Hacker News new | past | comments | ask | show | jobs | submit login
The Myth of Drug Expiration Dates (propublica.org)
609 points by danso on July 18, 2017 | hide | past | web | favorite | 263 comments

This issue is a conundrum. Most drugs don't fall off a cliff of efficacy when they reach their expiration date.

There are drugs such as tetracyclines that should never be used past their expiration dates because they degrade into toxic compounds. Certain classes of drugs such as anti-arrhythmics or drugs like warfarin are dosing critical to the point where I would not want them if they were out of date.

I worked in pharmaceuticals in a medically underserved community for a couple of years. At that time when drug samples expired, sales representatives had to return them to their companies for destruction. One doctor in the area made sure that all the drug reps knew he would accept short dated (but not outdated, which would have been against policy for reps) samples for a free clinic he ran. Everyone I knew participated when they had short dated samples. While reps could not distribute outdated samples, doctors had much more latitude in how they dealt with them. It was one of those rare and wonderful situations that was good for patients, created good will for reps and was all completely within regulations.

I should say this was some years ago and regulations may have changed since then.

Wouldn't the real solution here be to put real expiration dates on things that can harm you, and real expiration dates on things that haven't actually expired?

Like a "dangerous to use after 01/01/2019" label on things that actually are dangerous.

Other items could have labels like, "Safe for use until 01/01/2025 - Effectiveness may be degraded after 09/05/2019"

I worked on expiration dates studies in pharma. It's well known that you could probably always extend the shelf life to 3-4 years for most drugs that are supposed to expire within 2 only, because tablets for example tend to be very stable.

But nobody wants to run long stability studies, because that costs a lot of money (i.e. several batches that have to be analyzed in regular basis, documented, reported, QA'ed, inspected, etc... - and you need to place them in numerous conditions, too, ambient temperature as well as stressed conditions depending on where you intend to market the drug), and bring very little returns at the end of the day. You'd rather want your engineers to work on developing new formulations and regulatory teams to spend time on the application of new drugs rather than adding one or two years of shelf life to an existing drug on the market.

Also know that you can't "extrapolate" and need actual data based on regulations of the big 3 geographies (US, Europe, Japan). So "models" are not accepted. Also, these agencies are extremely conservative when it comes to what impurities should be considered acceptable and at what levels they should be. Not saying they are necessarily wrong, but this is another constraints. I have seen cases where it was difficult to extend one drug's shelf life because of a single impurity being a little over the tolerated limit.

How did they come to an expected shelf-life of 2 years in the first place?

Some medication that expires quickly can be extended simply by refrigerating it. I wonder what other environmental controls could extend the shelf life of common, expensive medication?

> How did they come to an expected shelf-life of 2 years in the first place?

By storing it for 2 years and testing it after.

> I wonder what other environmental controls could extend the shelf life of common, expensive medication?

Common, expensive is an oxymoron. The vast majority of all medication in use is practically free to produce. The packaging of drugs typically costs more than the drugs themselves. Drug discovery has long concentrated on small-molecular-weight drugs that can be mass-produced for a pittance. The reason drugs cost so much is that proving that they are safe and that they work costs so much. Because of this, there is little economic sense in trying to preserve drugs for longer (which would require more extremely expensive trials), when you can just trash the drugs and order up a new batch that has a combined msrp of $100M for $5k.

In their economics, small molecular weight drugs share more with software than with, say, bread. That is, practically all of the economic cost is upfront investment and the replication cost is non-existent.

> expensive medication

Apart from biologic drugs, most small molecule compounds are dirt cheap to produce. So cheap that nobody is even striving to produce them in a cheaper, more efficient way. If you compare this kind of industry with chemical specialty industries, where the mindset is to increase yield as much as possible, in pharma the production cost is not much of a consideration. So, what you pay for your drugs has almost no relationship to how much it costs to produce. You are paying for patents, market exclusivity and the whole system that goes with it (the pharma companies are not the only ones benefiting from it).

> How did they come to an expected shelf-life of 2 years in the first place?

Stability studies supporting 2 years data (submission to approval takes one year, and compiling all data before submission for filing is a 6 months process, so you can start submission with 1 year stability data in your hands, and tell the regulatory agencies you will provide the 2 years data during the second half of the regulatory review to get 2 years).

Because that's good enough for most drugs. 3 years is usually ideal, but not always necessary. If your drug is used in high volume most patients won't get batches that are produced long ago anyway.

The point is we don't necessarily know, or know to the degree of certainty the FDA requires, how many of these things degrade. Perhaps in certain conditions, they degrade at one rate, but in other conditions (such as a temperature difference), they degrade at a different rate. Testing how drugs perform years after production in multiple different scenarios is necessarily time consuming, because it takes at least that much time to be sure.

I'm a test engineer. Highly accelerated life testing (HALT) can probably be used to closely approximate the shelf life of medicine. You need to look at the parameters that cause degradation (humidity, temperature, light exposure, etc). HALT the samples at different temperature/humidity/lighting extremes and use mass spectrometry to check purity. Use the data to create an algorithm that approximates the product lifetime and then correlate the results with actual shelf life for verification.

Unfortunately, ALT is not a solution, because it's not accepted by FDA, except as a temporary measure.

Any drug submission that supports shelf life via ALT data will be expected to also say "and we're doing real time testing to support this", and FDA will expect the follow up real-time data.

Then perhaps the solution is for the FDA to allow HALT testing. Having said that, I imagine there is little incentive for pharma companies to willingly extend the shelf life of their products - perhaps the real solution is for the FDA to actually require HALT testing.

I think the real thing here is, "Better safe than sorry". It's fine to use things like HALT testing initially to say, "this drug should be effective for 8 years" so that folks can get it to market before 8 years have passed.

But if, in reality, that drug only lasts for a strong 5 years, after which time it loses effectiveness, especially in high humidity or with high UV exposure. We could catch this with real-time tests and adjust accordingly - before it actually affects the general public's health.

A good model really isn't good enough in medicine. Real data from clinical trials is needed, and nobody has an incentive to fund that.

The largest barrier to accelerated life testing right now is that we really don't know all the mechanisms that cause degradation (or understand their impact on stability). There are many lab instruments that already provide HALT metrics on how stable a drug sample is expected to be under different conditions. And the Arrhenius and Eyring equations are useful for empirical modeling. These tools are great for quickly identifying when formulations are unstable. But it's very difficult to say with certainty (at least in the pharma world) that something will be stable for X years unless you actually test it for X years.

For the vast majority of manufactured products, HALT makes sense because any uncertainty that remains after testing is not likely to harm users. But with some of these drugs, any uncertainty, even after HALT, could cause major problems for users. And it seems that the FDA is unwilling to accept that risk. But as far as I know, the FDA is actively seeking methods for performing HALT that produce accurate and repeatable results.

I see medicines in foreign countries wrapped in some kind of aluminum foil. Will that prolong the life of the medicine given there is no oxygen to react?

Given the amount of money spent on medicines, I really think there is a good moral business case to identify the correct expiration dates and save money for people.

We just went through the Health care ordeal. Money saved anywhere is money saved for people who desperately need healthcare to work for them.

Except the medicines themselves aren't expensive. The prices are that high because the pharma companies demand it, because of all the tests, which are their main source of income. That is why prices in the US are so much higher than elsewhere (because of the political power of lobbying and cultural acceptance of price screwing the public on medical things).

Given that it's the very same party that would be doing the testing for expiration dates, I kind of doubt they would research this out of the goodness of their corporate hearts. Effectively the same outcome could be realised without the effort of extra testing anything: by just lowering the price of medicine a bit; big pharma would make less money, the public would have more medicine for cheaper.

There may be other ways to approach this from a moral point of view. High prices for medicines are defended on the basis of high research costs, not high manufacture costs.

A case could be made for legislation that obliges free replacement of out of date medication for the original purchaser, perhaps with a small surcharge.

A case could be made for smaller packages so people are less likely to take out of date medications as well.

Seems simple enough, just extend the expiration dates on the meds that don't turn toxic and that still work.

The issue here though is that it's not in the drug manufacturers interest to have super long dates. Putting labels that expire sooner means they can push higher volume.

There's also the not-irrelevant issue that testing is slow and potentially other than free. How do you find out when a drug expires? You have a bunch of it sit around and test it periodically, right?

I assume that would be done after bringing a drug to market, otherwise drugs could easily be delayed 10+ years if they're shelf-stable.

Nobody said they need to find the absolute last day the drug expires before they can get it to market.

If it's a new drug and they can only say for sure that it lasts 6 months then put that on the label and ship it out. But assuming the drug will be out for 10+ years they should be continually testing the expiration date and updating as they've had more time to make that determination.

But like I said initially (and as you nicely pointed out) that costs money so why bother?

> If it's a new drug and they can only say for sure that it lasts 6 months then put that on the label and ship it out. But assuming the drug will be out for 10+ years they should be continually testing the expiration date and updating as they've had more time to make that determination.

This needs regulatory work. And this bring very little returns, so your regulatory resources are better spent somewhere else. There are other incentives in place.

If scientists are able to calculate the longevity of chemicals, food, and other products without having to wait the actual time period, I don't see how or why pharmaceutical drugs would be any different.

It's very likely it is not favorable to companies to look into this, just as it was revealed that the EpiPen expiration date was not really true.

You're absolutely right! Those calculations can be done without actual testing. The only wrinkle is that doing so comes with the risk of failing to foresee something or otherwise being wrong. So there are some error bars involved.

What kind of error rate are you willing to accept in models of pharmaceutical shelf stability? Bearing in mind that errors potentially translate into deaths, probably disproportionately of the less privileged among us?

I'm not sure how it would be any different than existing expiration dates for OTC medicines, food, dairy, etc. In the case of medication, it is most likely to become less effective over time rather than toxic.

Are they?

"able to calculate the longevity of chemicals, food, and other products without having to wait the actual time period"

I doubt that food companies let their canned beans sit for 3 years to figure out if that August 2020 expiration date holds true.

I would expect the chemistry to be characterised well enough to the point where you shouldn't need clinical tests to know it's degredation profile...

I don't know about that. There are variables beyond theoretical chemistry, like the packaging, the manufacturing process, etc.

I'm not anywhere near knowledgeable on pharma, but I worked on an FDA regulated medical device. If we made a claim, we had to test it. Knowing that e.g. every component in the machine was rated safe between temperatures X and Y did not mean that we didn't have to perform real environmental testing.

Regulations don't want you to rely on models.

From what I read, current expiration dates are produced by the sort of actual testing I described.

I would assume there's also liability to think about. If something goes wrong with a medicine after the expiration date it has to be better for the manufacturer. Also you probably want to leave some room for suboptimal storage

I'm sure there is some degree of companies wanting expiry to be like the dates on grocery items that are really "taste best by" dates but in speaking to manufacturing and development people their are some layers of complexity. Compounds do degrade over time in many cases and things like light, temperature and humidity play a role. In pharmacy warehouses they aren't just required to control climate they have to log temperature 24 hours a day in an auditable format. Because of these factors the target is a moving one and while many things would likely be fine, for prescription medicine the tolerance for risk is so low that it pushes things in a very conservative direction.

There is a great comment farther down about how complex label changes can be. Getting a revised label, or packaging through FDA is an enormous undertaking. This why sometimes dosing seems very weird. If research says we need a 2mg and a 4mg but later in the process patients need a .5mg it's a big mess.

Suffice to say it's a deep rabbit hole.

Any such labeling change would likely result in decades of litigation over the precise meaning of each term when someone inevitably comes to harm while taking the drug. From the manufacturer's point of view it's cheaper and possibly safer to use a single clear good/bad date.

The article states with absolute certainty that no patient has ever been harmed by an expired drug. Not doubting your point about tetracycline, but how much would a patient need to ingest to be harmed? Then there's the argument that the reason no one's been harmed is because the policy is working.

Harm due to unavailability of drugs or higher costs are much harder to track, of course.

Yet another case of a balance being hard to reach due to the visibility of costs on one side versus the difficulty of measuring on the other. So easy for the general public to see the cost of strictly enforcing overly conservative expiration dates as close to zero.

/offtopic: these issues are the achilles heals of democracies. Not that there's a better system.

See this comment for some contrary opinions on harm with citations in the literature:


Personally, I take all quotations like the ones in the article with a huge grain of salt. There's often enough of a communication barrier that even good reporters can misunderstand an answer, or for the researcher to be answering a different question than the reporter asks. And there's also the possibility that the researcher just had a failure of memory.

I thought antibiotics lose their efficacy part there due date

Other drugs lose efficacy at different times too, of course. But most still won't do you any harm (except for not helping the issue you're seeking to remedy), which was the OP's point, right? Or am I misunderstanding things.

> except for not helping the issue you're seeking to remedy

... which is harm. If you've applied appropriate medication to deal with the issue, but that medication has gone bad, then you're in a position where you think you've applied a remedy but haven't.

Even if you later find out that the medication was bad, you have no idea how much effectiveness there was left, which affects future dosage calculations.

I have to think they must, considering they need to be refrigerated. (Or at least the ones I've used did)

Some even become toxic, like tetracyclines!

I wonder if a single rule 'no consumption past date' is a safeguard to insure simple behavior from population. If people started to think some drugs can and some cant they might indulge into trying random stuff.

I agree keeping the message simple is important, but isn't the implication of the article that the FDA's requirements need more nuance? Some medicines could have expiration dates of 10 years, others should be kept to 2?

If people know that many expiration dates are untrue, wouldn't that make them more likely to experiment and accidentally take expired drugs that actually lose effectiveness or become toxic out-of-date? Expiration dates that are explicit and trustworthy would make people less likely to try random stuff, not more.

I don't think you make people less likely to experiment by making experimentation successful most of the time.

yes, that's a big part of it. It's a simple rule that's easy to follow, which can often be better in terms of health overall.

Yeah unless we put something better in place, maybe a dedicated QR code website to quickly verify if that box is 1, 2, 10, toxic

In principle this would be a god send. In practice I think this will be so drawn in legalese it will take incredible amount of effort to get valuable information.

There is already a super long information sheet in each of the drugs I take, and each time I read them to get a better grasp of how I should take them I'm flooded by useless warnings (could be all summed up by 'you might get worse taking this. If it happens, stop and call a doctor') and disclaimers.

Every time I needed specific practical info I had to go to a pharmacy and ask someone or search non official online resources. Even super basic stuff like 'should it be taken before, during or after a meal ?', no word what so ever in the notice.

Oh yeah, I agree on the fuzzy feeling when reading the folded sheet. It's mostly "potential side effect side effect annnd side effect".

It's hard to design middle ground solution for this. Between no information and too much and too complex information...

Come on worl, we have machine learning, we can find something.

But there are too many details in how the drug could fail depending on its individual storage.

A big chunk of this article's message is that those considerations seem to not have anywhere near as much of an effect as currently assumed. (EpiPens viable 50 months past their marked date)

I would push two prongs: reinforce that expiration dates are serious, and at the same time test and extend any/all dates to the actual behavior of each drug. turn a one year into 5, 10, 20? Hell yes.

Pretty obvious that big chunks of the health care system are rigged for profit, not efficiency or usefulness.

I don't completely agree with the article's message. I think people who deal with statistical quantities of product and do extensive quality control testing probably have some wisdom about what's really safe. The article refers to a paper (which I've skimmed) that leaves a bunch of unanswered important questions.

It seems like you could use that same space for the QR code to just specify its toxicity.

not web enough

I'll see your web and raise with a blockchain

The other issue (as mentioned in the article) is environment. For example I keep some drugs in a backpack that I regularly leave in a hot car. The drugs that were in the Pharmacy probably had near ideal conditions and didn't even move. Giving end users and consumers an algorithm to follow with regard to disposal simply would stand little chance of working in a practical sense. One size fits all is the way to go.

sirum.org (YC W15) helps redistribute surplus drugs that are often short-dated. Our online platform allows orgs to register, upload their surplus, and we connect them to charitable clinics and pharmacies. Check us out!

A 'mishandling indicator' for field dispensary seems like a good idea.

I think that a better approach to 'expiration dates' should be that highly controlled dispensaries (pharmacies) track lot and production dates. A given medication should have a standard expiration date based on /observed/ potency and reserving a safety factor (which should be known and recorded).

Drugs before the typical expiration date would have very infrequent random samples tested for effectiveness. If there is economic or technical incentive, medications nearing the expiration date would be tested and based on the results a new expiration date and testing schedule for the remaining units set.

> The findings surprised both researchers: A dozen of the 14 compounds were still as potent as they were when they were manufactured, some at almost 100 percent of their labeled concentrations.

What kind of reporting is this? Anything less than 100% is not "as potent as when manufactured", and the sentence implies some of those dozen weren't close to 100%.

> The idea that drugs expire on specified dates goes back at least a half-century, when the FDA began requiring manufacturers to add this information to the label. The time limits allow the agency to ensure medications work safely and effectively for patients. To determine a new drug’s shelf life, its maker zaps it with intense heat and soaks it with moisture to see how it degrades under stress. It also checks how it breaks down over time. The drug company then proposes an expiration date to the FDA, which reviews the data to ensure it supports the date and approves it. Despite the difference in drugs’ makeup, most “expire” after two or three years.

That seems to be the problem. There was a procedure in place to set expiration dates scientifically, and it was ignored for some reason, limiting the legal shelf life of even the most stable compounds to a few years.

I work for a company that has some periphery exposure to FDA-regulation, and as a result we have to do a lot of very similar scientific calculations to satisfy their archival records. A big part of why most medicines still expire (at least for medicines available over the counter) much earlier than they should is that a majority of the testing is done to determine the expiration date at a controlled temperature, pressure, and humidity.

They test how it degrades under stress to determine a maximum viability under perfectly stable conditions. They are not accounting for the typical medicine storage places: a bathroom cabinet where the temperature and humidity will vary wildly within a matter of minutes, a cool, damp fridge, or even the reading table next to your bed-- constantly exposed to UV light from the windows.

I'm relaying this secondhand from a former FDA contractor, but I can only surmise that for a company with the litagative risk level that any pharmaceutical entity incurs, the old adage of "Better Safe Than Sorry" is a major part of the story that's missing here.

Beyond that, what is the incentive on the part of the company? No one is evaluating medications based on shelf life. Longer expiration dates -at best- do nothing favorable for the company, and at worst both reduce the purchase of more of the drug, -and- open the company up to litigation in the even it's dangerous. Why take that risk? There is simply no economic pressure to.

I don't think there's an economic pressure to have a longer shelf life either.

It's not any more efficient to stockpile drugs, and it doesn't make sense to do so, as your need for prescription drugs is likely not going to be multiple years.

Well, although supposedly the Army specifically liked Cipro because of the shelf life, so that's one case counter.

Here's a startup idea:

Develop an "indicator pill" that changes color at the same rate as an individual medication.

It would be in (or part of) the bottle with the other pills, and be customizable to react to humidity, temperature and time in the same way as the particular medication.

The main problem is that there just isn't data on how much these drugs degrade. The expiration date is based on some preliminary measurements and conservative guesstimates. Neither the FDA nor the drug companies really cares or wants to see these dates extended.

There probably is no data in part because there is no good way to use such data.

> Here's a startup idea:...

How feasible would that be for thousands of different drugs with hundreds of thousands of different formulations and chemical compositions and various storage factors ?

In fairness, about 200 drugs account for the majority of stuff sitting on Americans' countertops. No need to hit thousands to get pretty good coverage.

I don't understand your comment. You say they test at a controlled pressure, whereas conditions in the typical medicine cabinet vary wildly, and then say that that's why the drugs expire much earlier than they should.

Do you mean that varying conditions make medicines last longer? You can't mean that medicines would last much longer if stored in perfect conditions, because that's not expiring "earlier than they should", it's expiring exactly on time for the poor conditions they are kept in.

Can you clarify?

I'll chime in.

Most chemical processes happen faster with a higher temperature. Some drugs decay into harmful substances (e.g. Aspirin).

Other factors, such as humidity or UV, may play a part as well.

Rather than testing every permutation of storage conditions over time, manufacturers put a safety margin. Individual consumers aren't in a position to remember exactly how they treated any given container of pills. Even if the manufacturer knew how a given sequence of storage practices affected their drug, nobody would be the wiser.

Ah, I see, thanks.

Most medicines are sold in blister packs these days. Won't that prevent a fair bit of the environmental stress from affecting them?

> Most medicines are sold in blister packs these days. Won't that prevent a fair bit of the environmental stress from affecting them?

Well, it keeps out humidity, and might prevent air-exposure effects. It won't prevent heat effects and some other environmental effects.

Also, quite a lot are sold in bottles of pills or other non-blister-pack forms.

I live in France and cannot remember the last time I saw pills not in a blister pack.

I remember vaguely having seen pills in bottles maybe 20 or 30 years ago. Again this is for France, I know that e.g. in the US one gets refills.

Which by the way I find great because the waste of pills is tremendous. It would have been much better to get the specific number prescribed.

It is a waste of pills- but in many cases the marginal production cost of a pill is incredibly small. See generic asprin, many hundreds of pills for a few dollars at an ordinary grocery store. More than you need, but the packaging & marketing probably still costs more than the pills.

Over here, benadryl is the main over-the-counter drug I see in blister packs. They say it's to slow down the meth cookers.

I think you mean Sudafed. Sudafed is pseudophedrin, which is a great decongestant and used to make meth. In most states, you have to go up to the counter and ask for it, as well as supply ID and sign a form promising not to do anything illegal with it. Benadryl is diphenhydramine, which is just an antihistamine, and you can totally get in bottles (in the states). I tend to always have both on hand in case of an allergy attack. Granted, I'm more likely to reach for the Sudafed than the Benadryl. Benadryl has the side effect of intense drowsiness (I guess you can't sneeze if you're in a coma). I also buy a year's supply of Zyrtec and benadryl at Costco. It might seem like a "waste" but when you suffer from bad allergies, it really isn't.

I'm a long time pollen allergy sufferer living in the hell that is South Florida. I know my allergy meds.

That's not the reason. Benadryl being in blister packs long predates its use in meth cooking, like by 30 years. The real reason is more likely marketing. If they sold them in a bottle, the amount they charge for those few tiny pills would seem outrageous. A blister pack gives the packaging more volume.

This varies quite a bit throughout the world (I think). I've had both blisters and bottles in New Zealand/Australia.

Yeah, I think USA is atypical for bottles. I was surprised to see bottles of 1000 Tylenol* tablets at the CVS chain.

In Australia the only thing in bottles are vitamins and similar, though it is possible to get small bottles.

* acetaminophen / paracetamol

Most medications at the patient-level are in amber vials in the US (all pills loose). BUT there are still plenty of meds still further up the supply chain that are in tamper-evident packaging and could be donated if short-dated.

Great comment, and I couldn't agree more.

When working with biological systems, most of the time your acceptable error margin is between 10-25%. There are a few compounds (like fentanyl or LSD) where the dosages are so small and the compound so potent that 10-25% makes a difference but for the majority, precision isn't that critical. The ones where it does matter are tightly controlled and clinically applied by a professional with active monitoring.

The only drugs that weren't within 10% of their original concentrations, according to the study, were amphetamines and aspirin. I can't think of any uses off the top of my head for those two where even 25% would make much of a difference to outcomes - amphetamines are mostly used as general purpose stimulants or to treat ADD and aspirin is not the blood thinner of choice for life and death situations. With amphetamines especially, what you eat and small factors like blood pH can have a massive impact on their effectiveness so we're already talking 10-20% differences day by day, let alone person to person.

Those 122 compounds were also over-representing publicly available products, as opposed to medically regulated compounds. You would need more rigor for a study where the compound efficacy actually mattered to make a general finding, across all classes of drugs. 25% difference in a beta blocker, glycerine, or blood thinner is a fatal change. These drugs are monitored with physical symptoms (metroprolol) and/or regular blood testing (sodium warafin/heparin/lovenox).

And to add into that, dosage studies are rarely adequately done for most drugs. The dose we take is often a best guess and may be twice or more than we need, so 10% is even more inconsequential.

This is completely wrong.

Dosage studies happens between Phase 2 and Phase 3 trials (and Phase 4 "post marketing" too) and are the most expensive, longest and most scientific process that arguably any business engages in in the entire world.

Phase 1 is a safety test in healthy ~men. These seek to find the maximum dose before adverse effects appear in the healthiest humans.

Phase 2 is a safety/dosage test in a much larger and varied sample. We know dose ranges and begin testing them for efficacy in a varied population.

Phase 3 is the big one. It's a randomized, controlled, double blind test of SPECIFIC dosages in SPECIFIC target populations to scientifically prove efficacy over placebo.

The dose we take is not a "BEST GUESS"

It is the result of a billion dollars of scientific study and that PRECISE NME (new molecular entity) has been studied, at that PRECISE dose, in a target population, for it's effectiveness, and it must be better than existing-treatments and better than placebo.

It should be highlighted that this process takes on average 10 years, costs $1 billion dollars, and has a success rate (success meaning FDA approval) of about 1 in 10,000 attempts (hence the price tag).

Source: Drugs: From Discovery to Approval, Second Edition http://onlinelibrary.wiley.com/book/10.1002/9780470403587

Please see papers like-

https://www.ncbi.nlm.nih.gov/pubmed/14714120 No difference in response between people taking 20 mg prozac daily vs. every third day.

Do all 10,000 attempts go through all the testing? Or are many weeded out before $1 billion has been spent?

The $1 billion dollar cost per drug is risk adjusted, so it includes all failed attempts at clinical trials and compounds that didn't make it through pre-trial screening.

Edit: The $1 billion number is meaningless outside the context of large pharmaceutical companies. The cost for phase 3 trials is in the tens or hundreds of millions of dollars so the floor is a wide range and there have been instances of companies getting incredibly lucky and developing a drug for much less than a full billion.

Why are drugs almost exclusively dosed in multiples of 5, ie 50mcg, 250mg? Surely this introduces a margin of error vs the "perfect" dose.

Come on, metabolism itself is reasonably variable even with controlled diet and perfect adherence. In reality, even more so.

Sounds like you don't know what you are talking about. Your source is a book written with heavy influence by the Pharmaceutical industry to justify the obscene costs of their drugs, making billions in profits off of sick, poor people.

You are the one who is completely wrong.

The dosages most certainly do start out with a "best guess" in the pre-clinical phase. For any given drug we would guess at the doses based on our calculations, and give a range on either end. If we thought the effective dose would be 1 mg/kg, we would run an LD50 of doses much higher than that, 10mg/kg, 100mg/kg. Sometimes our guesses would be wrong, and the low dose animals would die as a result. After figuring out a safe dose, we would then split it further into different dosing groups to test the effectiveness of the compound. Most people in the field are underpaid, overworked, and do sloppy work

Please talk to someone who has actually worked in the Pharma research field, they will tell you most of what they do is just guessing.. why do you think only 1 in 10,000 make it through?

It's all a guess until the end, and even then we're often not sure. Look at anti-depressants. We still don't know how a lot of those work but prescribe them to millions of people regardless. How do you think they determined the dose when they don't know how the drug works?

You and the GP talking about entirely different things. You're talking about how investigators choose the initial dose for the clinical trials, he's talking about the dosage that is actually prescribed by doctors.

Hence the 'The dose we take is not a "BEST GUESS"' (emphasis mine). Unless you're a lab animal or a person in a clinical trial, the does you take is the end result of lots of money spent on first figuring out maximum safe dosages (Phase 1) to prescribing guidelines that doctors are taught (Phase 3).

Antidepressants are a unique group of drugs because the dosage is highly dependent on each person's unique brain chemistry and a huge part of a good psychiatrist's job is working with their patients to find the right dosage and combination. The problem is that many psychiatrists don't have the time (due to insurance billing practices) or the patience to do the work but any decent psychiatrist will tell you that recommended doses for antidepressants are just a safe starting point for most patients, not the dosage that they will eventually find most effective.

For the record, I have worked in the pharmaceutical industry on pre-clinical drug development and Phase 1-3 marketing applications and the GP's description is largely accurate whereas you seem to have a chip on your shoulder. I wouldn't trust a single book about the drug development process that wasn't heavily influenced by the pharmaceutical industry in one way or another just like I wouldn't trust anyone without semiconductor industry experience when talking about Intel's cutting edge fabrication processes.

> Antidepressants are a unique group of drugs because the dosage is highly dependent on each person's unique brain chemistry and a huge part of a good psychiatrist's job is working with their patients to find the right dosage and combination.

How does one accomplish this absent trial and error?

This fact undermines the entirety of your argument.

"Best guess" is exactly what they do.

You do know what "unique" means in this context, right? Hint: it means that antidepressants are the exception to my argument because psychopharmacology is so much more dependent on genetic/environmental factors than the rest of the pharmacological field that we cannot draw conclusions on the effectiveness of dosages pre-treatment without spending 100-1000x more time and money on the clinical trials than we do now.

Please take your clearly ignorant bias and anti-pharma prejudice somewhere else - you have no idea what you're talking about.

It is not a guess. It is statistics. Typically, the dose is either effective or partially effective and often on the smaller side as Phase 3 also looks at side effects.

>It's all a guess until the end, and even then we're often not sure. Look at anti-depressants. We still don't know how a lot of those work but prescribe them to millions of people regardless. How do you think they determined the dose when they don't know how the drug works?

We know how antidepressants work, the chemical reaction is well known and can be quantified at different doses. We just don't know why they work because we don't understand why/how the brain works.

Pedantic but important difference.

No, we don't know the antidepressive mechanism of antidepressants, therefore we don't know how they work. Understanding some neurochemical reactions and biological adaptations they cause is not enough. (We also don't have a complete understanding of depressive disorders, we don't really have very effective medication/therapy for many depressive people, etc.)

SSRIs, SSRIs, MAOIs and TCAs all have reasonably well understood pharmacology. As an extreme example, there are very detailed guidelines on how to prescribe venlafaxine, to whom, and who should supervise. You're free to dispute these claims but I think you need to produce more than a blanket dismissal, especially as the OP offers a seemingly reasonable source.

You're talking about the study linked as "2006 study of 122 drugs", right? I can't find amphetamines in the tables or the text, what page is it on? (or was it maybe another paper linked in the article?)

I have a prescription for dexamphetamine, but I don't use/need it a lot, so I have some bottles left that are a bit old--not past the expiration date, which appears to be slightly less than 3 years (after the date of the recipe), on the bottle I'm currently looking at. So I was curious about the amphetamines in particular, and if it's just efficacy deteriorating a bit that's fine, because in my personal experience the effect (which is quickly and clearly noticeable) varies easily by 25% already, depending on so many other factors (like what/how much I eat, how well I slept, stuff like that).

>your acceptable error margin is between 10-25%

10-25 percent of....what?

> There are a few compounds (like fentanyl or LSD) where the dosages are so small and the compound so potent that 10-25% makes a difference but for the majority, precision isn't that critical.

Wrong. Therapeutic ratio, not dosage.

Wrong. LSD and fentanyl have polar opposite therapeutic indexes - the therapeutic index is specific to toxicity whereas I'm talking about clinical outcomes in general. There are drugs that have threshold doses below which there is essentially no effect on the patient but aren't toxic until 10-1000x that threshold dose. Fentanyl just happens to have an effective threshold dosage that is dangerously close to its LD50.

LSD has an exceptionally wide therapeutic index, but fentanyl's therapeutic index of 270 is pretty good as prescribed drugs go. That's a far wider TI than over the counter painkillers like ibuprofen. Warfarin, a common narrow-therapeutic-index drug, has a TI of ~2.

Fentanyl's therapeutic index is NOT 270 except when used as an opioid supplement for surgical anesthesia cocktails, where the dosage is less a microgram per kilogram and it's balanced with several other drugs and opioid antagonists. The maximum recommended dose for fentanyl is 50 mcg/kg and that requires ventilation because it is well above the estimated LD50 of 30 mcg/kg. The recommended low dose for fentanyl on its own is 1 mcg/kg so at best we're talking about a therapeutic index of 30.

LSD's therapeutic index is unknown because its LD50 has never been established in humans or monkeys and you can't extrapolate from non-primate animals to humans (for example, rats can handle a dose of fentanyl that is a thousand times higher per kg than monkeys can, which is where many of our LD50 estimates come from).

I didn't know that extra context about the numbers for fentanyl. Thanks for correcting me.

I was going to post something long about LSD but it turns out I am just wrong. So thanks twice, and have my upvote.

Glad I could enlighten. The therapeutic index is a very subjective measurement that makes some drugs sound safer (or more dangerous) than they really are. Both the numerator and denominator in the ratio are very situation dependent. For example, trazadone is a hypnotic at doses in the tens of milligram but an anti-depressant at doses in the hundreds of milligrams so depending on what you're using it for, its therapeutic index could be different by up to 100x.

Likewise, in a clinical setting (i.e, with a trained anesthesiologist in fentanyl's case), the dose at which 50% of patients die from a given drug is usually much higher because trained professionals can quickly receive feedback by monitoring the patient's symptoms and apply drugs or medical devices to keep them alive well beyond what could kill someone on the street.

Excuse me, index. A dimensionless quantity. Nothing to do with potency. And you could hardly pick worse drugs to spread misinformation about.

I do not believe that you have the requisite knowledge of pharmacology to talk intelligently on these topics so you should stop correcting people who do, especially when your other comments suggest that your knowledge is based entirely on recreational drug databases. Anyone who has spent time working in the pharmaceutical industry knows that "therapeutic index" and "therapeutic ratio" are used interchangeably and that I never said anything even remotely related to either of those until you brought it up out of nowhere.

The therapeutic index is the ratio of the therapeutic dose to the dose at which the drug starts becoming toxic in 50% of the sample. I repeat, for the third time, I was not talking about toxicity. We are discussing drug expiration dates and you will not experience toxicity from a drug that has been degraded over time to a dosage lower than the one you were prescribed, except in extremely rare circumstances. A drug's therapeutic index is entirely irrelevant to what I am talking about - I don't know how I can make that any more clear.

But what you said was still totally wrong, regardless of whether you're talking about toxicity or anything else. LSD is not so finicky that a 25% change in the dose makes a noticeable difference -- most people would be hard pressed to tell the difference between a 100 ug tab and a 125 ug one.

It also does not follow that because a drug is active in small amounts, i.e. micrograms instead of milligrams, that a small percent change in the dosage will be more important. Full stop illogical, and also not true in practice.

So yes, I'm not a doctor and I don't know the jargon. And when I read your comment I forgot the article was about expiration dates, i.e. percent decrease. And yet you were still wrong on that particular point.

LSD is an agonist for most classes of serotonin, dopamine, and adrenergic receptors - most of which have sensitive, nonlinear responses and all of which uptake the agonist at very different rates depending on individual biochemistry. We don't have the data to say for certain because of prohibition and the DEA's reluctance to allow researchers to study LSD but every other drug combination that operates on such a diverse set of receptors is extremely sensitive to dosage because all of those receptors have interdependent feedback loops - which makes them mostly useless for therapeutic uses. This is also why monoamine oxidase inhibitors (MAOIs) are so dangerous with such a wide array of drugs - the enzymes that break down monoamine neurotransmitters are regulated by serotonin+dopamine receptor interactions and even a tiny amount of MAO inhibitors can wreck havoc on that balance, causing some severe symptoms up to and including death. This happens because, even though MAOs are rather indiscriminate on which monoamines they target, their effect on each different neurotransmitter is variable and nonlinear.

In my personal experience (a decade of Burning man), when you do have a known concentration of liquid LSD (measured with an LCMS) that you carefully handle, store, and dilute for dosages, 100 mcg and 125 mcg can mean the difference between a good trip and a bad one in at least a tenth of my sample size and can drastically change the intensity of hallucinations (going from zero to fractal visuals to aliens) in a good quarter - not to mention the effect on the body high and introspection. If you're buying tabs, the chances that you are even close to the original concentration falls rapidly the further removed you are from the original source. Most people can't tell the difference between a 100mcg and 125mcg tab because those dosages are exaggerated and many people feel stereotypical effects at lower dosages. You can easily test this by eating a significant fraction of a blotter: you are extremely unlikely to experience the effects described by the literature of 50 ng/mL blood concentrations even if you're lucky enough to get to 5 ng/mL effects with a few tabs.

IME bad trips are mostly determined by how neurotic the tripper is. If you can accurately say that a 25% increase causes a 10% increase in bad trips, you must have quite a lot of experience. I at least know that if some people take one tab and others take two at the same time, the ones who took two don't go flying off the walls.

Saying that a 25% change is more significant for more potent drugs is illogical on the face of it, because a % is a dimensionless quantity. If there is some subtle reason why this trend exists I'd love to hear it.

> Saying that a 25% change is more significant for more potent drugs is illogical on the face of it, because a % is a dimensionless quantity. If there is some subtle reason why this trend exists I'd love to hear it.

Actually, the reverse is illogical because few biochemical systems have linear responses, especially when you're talking about something as complicated as the neurotransmitter systems that LSD effects. Potency depends on two factors: the affinity of the drug, or how well it binds to its target receptors, and efficacy, or the relationship between the concentration of the drug (and second order neurotransmitters) and the ability of the receptors to initiate a cellular response. Neither of those two factors are linear and they both change as the concentrations of the drug and its byproducts change. As neuron receptors are activated by LSD, they cause cells to release a flood of other neurotransmitters at varying concentrations (dependent on LSD concentration and individual brain chemistry) that start interacting in complex ways like preventing the LSD and other neurotransmitters from binding as effectively (lowering their affinity), potentiating the cellular response (increasing its strength aka increasing efficacy), or building tolerance (lowering efficacy due to exposure). Each receptor and neurotransmitter pair behaves differently. Neurotransmitters in general can't activate cellular responses before they are above a threshold potential that causes the neuron to react, after which the cellular response is never linear. A 2x increase in concentration will rarely achieve a 2x response unless it falls in a small range where the curve is mostly linear, and even then the complex interactions between all of the neurotransmitters will usually compound into a nonlinear response anyway.

These interactions get so complicated, for example, that you get many cases were an opiate A, which is technically 10x more potent than an opiate B, can actually be less potent at treating mild pain because it has a steep response curve that doesn't ramp up until a certain dose. 10mcg/kg of opiate B might be 10x stronger than 10mcg/kg of opiate A (which could be too low to even feel the painkilling effects of opiate A) but once you hit 50mcg/kg concentrations, opiate A might be 10x more potent than opiate B, whose effects plateau with doses higher than 20mcg/kg. Potency is typically expressed as the [A]50 - the concentration of the drug at which you reach 50% of the maximum effect, which depends on the therapeutic effect you're looking for. So in this example, the [A]50 of opiate A in mild pain scenarios can be 25mcg/kg versus opiate B's 7 mcg/kg while with severe pain, the [A]50 of opiate B can be above its LD50 and opiate A's can be 30 mcg/kg because from 25 to 30 mcg/kg opiate A has an exponential response curve. Like I said, it's very complicated and when measuring potency, you're actually measuring the effects of concentration on "arbitrary units," which could be something concrete like a chemo drug's effectiveness at killing cancer cells or something subjective like pain relief. When it's the latter, especially, potency has a very specific meaning in pharmacology that is very different from how the word is used colloquially.

This is just throwing around a lot of irrelevant details to obscure the fact that your original statement violates the dimensional analysis smell test -- I could make a shitty prodrug of LSD where the threshold dose is 1g instead of 50ug, but you wouldn't expect the 1g version to be less sensitive to a 25% change because it's "less potent." If you wanted to give examples of a drug where a 25% change is a big concern then warfarin or synthroid are much better examples than LSD. As it turns out, you know your shit, but it is still a sloppy statement.

I tried explaining why you are wrong but I give up. You are incapable of accepting basic facts about biological systems and neuroscience that have been known for many decades, no matter how hard I try to explain it to you. Your "dimensional analysis smell test" is entirely idiotic in the face of those basic facts as accepted by the entire medical and biological research fields.

Please don't talk about pharmacology. Period. You don't know what you are talking about and someone might make the mistake of believing otherwise and do something dangerous.

This is what you said:

> There are a few compounds (like fentanyl or LSD) where the dosages are so small and the compound so potent that 10-25% makes a difference but for the majority, precision isn't that critical.

As in, typical dosages for fent or LSD are on the order of micrograms, so we expect them to be sensitive to a 25% change, for other drugs where a typical dose is measured in mg or grams, 25% is less likely to be a big deal, etc. This is the only way to interpret what you wrote, and it doesn't make sense. No such trend exists. None of the "basic facts" which you "explained" above (which I already knew) go anywhere near supporting this assertion.

Put another way, if you know the [A]50 you know one point on the dose-response curve, but in general that tells you nothing about the slope or shape of the curve near that point.

Judging from your comment history, you are not an actual doctor or biomed researcher, just a programmer in that industry. Which makes your teeth-gnashing about what an authority you are a tad less persuasive.

For the former (Amphetamines, or drugs derived from them), wouldn't the potency reduction of 25% be a big deal?

Any idea whether the chemicals they decompose into are ever more harmful than the original?

Aspirin is a bad actor in this regard. It tends to revert back to salicylic acid, which is ok for some specific topical uses, but has potential to be pretty harmful orally. Sniffing an old bottle of aspirin will tell you whether they're still good enough

For the record, what you're smelling for is acetic acid (vinegar). When aspirin (acetylsalicylic acid) hydrolyzes into salicylic acid, the other product of that reactions is acetic acid, which is what's in vinegar.

Thanks! It's been 25 years since my last pharmacology class, the details are all getting a bit vague.

It's not that harmful if you know what you're dealing with. Have to be careful about liver toxicity and dosage but in a pinch it will work just fine as an antipyretic. It was used before aspirin has been invented.

From the study:

> Twelve of the 14 drug compounds tested (86%) were present in concentrations at least 90% of the labeled amounts, the generally recognized minimum acceptable potency. Three of these compounds were present at greater than 110% of the labeled content. Two compounds (aspirin and amphetamine) were present in amounts of less than 90% of labeled content. One compound (phenacetin) was present at greater than 90% of labeled amounts from 1 medication tested, but less than 90% in another medication that contained that drug.


Sounds like maybe concentrations are allowed to vary a bit from the label at manufacturing time? Not sure, maybe someone who knows more can comment.

My brother-in-law works in pharma manufacturing, this stuff is really hard and it's basically impossible to create exactly identical pills, apparently in large part because the active ingredient is frequently <1% of the volume of the pill. The best they can do is control the range of dosage in pills, and there's variance in what that range is across drugs (i.e. some compounds can't be spray dried, can't be heated hot enough to make binders liquid, etc).

Depending on necessity and value they can typically engineer more precision, but it gets very expensive very fast.

> What kind of reporting is this? Anything less than 100% is not "as potent as when manufactured"

Relax, this is the real world, not a computer based series of 0s and 1s. Any such compound begins drifting off from the original specs as soon as it leaves the factory - and in a very real sense it's not exactly at 100% at any time post the moment of manufacture. The question is what kind of deviation can be tolerated.

Offtopic, but even the world of 1's and 0's is subject to real-world constraints - learned in my 90s computer hardware class as I spent weeks in summer after semester debugging why my "basic 8-bit processor" kept crashing after only a few hours of uptime: faulty silicon that periodically read median-voltage differentials cascaded eventually into bad instructions.

The world of logic is built on the shifting ground of reality. Only redundancy and checksums keep it from crashing down in compounded errors.

Well said. Another aspect: Having the option of a 90% effective drug vs a 100% effective drug you DONT have because you cannot afford.

There was a time in my life where our family lived dollar to dollar and skipped medical care due to cost. We would easily settle for 90% efficacy.

It is not even 90% effective. You could adjust the dosage in many cases.

The paper discusses this, they say that 90%-110% is the usual range which is allowed when the drug is sold new. The FDA can set limits on a per-drug basis, and there are some where they require 95-105%, but 90-110% is the most common. (C.f. https://www.fda.gov/ohrms/dockets/ac/06/briefing/2006-4228B1...)

Several of the drugs they tested were > 100%, so it seems reasonable to assume that a lot of this is due to variation in manufacturing rather than to the drug degrading.

It appears that potent is properly used in this article:

producing powerful physical or chemical effects. ( http://www.dictionary.com/browse/potent )

chemically or medicinally effective ( https://www.merriam-webster.com/dictionary/potent )

exerting or capable of exerting strong physiological or chemical effects ( http://www.thefreedictionary.com/Potent )

I think it would introduce a huge amount of legal liability to guess that a drug would survive on a shelf for 10 years based on 6 months of testing. Especially if it's something critical like an EpiPen or cancer med.

At the end of the day, there's no really reliable way to speed up time, and it would be a waste of time to add 10 years to a drug development timeline just so you can leave it on a shelf for 10 years to test if it still works.

And yeah, we could probably go through and retroactively test some Ibuprofin, but also remember that these drugs are repackaged / reformulated slightly every couple years for various reasons, and without a lot of legal protections, no company will consider it worth it to guess about drug lifetime.

I wonder for certain drugs how much of a difference 98% potency is vs 100%. Surely that 2% must matter. Also, what does it mean to be less potent? It's still the same drug right? So does that just mean the pill contains less of it somehow?

No drug where a 2% difference matters can be administered outside of a monitored hospital setting. There are too many variables that affect the drug uptake far more than 2%, such as body weight (100 pound person will have 3x the drug dosage of a 300 pound person), drug absorption (based on body chemistry, other food eaten, etc.) and the actual intake dosage itself (Eye drops, for example, can have a 500-fold variance in how much actually gets administered). That's why over-the-counter drugs have to have very high therapeutic index. https://en.wikipedia.org/wiki/Therapeutic_index

> There are too many variables that affect the drug uptake far more than 2%, such as (...)

...or in case of tablets that come in 2x/4x the dose and you need to cut them in half/four pieces, the precision of your cut and amount of tablet dust that is lost.

And the distribution of the active ingredient vs filler.

It's a lot of guess work.

IANAD but it probably means some of the drug molecules have been rendered chemically inactive.

For many drugs a 2% difference is probably unnoticeable. Why are Ibuprofen pills 200mg? Because it's a nice round number that happens to work for many people. If I have a headache I often break the 200mg pill and only take half of it. It still does the job fine for me with fewer side effects.

I'm pretty sure for almost all drugs 20% doesn't matter. Think about it. I've never gotten an over the counter or prescription drug that was adjusted for my weight. The closest thing I can think of is painkillers (tylenol, etc) giving different dosages for children.

So if a 110lb wife gets the same dosage as a 200lb husband, how on earth would even a 20% difference in dosage amount make a big difference?

2% can only be done in fantasyland, like in a hospital you and the medicine are both closely weighed and administered.

> I've never gotten an over the counter or prescription drug that was adjusted for my weight.

How do you know that your doctor didn't take your weight into consideration when selecting the dose for a prescription medicine?

OTC drugs are often generally provided at a dose that should be safe for most adults, but may be significantly different in effectiveness based on weight.

Because my dosage has never varied, while my weight has and by more than 20%.

>I wonder for certain drugs how much of a difference 98% potency is vs 100%. Surely that 2% must matter.

Why is this a "surely"? On a scale of 100%, 2% is 2% because it's not hugely significant.

You can get a rough idea of the required precision by checking if the dose is weight dependent. If it isn't, 2% won't matter.

And if 2% does matter, you'd better make sure you can measure both the dose and your weight to better than that accuracy.

Chemical reactions between air, mostly, and to a lesser extend between the drug and the pill substrate itself, I would imagine. I'm sure on such small size and long time scales the pill does simply ablate some of it's material.

Disclaimer: this is conjecture.

> That seems to be the problem. There was a procedure in place to set expiration dates scientifically, and it was ignored for some reason, limiting the legal shelf life of even the most stable compounds to a few years.

To investigate how a drug breaks down over a ten-year timespan, you'd need to do at least one ten-year study of the drug. That seems difficult to justify.

I was thinking about this today after hearing a report on the radio this morning on this topic. Hospitals end up throwing away large amounts of drugs each year due to expiration, potentially hundreds of millions of dollars - although there is no real data currently.

It seems like it would be a good long term investment for a group of hospitals to partner together and begin a systematic study of the actual lifetimes of drugs. Instead of destroying all expired drugs they could send off samples to a lab for testing. They could even intentionally keep expired drugs in an off-limits area for 1, 2, 3 years past their expiration dates to gather more information. With enough hospitals participating they could gather a significant amount of data without any one hospital shouldering all of the burden. Once there is enough evidence that a particular drug is still potent the FDA could extend the allowed shelf life. This would save the hospitals money over time since they could stop paying to dispose of and replace as many expired drugs. It would also provide a general benefit to the public since people would have solid scientific evidence for the actual effective lifetime for particular drugs. This would include poison control centers, the hospitals themselves, etc. that would be able to better handle cases where someone has already taken an expired drug.

Of course an alternate way would be to require the drug companies to accept returns from hospitals of small amounts of expired drugs that they would then be responsible for testing and providing the data to the FDA. Limiting the returns to large institutions would keep the costs down (reducing push-back from the pharma companies).

> Anything less than 100% is not "as potent as when manufactured",

Since the % is of labelled strength, it might still be “as potent as when manufactured” if the drug was at less than labelled strength when manufactured, though that, too, would be a serious issue.

I think you're confusing percentage of concentration with the threshold of potency.

Don't Drug MFGs know this and are allowed to "donate" to weak economies.

This article has some shoddy logic. For the vast majority of drugs, the actual marginal manufacturing price per pill is essentially zero. That means that when the expired pills are thrown away and new pills are manufactured, there is no true economic loss. Rather, all that potentially happens is a transfer between the consumer and the manufacturer.

If pill prices were set by some external force, this could at least be an important society-wide transfer. But in fact, in equilibrium pill prices will be affected by the expected rate at which pills expire without being consumed. Even when manufactures have a monopoly, the manufacturer-surplus maximizing price is determined by the demand curve of the consumer which takes into account the expiry rate of pills. (If 10% of pills expire before I consume them, they are worth 10% less to me in expectation.)

Yes, I'm sure there are market model where expiry dates create net economic drag, or a net value transfer between manufacturer and consumer, but it's not even clear which direction the transfer goes. Such an analysis depends on the details of the world and how they are reflected in your model, which are completely absent in this article. Most importantly, the intuition that "letting $100 pills expire for no good reason must cause $100 of damage" is completely false when the marginal cost of manufacturing is low.

(There are exceptions where the marginal manufacturing process is expensive, but the article doesn't focus on these.)

>This article has some shoddy logic... all that potentially happens is a transfer between the consumer and the manufacturer.

The point of the article is that this is an expensive mess for everyone except the manufacturer, which means approximately everyone. As a consumer, having the manufacturer "transfer" money from your pocket into their own sucks. Seems like the article's logic is actually pretty decent.

Also, even if we accept that the marginal cost of producing a pill is near-zero, that's ignoring the entire rest of the supply chain. It's not like drugs are magically teleported. There are real costs in time, money, and resources for creating the packaging, transporting drugs, stocking/shelving them, unstocking them, sending them to a disposal location, and then disposing of them. All of that is a complete loss. Most of it generates various forms of pollution, waste, and other externalities. This also sucks for ~everyone.

Basically, outside of theoretical econ-land, wasting perfectly good things is wasteful.

> As a consumer, having the manufacturer "transfer" money from your pocket into their own sucks. Seems like the article's logic is actually pretty decent.

As I addressed in my second paragraph, it is not actually a transfer unless you introduce new assumptions into the standard model of a monopoly. Please describe them.

> There are real costs in time, money, and resources for creating the packaging, transporting drugs, stocking/shelving them, unstocking them, sending them to a disposal location, and then disposing of them.

These are, in almost all cases, trivial compared to the development cost of the drug. You can group them under the marginal cost of the pill.

> Basically, outside of theoretical econ-land, wasting perfectly good things is wasteful.

Much worse than theoretical econ-land is non-quantitative land where all "waste" is bad and cost-benefit analyses are never done, not even at the order of magnitude level.

Please, by all means, construct a more accurate model. But just criticizing the imperfections of a model, which always exist, and then making non-quantitative value judgements is how we get recycling programs that cost 10 times the value of the material they recover.

>Please, by all means, construct a more accurate model.

Gee, you get to just hand-wave an entire manufacturing process and supply chain into insignificance, but I need a model, huh?

Fine. Let's use a generic drug (ie, not paying patent/development costs) as a proxy for how much the entire supply side costs. How about loratidine. You can get that at Walmart (stocked on the shelf, since that's mostly what the article was discussing) for about 10 cents/pill.[1] Of course that's completely ignoring the disposal costs. I strongly doubt that disposing of hospital/pharmacy quantities of drugs is as easy as chucking them in the municipal landfill. Let's say it costs 10% of the purchase price. So 11 cents/pill just to run a pill through the supply chain and then dispose of it.

Intensifying the napkin math... CDC says there were ~3.4 billion visits that resulted in drugs "provided or ordered". [2] Let's say that those were an average of 7 days worth of drugs ordered, so ~30 billion doses. Let's say that 10% of that much is either wasted by the consumer, or wasted while sitting on a shelf before anybody bought it - 3 billion pills.

3 billion pills/year * 0.11 cents/pill = $330 million/year. I dunno, seems potentially significant, at least to an order of magnitude level. Especially since the costs of the FDA program to extend the shelf life was around 1% of that much.


[1] - https://www.walmart.com/ip/Equate-Loratadine-10-mg-300-ct/20...

[2] - https://www.cdc.gov/nchs/fastats/drug-use-therapeutic.htm

> Gee, you get to just hand-wave an entire manufacturing process and supply chain into insignificance, but I need a model, huh?

Nope, for this you just need a number! (In my comment, the "model" part referred to how monopoly pricing was done.)

> So 11 cents/pill just to run a pill through the supply chain and then dispose of it.

Good, your estimate process is a reasonable one, and I accept this number.

> $330 million/year...I dunno, seems potentially significant, at least to an order of magnitude level.

Ok, let's compare to what the article says:

> Experts estimate such squandering eats up about $765 billion a year — as much as a quarter of all the country’s health care spending.

So we see that our cocktail napkin math has been extremely useful, and that the article has overstated the size of the issue by some three orders of magnitude! (25% --> 0.01%) The many suggestions in the article that this is important on the level of the entire health care system is absolute bunk.

Thank you for engaging in this exercise with me!

$765B is their estimate for all waste, not just wasted pills:

    ProPublica has been researching why the U.S. health care
    system is the most expensive in the world. One answer,
    broadly, is waste — some of it buried in practices that
    the medical establishment and the rest of us take for
    granted.  We’ve documented how hospitals often discard
    pricey new supplies, how nursing homes trash valuable
    medications after patients pass away or move out, and how
    drug companies create expensive combinations of cheap
    drugs. Experts estimate such squandering eats up about
    $765 billion a year — as much as a quarter of all the
    country’s health care spending.
Later on they say:

    Tossing such drugs when they expire is doubly hard. One
    pharmacist at Newton-Wellesley Hospital outside Boston
    says the 240-bed facility is able to return some expired
    drugs for credit, but had to destroy about $200,000 worth
    last year. A commentary in the journal Mayo Clinic
    Proceedings cited similar losses at the nearby Tufts
    Medical Center. Play that out at hospitals across the
    country and the tab is significant: about $800 million
    per year. And that doesn’t include the costs of expired
    drugs at long-term care pharmacies, retail pharmacies
    and in consumer medicine cabinets.
This isn't estimating the same thing as the GP's $330M (hospital waste instead of consumer waste) but $800M is on the same scale, suggesting to me that the marginal cost of drugs might matter some.

(I do agree with your point that most of the cost, and especially the most egregious examples like the wasted epi-pens, don't represent real costs because of monopolies.)

Hey thanks! This is a good clarification. I'm happy to modify my criticism from "the author is strictly off by a factor of 1000" to "the author strongly leads the reader to believe the size of the effect is 1000 times stronger than it really is, but he deftly avoids saying anything that's technically false, and in any case the reader shouldn't care about this since it represents an dead-weight loss of about $2 per person per year, which is less than the value of the reader's time spent reading this article".

Your logic would be fine if manufacturers were required to replace expired pills with new ones. But they aren't, so early expiry dates impose enormous economic costs on consumers of drugs, while manufacturers get the massive benefits of selling drugs twice to same consumer.

A lot of manufacturers accept expired drug. And even if they don't, distributors often do as well.

And expired drugs comes down to inventory management. Don't buy more product than you can sell.

If lots of manufacturers were all manufacturers, this article wouldn't have been written. In truth, the vast majority of expensive drugs aren't being replaced by distributors or manufacturers, or the Shelf Life Extension program for the government would not be saving $2B a year.

And don't blame the victims, which is what your sentence on inventory management does. If a hospital over-estimates patients usage of a drug and gets stuck with expired drugs due to excessively shortened expiry dates, that's on the maker. The maker knows the real expiry dates, usage can never be anything but an estimate.

I would suggest you read up on the Shelf Life Extension Program. It's a specific program around either limited use products or products where the gov't hold such large stockpiles that companies won't take returns, it just wouldn't make sense. The most likely scenario is one where the products are never used.

I had a friend who worked on the federal Tamiflu stockpiles. The gov't and the manufacturer have a deal where the gov't gets a significant rebate on unused product that is either put back into circulation or the manufacturer buys it back and destroys it.

Remember, the gov't has companies competitively bid on manufacturing these products (particularly for generics which were most of the examples in the article). Is the company going to give a 100% money back guarantee if it's not used? Would you? I wouldn't.

So yes, there are certain scenarios (stockpiling) where manufacturers don't take back product. That's not most scenarios.

Most scenarios are hospitals holding 30-45 days on hand of product. If they screw up their inventory they should pay a penalty not the manufacturer.

This is addressed in my second paragraph.

No it's not. Essentially you are saying it's ok to have an "expiration tax" and that it doesn't change the market. I'm saying the expiration tax does change the market, and raises the cost for the consumer, even if it's factored into their purchase. Typically consumers don't have the ability to substitute complements, and manufacturers products are patent protected, so it's another way to raise prices and increase manufacturer profits.

I explicitly point out in my original comment that this applies for monopoly markets! To spell it out in gory detail: in a simple monopoly, the price is that which maximizes the supplier surplus for a given demand curve, and the demand curve will be proportional to probability that any single pill is consumed. If you have a more complicated model, then by all means explain it to us, but do not just assert the contrary with no explanation.

I had exactly that thought when listening to the NPR interview this morning. Unfortunately, the value was taken for granted (even extrapolated x4000 to estimate total value loss across all hospitals in the US).

Pfizer behaves very differently with an increase in marginal revenue than a patient (who is passed the cost from the distributor) would behave with a similar decrease in health care costs.

Anticipating pill expiration and what will be excess supply is a difficult predictive modeling problem. At a practical level, it's difficult for pill expiration date to make it into the price, especially when the purchasers of the pill are different people than those who throw out the pill years down the line.

The cost model is like a time limited software license. If there is some new trick to extend the license, the drug companies would likely just charge more up front.

Exactly. The price is set by the willingness to pay (the demand curve), which accounts for the fraction of purchased pills that will actually be consumed.

It seems like this would be true if everything were working perfectly and there was a gradient of purchasing options across price/expiration. As it is consumers really don't have any choice wrt expiration date since as far as I'm aware the option to pay more for longer lasting drugs doesn't really exist.

It doesn't require a gradient. Suppose that, instead of expiry issues, half of all pills were inert duds for whatever manufacturing reason. (You look in the bottle and the duds are black.) This would make the pills half as valuable to consumers, and manufacturers would simply sell twice as many for half the price. It does not require that consumers have access to pills that have a higher or lower dud rate. And yes, it's possible to construct a model where there are frictional losses from duds (increased spending on pill bottles!), but to assert that you'd need to try to write down some numbers.

Pharma companies get the benefit of near-perfect inelasticity in demand, i.e., if you don't get your meds means you die or have extreme pain, you will likely be willing to pay inordinate prices.

> But neither Cantrell nor Dr. Cathleen Clancy, associate medical director of National Capital Poison Center, a nonprofit organization affiliated with the George Washington University Medical Center, had heard of anyone being harmed by any expired drugs. Cantrell says there has been no recorded instance of such harm in medical literature.

I am more than a little surprised by this. Fanconi syndrome, a kidney disease that can cause bone damage [1], has been repeatedly found in the medical literature to be caused by taking expired pills of the antibiotic tetracycline. [2]

That these people completely failed to remember these cases is distressing at best.

(I heard an interview with the author of this story on NPR Morning Edition today, and also scanned this story webpage. The only mention of tetracycline was by a web commenter on the story.)

[1] https://medlineplus.gov/ency/article/000333.htm

[2] https://scholar.google.com/scholar?hl=en&q=tetracycline+fanc...

I was going to post a similar reply. However, outside of tetracycline, there have never been reported adverse effects from expired drugs. Furthermore, there were only three cases of expired tetracycline related Fanconi Syndrome and that preparation responsible is no longer made. Of course, if it's old...

Otherwise I agree I'm surprised by the reporting ignoring this because this is a classic board question in internal medicine.

Source: http://www.mdedge.com/internalmedicinenews/article/113748/sh...

Edit: Apologies that the above link has a paywall. When I accessed it via google it worked once, then shot up a paywall. If it doesn't work at all I'll remove it.

I have a lot of respect for this kind of reporting from ProPublica. It's a really good public service to report news that isn't actually new.

In this case, part of the news is that the FDA's 1986 Shelf-Life Extension Program ( https://www.fda.gov/EmergencyPreparedness/Counterterrorism/M... ) has been working fine for decades. This is important for people to know and it's worth re-publishing from time to time as a reminder.

They had another great article like this last month, https://www.propublica.org/article/hundreds-of-judges-new-yo... , which basically reported "local courts in New York state had terrible problems according to an in-depth 2006 New York Times report and are still terrible in 2017". Great stuff.

Likewise a lot of the reporting on "civil forfeiture continues to happen and continues to be unfair, here are more examples" provides a valuable civic service.

I know it can be tough in a news organization to re-report something that everyone already knows is true. It can certainly be tempting to pass over truly important stuff in favor of seeking out brand-new news -- a fair amount of science reporting is driven by "what's in the journals this week" -- but this kind of long-term focus on "what's still true that needs your attention" is equally valuable and great work.

"Berkowitz picks up a box of sodium bicarbonate, which is crucial for heart surgery and to treat certain overdoses. It’s being rationed because there’s so little available."

Huh? US hospitals have a shortage of Baking Soda?

Oh yes... it's a pain in the ass. Recently in the news, too [1].

There weren't many days where the hospitals I've worked at weren't short on something. (And these are like, big name academic medical centers in big U.S. cities.) I've run into a shortage of normal saline before. Yes... salt water.

Sometimes hospitals will even stockpile basic supplies to keep from running into shortages... which only makes the whole situation worse.

It causes a lot of patient harm, even if it's usually not catastrophic because we usually have redundant meds. Recently my hospital was out of a common, generic, cheap antibiotic. For many infections in many patients, this meant we had to switch to the second-line drug -- which was more expensive, had more side effects, and was less effective against the bug. This type of thing happens to chemo drugs, to drugs you give after a heart attack, to drugs you need for planned surgeries or procedures.

I don't know how to solve this. Less regulation? More regulation? Subsidies for manufacturers of essential medications / supplies? I mean, normal saline is already ridiculously expensive enough [2].

For more reading: http://news.medill.northwestern.edu/chicago/hospital-drug-sh...

[1]: https://www.nytimes.com/2017/05/21/health/sodium-bicarbonate...

[2]: http://www.nytimes.com/2013/08/27/health/exploring-salines-s...

Your second link says including $546 for six liters of saline that cost the hospital $5.16.

That doesn't sound like the cost of the saline is much of a problem. It sounds like hospital billing is the problem.

It sounds like the labor and other costs of tracking, delivering, etc. the saline, rather than the material costs of the solution itself, are the issue.

Sure, if you think $100 a bag is reasonable for that, then that would be a reasonable explanation.

Remove patent protection for any drugs that are in shortage.

Do you honestly think that baking soda is in shortage because of patent restrictions? The US drug system is messed up, but not that messed up.

They are in shortage because it takes a long time (six-ish months) to certify a factory as producing something safe for consumption[1], and in the normal course of things, there is very little point in setting up a second line to produce medical grade sodium bicarb- no company is going to show a profit from it. So you get one place producing the national supply, just large enough to handle the normal demand. Then they have a problem (in this case, Pfizer was unable to source the glass ampules) and now there is no production.

[1]: And with good reason. Producing mass quantities of sterile, consistent, things with the correct amount of material is hard.

Source: wife is a pharmacist tasked with doing her hospitals contingency plans on bicarb.

> there is very little point in setting up a second line to produce medical grade sodium bicarb

Decentralization and redundancy of medical production lines would make our country safer and more robust to unforeseen events. It would certainly be a benefit. I agree, however, that it would be more expensive.

At least in government, that sort of thing is usually handled with second-source agreements as a contingency of a supply contract. Can have the side benefit of introducing some competition. See AMD, once-second-source-supplier for Intel, Fairchild, and National Semiconductor.

Sure, and for some stuff this is done now. But there are something like 9,000 different pharmacy items (leaving aside medical supplies like sterile bandages, which are procured through a totally different system) in my wife's very small hospital (that ships out all the difficult cases, requiring specialized medicines, to larger nearby hospitals). Should there be second-source agreements- and extra productions lines- ready to go for all of them?

Most of these medicines are like sodium bicarb- the national population doesn't need that much, but if you need it, it's hard to get a good substitute. Snake antivenom or daraprim are good examples of these niche products that are even more crucial than bicarb, but still probably not used enough for any one organization to be able to justify a second-source contract.

This article isn't just about baking soda. If society freely allows you the massive competitive barrier of a patent to ensure your riches, you should also have some obligations to society as part of the contract.

The article isn't, but most of the really egregious price increases we've observed (e.g. epipens, all the drugs that Martin Shkreli bought, etc.) have likewise not been on drugs under patent protection.

In point of fact, most drug companies have faced what are called patent cliffs, where due to troubles in the R&D pipeline[1] drugs lose patent protections far faster than new drugs come online to replace them. While this is probably bad for society overall- we really want lots of new drugs getting discovered and making us healthier![2]- it means that patents protect smaller and smaller portions of the overall drug market. IOW, playing games with them has less and less overall value as the years have progressed.

Need to find new levers to have influence on this market.

[1]: A totally different subject with a lot of theories as to why, but not much in the way of solid evidence.

[2]: Yes, the 1k USD/pill Solvadi is amazingly expensive. By the same token, it actually works, and cures you after three months. This is the sort of medicine that improves humanity, even at that price. There are plenty of other drugs (especially cancer drugs that extend lifespan by averages of <4 months) that do not improve humanity, but Solvadi does. It just costs a ton.

Yes. One of 2 (or 3?) manufacturers in the US making it as a medical product had manufacturing issues, so there isn't enough of it that's cleared for medical use available.

EDIT: https://www.usatoday.com/story/money/2017/05/31/how-baking-s...

Medical grade baking soda, yes: https://arstechnica.com/science/2017/05/baking-soda-shortage...

You can't just take a box from the baking supplies shelf in Walmart.

You could but shouldn't outside of battlefield or other critical conditions. Direct solutions have been used some hundred years ago, prepared on place by a chemist. (They tended to use better than food grade, but not much better.)

Yes, and it's even more bizarre when you realize what we need it for-- critically ill patients who are so acidotic that drugs like epinephrine (adrenaline) don't work. Bicarb reverses that acidosis, it doesn't cure it, but it does make it so your epi and other drugs have a chance to work. The main alternative is sodium acetate, which depresses heart function, causes low blood pressure and reduces respiratory drive.

It's an injectable product. That's not easy to manufacture at all and has high capital costs. You need to make sure it's sterile and remove any pyrogens. Manufacturing injectables requires a much more strict adherence to quality than oral products.

(former pharmaceutical developer here). This article makes it sound like there's a big conspiracy, which there isn't.

A very few drugs (like erythromycin) become toxic so should be discarded.

Some drugs become less efficacious over time, but nobody really knows the shape of the curve as they are simply tested to see if they have the same efficacy on day E as they did on day 0. Well of course all drugs will become worthless as t approaches ∞ but you can guess that since tablets have a very low moisture content, if they are kept in a cool dark place it's likely they'll last a very long time. I also happily keep expired drugs in a controlled environment and use them; all drugs in my car's and backpack's first aid kits get replaced annually because they are exposed to harsh environments.

(Stockpiling drugs doesn't prove anything BTW: if you are stockpiling them against an emergency the presumption is that some efficacy is better than none).

Nobody is going to do accelerated life testing beyond what they have of course. I think extending the required lifetime is a good idea, though I question the size of the economic return claimed in the article.

The expiration dates on food are slightly more scandalous: the FDA doesn't require the same level of testing as they do on the medical side so they are mainly set stupidly short. Last week purchased some vacuum-packed lamb that had a manufacturer label with an expiration date a month away, but was prominently labeled to expire this week. And of course US egg producers take steps that reduce the storage time of eggs, which can be months old when you get them -- and then "expire" a week after getting home.

> Last week purchased some vacuum-packed lamb that had a manufacturer label with an expiration date a month away, but was prominently labeled to expire this week.

Are you sure this was an expiration dates, and not a "sell by" or "best before" date?

A sell-by date is the date at which the grocery store will voluntarily throw out the product—not because it has gone bad, but rather because its unprepared appearance/texture will have changed enough that people will either avoid the product on the shelf, or will complain and return the product, even though those differences disappear once the product is prepared. I'm talking here about things like "bruised" bananas, "slimy" baby carrots, raw chicken that's turned a slightly deeper shade of pink, and anything that "smells off" when taken out of the package. The grocery store can't make a profit keeping these products on the shelf past a certain point—even though they're perfectly fine and safe to eat—so it does some Business Intelligence stuff to optimize the profit-curve given (shelf-space opportunity-cost + complaint PR-cost + return-cost), and out comes a number they print on the label for when they should remove the thing to replace it with a newer one if they want to make the most money.

You might think that this makes for a lot of waste. But grocery stores are aware of this waste, and how it translates directly to lost profit opportunities, so they're usually on the hunt for ways to solve this. This is why so many groceries have in-store delis that make ready-to-eat meals: it's a way to turn products people aren't willing to buy, into products they are, by doing the preparation step themselves.

Often, also, grocery conglomerates have "up-market" and "down-market" marques they operate under. Down-market customer bases have higher tolerance for these "red-herring signs of spoilage" (probably because they've at some point tried these foods out of necessity and found out they were fine after all), so the conglomerate can take advantage of this when building a centralized logistics pipeline for its stores: when it acquires produce/dairy/meat/etc., the stuff that already has some of these signs (or will likely acquire them sooner) is sent to the down-market-branded stores, while the up-market-branded stores get the products that appear fresher and will continue to appear fresh for longer. This is a necessity—if they didn't do this, they wouldn't be able to keep the up-market stores in stock given said stores' customers' intolerance for signs of spoilage—but it conveniently allows the up-market store brand a cachet of having "the freshest" produce, allowing them to justify charging more for food from the same farms.

And, finally, some grocery chains have deals with e.g. homeless shelters, to provide them these voluntarily-expired products as ingredients for free-meals programs. Grocery stores get people to feel good with "can drives" for food banks, donating non-perishables, but perishables are in much greater demand, and it's often the stores themselves that are doing the most good there. (Why don't they talk this up for PR? Because the fact that this is "expired" food would have to come up, and the public will never bother to read and understand what I just wrote above.)

> (Why don't they talk this up for PR? Because the fact that this is "expired" food would have to come up, and the public will never bother to read and understand what I just wrote above.)

The deals for grocery chains are usually made with tax deductions as an incentive, as there are logistical costs associated with it, and they still dump/grind a very large volume of perishables every day. Talking this up for PR could be detrimental to grocery chains, as theirs clients might not appreciate the "inefficiency" of their distribution chain, as they are the ones paying for the donated/dumped perishables.

> Are you sure this was an expiration dates, and not a "sell by" or "best before" date?

In the case of this lamb (sold by Trader Joe's FWIW) both were labeled "use before". Hmm.

> And of course US egg producers take steps that reduce the storage time of eggs, which can be months old when you get them -- and then "expire" a week after getting home.

I've yet to have a US carton of eggs go bad, and they often spend a few weeks in my fridge.

Refrigerated, store-bought eggs may be alright for 6-7 months according to this experiment[1].

(Note: This is the only practical experiment on this subject that I am aware of. If anyone knows of a better study, please share the link.)

[1] http://www.motherearthnews.com/real-food/how-to-store-fresh-...

This is why I put "expire" in quotation marks -- it is bogus.

Eggs can indeed go bad, but not immediately, and last a lot longer if the cuticle is not washed off.

I've stored eggs for 2 months out of the fridge and they did not go bad.

As far as I know eggs will not go bad at all - they will just slowly dry out.

One time I accidentally opened an egg that an absentee roommate-landlord had left at a girlfriend's house. They had expired 14 months before, and inside the one I cracked was a very foul smelling blue moldy glump.

A silicon valley startup by the name of SIRUM (http://www.sirum.org/) has done some pretty incredible work in this space. They re-distribute medicine nearing it's expiration from hospitals/pharmacies/etc. with surplus to areas that need it and can use it before it expires.

I'm sure they do more - I only have a slim memory from hearing one of their founder's speak.

They do have a pretty incredible impact and deserve mention in this conversation - they've been attacking this problem from a "what can we do now" perspective for the last 5 or 6 years while others have been spending time debating what to do or if anything really needs to be done. They are good people.

SIRUM Co-founder here! Super pumped and happy to chat with anyone!

We help orgs donate surplus meds that are often short-dated. We're tackling the $5B of unused, UNEXPIRED meds that go to waste every year in the U.S.

Another facet to this is "Medication Adherence" which is defined generally as taking medication properly, but which the drug industry defines as taking all of your medication, and getting a refill for more. Speaking only as a lobbyist who has (regrettably) worked for a Pharmacy Benefit Manager in the past, there is tremendous amount of money and thought being put into how to get patients to take all of their medication. The belief is that this will lead to patients refiling that prescriptions, oh and probably help them overall, but refills are the number one priority for the drug industry.

Along with some other people in the healthcare industry (process person at a large provider, product person at a large EHR), I'm trying to make a map of the "systems truth" of healthcare in America.

We'd love to talk to you (in confidence!) about how the PBM part of the world works. Please reach out to me if you'd be interested - contact information is in my HN bio.

Most of the time the drug handles time well but the excipient degrades. This makes it very hard to control how those drugs will work. It opens so many variables.

It is like not using the seat belt could save your life if you are thrown away of the car at some specific circumstances, like happened to a person I know. But they are there because it works better most of the time as engineers could design safety with less variables.

My father had an emergency immune problem(that we have already identified as it is recursive) and the only required drugs in house were expired for some years. He took those before we bought the new ones and the old ones were like 1/3 potency of the new ones.

Given that most of the price of drugs come from intellectual property and patents and each pill cost dollar cents to make, I don't see the urgency of taking expired drugs.

If Hospitals trow away expired drugs then it is a good reason to take into account in the global negotiation process, and I bet they already do in countries that buy drugs in bulk.

I'm glad people are contacting the FDA to try to get them to extend drug expiration dates, but this article gives me the impression that after the FDA fails to reply, or replies non-committally one way or another, people give up. Have any of the efforts involved tried to move up the chain of command to Congress? The FDA might not seem to be willing to move on the issue, but if one could get one's senator or representative to care, that congressman probably has more means to get the FDA to act than the average person.

The cynic in me suspects that if this gains too much traction, drug manufacturers will simply start adding self destruct compounds to all of their formulations "for our safety".

In fact, some of them have track records of such "ethical" behavior, it wouldn't even surprise me if they turned them poisonous vs simply inactivating them.

The only solution is to fix the economic incentives so that drug manufacturers want to make less wasteful medicines. Same with preventing planned obsolescence. The only solution is to encourage more competition.

Or we could just make it illegal. Then again, with today's political climate it wouldn't surprise me if someone was unironically opposed to regulating drug manufacturers.

Laws are merely barriers to work around, incentives guide what's going to be attempted. Making something illegal adds some negative incentive, but isn't nearly as good a solution as simply having positive incentive to do the right thing.

What makes you think laws aren't incentives? "Merely barriers to work around" kind of misses why laws exist in the first place. Making something illegal is what society should do if we collectively agree that behavior is to the detriment of society without any positive gain. What do we gain by allowing drug manufacturers to put false expiration dates on their products?

As I just said laws are negative incentives, your response makes no sense, care to try again?

Require manufactures to send back fresh pills (at their own expense) if you mail them expired ones.

In Denver there is a program that supplies hospitals internationally with expired drugs and equipment that would otherwise be without. It's been going on for decades and saved many lives.

I'd think there's a business opportunity in recycling the still-active chemicals in those pills.

A lot of cost of drugs comes from R&D, quality control etc. The chemical synthesis itself generally isn't that expensive. And expired pills would be a quite unpredictable precursor, which isn't very desirable in the pharma industry (or anywhere else for that matter). In other words, I doubt you could make much of a profit (if any) from purifying non-decomposed chemicals from pills.

I wonder if you could get around drug patent laws with the first-sale doctrine.

I always assumed that the expiration date was partly legal, in the way that you only have 1 year within which you can legally have the drug in your system.

I mean, if I am prescribed oxycontin, can I legally have the drug in my system for the rest of my life?

In the United States, you can be prosecuted in some states for drug possession if your prescription has expired and you still have the drugs [1]. This has not been sufficiently tested with the Supreme Court so you could eventually win on appeal but not before spending lots of money and time behind bars.

[1] https://edfolsomlaw.com/2013/01/illegal-possession-of-your-o...

> This assistant district attorney operates on the theory that if your prescription was expired on the date of the alleged conduct, you no longer “possessed a valid prescription” at that time and therefore cannot avail yourself of the affirmative defense.

Christ. That is insane. What are people supposed to do for medicines that are used only in the case of an emergency? I really wonder what goes through some of these prosecutor's heads sometimes.

> I really wonder what goes through some of these prosecutor's heads sometimes.


I'm pretty sure it's never illegal to have a drug in your system (besides something secondary like a DUI). It's only possession that's illegal.

Depends on the country. In Sweden it's illegal to have certain drugs in your system. E.g. if THC is found in your blood you can be convicted for a minor drug offense. We're sadly pretty extreme with regards to drugs though.

Is it illegal to have THC in your system or is having THC in your system admissible as rebuttable evidence of possession/use of illegal drugs? Either would support the outcome you describe, but they are meaningfully different.

E.g., with alcohol, the “in the blood is illegal” rule would criminalize auto-brewery syndrome occuring in those legally denied alcohol (e.g., due to age), while the “in the blood is admissible as evidence of use” would not, though those with the syndrome might need to present evidence of it to avoid charges.

It's the latter: (intentional) use of (illegal) drugs is illegal, and having THC in your blood is proof enough to convict you (even if you've consumed it in a different country where it's actually legal; there's precedent). Typically you'd just get a fine though, if you admit guilt and don't go to court. See § 1 here for the exact law: http://www.notisum.se/rnp/sls/lag/19680064.htm

Not really, there are lots of countries having "zero tolerance" policy laws which include ingesting and being under the influence of illegal drugs.

Ah, ok. I was thinking specifically in the US. My apologies!

I think it is legal in that the real expiration date is guaranteed to lie after the date on the box.

The myth is that "lie after" should be read as "equals", which is often far from the truth apparently.

I always assumed that the expiration date was partly legal, in the way that you only have 1 year within which you can legally have the drug in your system.

Are you implying that you still hold this assumption and are looking for someone to convince you otherwise? If so, I can't even conceive of where you got this idea in the first place, let alone argue against it. Let's slap a little Occam's Razor on this: how would you even begin to enforce this?

According to my research, in America it's mostly an unknown (unenforced or depends on the state's sometimes conflicting verdicts) whether having a drug in your system means you are in possession of said drug. There are other laws that relate but I cannot pretend to understand the intricacies of which laws might apply to this situation.

But yes, I am interested in a definitive answer.

Laws around prescription pills are sometimes unexpected, like every prescription pill in your Mon-Tues-Wed-etc pill box (outside their prescription container) being a felony. So my grandma has numerous unexpected felonies in her purse.

I thought it it was illegal to have certain controlled substances past the prescription. So if you were prescribed a month's supply of oxycontin and got pulled over with it on your passenger seat a year later, you were breaking the law.

I hope not because my doctor prescribes me with a "30 day" supply, as wrote on the prescription, which he tells me to use over 3 months. Either way, the unknowns about potential felonies scares me...

At one time we were required to keep an extensive list of emergency drugs at work.

Off the record, it was generally accepted that we should never, under any circumstances, use any of those drugs, bar adrenalin, for fear of doing more harm than good.

The cost was substantial and one company seemed to have the monopoly on supplying them. They also kept a record of the expiry dates and supplied replacements as stocks went out of date.

The most annoying thing was that they seemed to deliberately supply drugs with most of their lifespan expired.

They of course denied that this was a policy.

I was once sedated for a procedure, and heard the Doctor ask for a certain type of needle. The Nurse replied that they were out; she had just thrown them all away since they were expired. The Dr asked "how expired", and she said "2 weeks". He replied that that was fine, he wanted them anyway. And then I heard the lid on the trash can opening...

I found it quite humorous; I couldn't care less that their supplies were a bit 'expired,' nor that the sealed packages had sat in the trash can for a bit!

Allowing drug makers to set expiry dates is a huge conflict of interest, they have massive incentives to expire their drugs as soon as possible. I'm surprised the FDA doesn't do the testing and set the dates itself.

The best solution is to require drug makers to replace expired drugs with new, for free. Given that manufacturing costs are typically a tiny fraction of sales price, this is not an expensive warranty. This will also give them incentive to make expiry dates as reasonable as possible.

Due to complexity it doesn't seem realistic to try to save money by using legacy drugs.

On the other hand, the economics of pharma industry are such that manufacturing cost is usually a very small proportion of price.

Therefore a better solution would be to introduce a mandatory new-for-old trade-in policy. So they wouldn't lose money on the deal, pharma companies would be rewarded at mfg cost. I.e. sans profit, marketing or research cost.

Great idea, never going to happen.

Why would they sign up for that when they can just as easily charge full price for those replacement drugs (as they do now)? The only people out there who can change this are making huge profits from the system currently in place, so why would they?

There might be some up-side in terms of reputation gain, but it would have to be imposed by a regulator.

Degraded tetracycline has been documented to be dangerous in those who are suffering from renal failure.

There's also the possibility that a person may be suffering from a new undiagnosed condition, which hadn't yet onset when they were initially prescribed the drug. In this case you could think of it less as an expiration for the drug, and more of a suggestion to seek additional medical guidance on continuing after a certain date.

There's also the remote possibility that a new drug is developed which could potentially become dangerous when expired. The general population would have to unlearn little morsels of knowledge such as this, in the meantime people could be harmed.

Edge cases like this are enough in my opinion to warrant not spreading blanket advice like this, even if it's nearly always true. Erring on the side of caution is the best approach with medical affairs, even if it costs a bit extra monetarily. Giving patients potentially dangerous advice so they can save money is ethically questionable at best.

3 cases in 50 years? That's all you've got?

I would be more concerned about the second point. A patient developing additional complications isn't unheard of.

Read similar article on HN before. U.S. Army was able to save millions after FDA okays them to use expired drugs.[1]

[1]: https://www.thepharmaletter.com/article/fda-tests-let-milita...

yes. For the scientists in OP to say they were surprised, they completely failed sat lit review.

It seems that a law simply allowing private testing of drugs for efficacy, the same way the federal government is allowed to, would go a long way to saving money here.

Hospital systems and HMOs are large enough to save millions for a bit of testing, and once enough data is gathered they can easily and safely adjust their retention policies.

Good idea but I suspect, the FDA won't give up their ability to pick winners and losers anytime soon.

Hospital systems and HMOs turn their inventory fast enough that inventory control can probably handle their problem.

Small hospitals and pharmacies probably suffer more from expirations.

Aren't there many drugs that hospitals need to have at hand, but are rarely used? Things like antivenins, anti-rabies serums, or other agents that are only reached for in extreme cases. It may not be a high proportion compared to what hospitals are using on a daily basis, of course (but they might be proportionally more expensive, too).

As far as antivenins, it is my understanding that most hospitals don't keep them on hand for the most part, and instead rely on an "antivenin bank" to fly it in when needed. Interestingly, the main bank in the US is run by the Miami-Dade Fire Rescue [1]. They have a 24 hour hotline and will arrange to have antivenin immediately flown to anywhere in the world.

[1] http://www.miamidade.gov/fire/about-special-venom.asp

That's very interesting, thanks!

> Hospital systems and HMOs turn their inventory fast enough that inventory control can probably handle their problem.

Doesn’t sound like it from the article...

The "extended use dates" that Pfizer published mean that we now have 'older' drugs with expiration dates that are past the expiration dates of newer batches. They didn't just say "add 1 year to all expiration dates" they said "for these specific lot #'s, the new expiration date is: XYZ".

They only published extended dates for some lots, since they expect more recent lots to still be in date by the time the shortage is addressed.

I may have spent the better part of a day a couple weekends ago relabeling a bunch of vials with a sharpie...

Drug Expiration Date Excuses

FDA requires testing. testing expensive... really ... how hard would it be to grab 50 tablets per drug for each environment (high UV, extreme humidity etc..) let them sit and once a year grab a pill and run chem analysis. Sounds like it could pretty much be automated, except for maybe the maintenance of the environments. Maybe 500k to 1m per year with the work out sourced with various statistical validations. (Yes I realize this is hard part, damn humans) Take all the generic tablets and voila..

Drugs are cheap to produce

Really though, medicine doesn't cost much. It costs because society deemed it so. As it's been explained above/below, but, paying for research, marketing and regulations is probably 90% of the cost.

The hard question is

A more interesting question is how much can we afford to help others when they are sick? If the difference between health is as trivial as getting the right medicine I know I'd feel guilty if I didn't do something. This is especially true for chronic disease, as the cost benefit is tilted.

Seems like an easy fix would be to require the manufacturers of patented drugs to replace, free of charge, any unused expired drugs.

Incentivizes the manufacturers to accurately measure expiration times and the marginal cost to them of replacement is much less than to the consumer.

Can't there a company which buys up all the "expired" drungs and get the appropriate approval from FDA and resell it for a cheaper price?

What's going on with our healthcare system that pharmaceutical grade baking soda is being rationed?

W T F.

More discussion here: https://news.ycombinator.com/item?id=14798615

It seems the problem is that the suppliers' supplier had a problem supplying a component (not the bicarbonate itself, but the vial it comes in, I think) and this caused production to come to a halt while the issue was being sorted out. Presumably this particular component has been certified in some way and replacing it with an alternative requires recertification, though the news articles merely say that the companies could not divulge additional details.

That's the proximate explanation.

I wonder more about whether such a fragile supply chain is inevitable in a mostly capitalistic society or whether we could do more to ensure stability.

It's the generic drug business. Competition is purely though price. That squeezes margins significantly, even for more capital intensive product like injectables. If you're a manufacturer who has a sudden quality control issue, it often makes more sense to just shut down production than try and fix it and lose a ton of money.

It seems that when it comes to business consolidation, what is good for the goose is not good for the gander (apparently there are 2 suppliers left in the US).

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact