First, the expiry date on medications is not so much “arbitrary” as it is “this is how far we have tested it, we meet the legal requirements for a stability program and since we’d like to be able to make/sell more lots of this drug in the future, we will use this date so that the patients have to come back and get more”.
Second, by the time you hit your expiry date (2 years is standard, IIRC) your drug has not degraded 15%. Or 5%. At most, as required by law, it has degraded 2%; in fact, it must be within 98-102% (or 98.0-102.0%, I don’t have the USP/EP/JP etc in front of me) of it’s original potency. In addition to that, any harmful (to any extent) impurities that may have formed cannot exceed more than a given amount, which depends on the impurity, so while your drug can still be 99.8% active, you might have reached a 0.2% impurity value for something by the time you hit 2 years, and so you have to stick a one year expiry on it. Basically, the expiry date is as close to a guarantee that the drug is the same as the day you bought it as you can get (provided you stored it correctly).
Stability programs: There’s the initial research stability programs, which establish what kind of storage conditions the drug can be kept under, and long-term stability programs, which is what the OP is talking about. The first one will test the drug at 0, 1, 2, 3, 6 months (maybe more) stored at 40C/75% relative humidity, 30C/60% RH, 25C/60% RH, 5C/Ambient RH. This way they can tell you to keep it in a cool place, to not exceed 30C, or keep it in the fridge. The later studies are usually done at something like 25C/60%RH or 25C/Ambient RH unless it had to be stored in the fridge or freezer, in which case you get 5C/Ambient RH or -20C/Ambient (very, very rare!).
The requirements for the long-term stabilities will vary from drug-to-drug, but in essence, they are all the same. The drug must pass the same criteria at the end of the stability period as it did at the beginning. When a lot is manufactured and is tested for stability, it is fully tested at T=0. That means that potency, impurities, colour and appearance, pH, density (if applicable), dissolution profile (if applicable), microbiology/bacteria presence and more will be tested. Samples of that lot will be stored as above, and removed for testing at 1,3,6,12, (18), 24, (30), 36 months. The variations depend on what previous/expected stability results have indicated is probable. Some months only some things are tested, others get the full range, but potency is almost always the main point of investigation. The colour might change, but if there’s no reason to believe it’s unsafe, that might still be ok (Listerine has a warning on it’s label about it becoming cloudy… technically it’s a “fail” for physical appearance, but it has no affect on the product or its efficacy, and so it is allowed to continue to be sold).
Now, if something tests out of range, there is a whole investigation that happens to make sure that it is a product problem and not an instrument/analytical one (like dissolving in methanol instead of ethanol or something). An OOS (out-of-specification) during a stability study is not necessarily the end of the study, since there’s still a lot to be learned by continuing. Anyways, with all that compiled stability data, the risks vs benefits of a 12, 18 or 24 (or whatever) month expiry date are weighed, and a decision is made. So the product is manufactured, packaged and sold.
Here’s the thing; once a product is on the market, it will continue to undergo stability trials. I forget the frequency, but something like 1-2 lots of every drug manufactured at my former company was put into a 24 month stability cycle, in order to ensure that the product continues to be safe for the public. This is a legal requirement. This is also the way in which product defects are often found, and the triggers for recalls (people panic at recalls, but IMHO, they are kind of a good thing!)
North American law prohibits the repackaging and/or reselling of unused medication because there is no way to check that the product hasn’t been tampered with by whoever had it before. Even when a pharmacy chain returns boxes and boxes of unopened…uh…boxes of drugs because they are close to their expiry and they want a new shipment, they are typically destroyed (unfortunately).
If your drug has special storage conditions (fridge, keep away from light, etc), then I wouldn’t recommend using them much past the expiry date… keep in mind that a 1 year date usually means the drug passed at 18 and/or 24 months…there is a margin of error in this! Drugs which are more risky to use past their expiry dates are hormones (e.g. the Pill, HRT, etc), vitamins (e.g. cyanocobalamine, Infuvite), and antibiotics (penicillin, clindamycin, etc) and several of the new “biologicals” (Enbrel, Humera, etc). If the drug has no particular storage conditions (which is true of most tablets and capsules), then using them a year or so past the expiry date is probably OK (the drug is unlikely to have degraded), but there is no guarantee, because after about 36 months from manufacture, it is very, very unlikely to have been tested.
Somewhat rambling post, but I think that covers most of the questions brought up in the thread as of whenever I opened it. Feel free to ask more questions if you’d like. BTW; my qualifications are simply that I was an analytical chemist for two different pharmaceutical companies for about 4 years and did many, many stability assays!
[/Wanders off muttering about N*****3 and their request for 64 8-hour dissolutions in 10 business days]