of the 20 years, though I think you can get some time back if it takes a very long time.

Correct. There IS a "patent term adjustment" if the PTO takes longer than it should.

The term you want is "prosecute" not "litigate."

It should probably depend on the product type. I get 20 year parents for medicine where research is expressive and time-to-market is long. But for most fields 5-10 years is plenty to make back your R&D investment and make a profit. And after your patent expires you still have the first-mover advantage

Pharma "research" categorization of expenses is maximized. In part to create sympathy for long patent terms, in part to obscure the role of government funded research underlying many pharma products, and in part to maximize tax benefits.

And in large part because FDA multi-stage trials are very expensive to get through and the risk of failure is high. I know it is popular to hate on the pharmacy companies right now - and deservedly so - but the research costs are not totally made up like you are implying.

Maximized is not "totally made up." You will find software companies maximizing R&D tax credits, too, for example.

A JHU study found median costs for clinical trials of $19 million.

That's expensive relative to software. If I piled HIPAA, payments processing, and whatever telcos might require for hooking up to SS7 in one app (call it "X") I doubt I could get halfway to $19 million.

But $19 million is peanuts compared to pharma revenue. I stand by the assertion that pharma gets undeserved sympathy for long patent terms.

> I stand by the assertion that pharma gets undeserved sympathy for long patent terms.

Sure. That's an assertion that's easy to stand by because it's so easy for it to be true because it's so open ended and unbounded.

I think most people believe the pharma companies are playing it up some and getting some undue sympathy, but 10x the sympathy and 10% more sympathy are vastly different situations and your assertion is "true" for both. It's less a question of whether that assertion is true, and more about to what degree, because that's what really matters in a discussion like this.

A statement about the median cost of a single study across an industry compared to pharma revenue (across an industry, across all studies done, whether resulting in a successful product that can make it to market and earn revenue), is not something we can make useful assertions from, whether or not I believe your assertion to be true (I do). It's just throwing disjointed numbers around and using implications instead of data to link them, which isn't a useful way to convince people who want to understand what's going on.

How is an industry-wide median cost "disjoint?" What number would better represent typical clinical trials costs?

Is it the median cost for a trial or for all trials that a company does? A company might have many trials going on at once, of greater or lesser costs, but revenue is an aggregation of everything over a year. A company with $100 million in revenue and 10 trials ongoing and a company with a $1 billion in revenue and 10 trials ongoing would possibly lead one to specific conclusions, but even when it's broken down that much it doesn't really mean anything for the examples presented, because the mean cost of the trials that the company in question has ongoing could be entirely different. The 10 trials the $100 million revenue company is administering could average $1 million, and the 10 trials the $1 billion company is administering could cost 80 million each, by nature of the types of drugs they and markets they are aiming for.

It could also be that there's lots of low cost trials around $20 million, but slightly less than half of the trials that are done are very expensive, and 4-5 times the cost. This would be more obvious if there were values for bot mean and median, instead of just median.

The looser the data, the less you can actually infer. Any time someone tells me the median cost for something across an entire industry, my immediate reaction is to wonder what that means, not assume I actually know what it means, because it's far too little data to really know. How many trials are done in a year on average? How many a year do companies average based on revenue? Or simple, what's the actual amount spent on trials compared to revenue, and over how many years are those trials run? Is yearly cost of a trial really a good indicator, if trials take 5-10 years to run, so a failed one results in not just loss of money spent but opportunity cost over many years?

Note: If you provided a reference for that median number I would have actually looked it up to see if it made sense to me, or if I had additional questions about how they went about their measurements. Instead, all I have a questions about what the numbers mean because I have no context.

Here is a link to the JHU announcement: https://publichealth.jhu.edu/2018/cost-of-clinical-trials-fo...

You will like some of the numbers on that site even less than the $19 million median cost for a trial.

The study was published in JAMA, link on the JHU page. I hope this does not sound like an appeal to authority, but good luck discrediting the numbers.

Thanks!

> You will like some of the numbers on that site even less than the $19 million median cost for a trial.

> good luck discrediting the numbers.

I'm not sure why you would think I wouldn't like the numbers. I'm not pro-pharma, I'm just pro-informed reasoning.

That said, the article you reference is a bit more nuanced than your take. By the studies they're citing, while the trials for drugs that were approved have a median cost of $19 million, they are clear to outline that it does cost a lot to develop a new drug:

The $19 million median figure represents less than one percent of the average total cost of developing a new drug, which in recent years has been estimated at between $2 to $3 billion.

“The cost of generating this fundamental scientific information is surprisingly low given the total cost of drug development and the high price tags on many drugs,” says study senior author G. Caleb Alexander, MD, MS, associate professor of epidemiology and medicine at the Bloomberg School.

So, while studies may not be a large cost center for new drug development, they state that new drugs are very costly to develop. Or maybe studies are costly, and it's just that it takes very many of them until one shows success. Ten concurrent studies on ten variations of a drug to look for something promising would cost ten times the amount, and there's still no guarantee that any one study will end in a positive outcome.

If we're talking about the justification for patents for pharma companies, it seems like we should focus on the "total cost to develop a new drug is estimated at $2 to $3 billion" part and not the "$19 million median cost to run a study" part[1], which as I noted is lacked enough context to know exactly how much it relates to drug development cost.

That said, I'm not sure the total cost number referenced above is accurate (to my taste) either. How much of that is advertising? How much of that is kickbacks and trips and drug rep lunches and free samples to doctors? I'm not sure because I don't have access to the full paper, so I'm not sure what their methodology includes. It wouldn't surprise me to learn that a few "blockbuster" drugs had 80% of the cost in massive advertising and good-will pushes to seed the public consciousness about them so they could make more money in the end, so I'm really skeptical that $2 to $3 billion is an accurate indication of drug development cost just as much as I'm skeptical that a $19 million median trial cost equates directly to the drug development cost in a way that's obvious from that number alone.

1: Referenced and linked in JLU article, but here's the link as well: https://www.sciencedirect.com/science/article/abs/pii/S01676...