[Kalium]

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.

[techthroway443]

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?

[ekianjo]

> 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.

[TheKarateKid]

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.

[Kalium]

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?

[TheKarateKid]

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.

[larelogio]

Are they?

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

[TheKarateKid]

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

[yomly]

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...

[EpicEng]

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.

[ekianjo]

Regulations don't want you to rely on models.

[Kalium]

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