Quality control has limits, if you only test some samples you are likely to miss issues. Especially as there is a wide variety of potential problems, and you can't easily test for all of them. That is why the processes themselves are regulated, and not just the products.

And the issues here are not just onerous paperwork. This is the company shredding documentation that would otherwise show that they had data that indicated potential issues, but which they ignored.

I think what you’re missing here is that generics are already approved with pretty large deviations in e.g. serum response from the underlying approved drug, but without re-testing.

So the current system is already producing bad-for-consumer outcomes at a rapid pace.

Can confirm, the generic for the adhd med a couple of my kids take is super obviously not even close to as good as the real thing. Took us a little bit to put 2 and 2 together the first time we got that instead of the "real thing", but that was definitely the problem. Of course we still end up with the generic a lot, what with the shortage, and just have to deal with it barely-working instead of being wonderfully effective, until we get to roll the dice again 30 days later and hope we luck into name-brand pills being available.

Makes one wonder about, you know, all the other generics.

Is this a time release drug, like Concerta?

Yeah. The generics use a different ER mechanism, which is surely part of the problem, but others report issues with non-ER versions, too, presumably due either to purity/consistency issues or different filler material, so I doubt the ER difference is the only problem—we did use name brand non-ER at first before switching to extended release, and that wasn't even close to as ineffective as generics of the ER. I've read (but not confirmed) that some adhd generics are also allowed to use cheaper-to-manufacture almost-but-not-quite-identical isomers that haven't been tested and may have significantly worse efficacy, though, no clue if that affects ours (or if that's even true, really).

I wouldn’t be surprised. Concerta (osmotic time release) is so much more effective than normal generics (non-osmotic time release) that they shouldn’t even be considered the same drug.

Yeah, our experience is something like:

    unmedicated << generic <<<<<<<<<<<<<<<<<<<<<<<<<<<<< name brand

in terms of perceived effect. They're sure as hell not close enough to be considered equivalent under any reasonable classification scheme (but I assume insurers really, really, really want the generic to count as equivalent and work to ensure it is so-regarded—and, anyway, availability's so spotty that we just have to take what we can get regardless).

As a patient this was also my experience.

> Especially as there is a wide variety of potential problems, and you can't easily test for all of them.

Such as?

Most quality issues in drugs -- purity, composition, dissolution rate, presence of mycotoxins, etc. -- are very, very easily testable.

What's more, most of the things that are not easily testable would not be picked up in a facilities inspection.

You don’t mention quality controls. What should FDA do if upon said random testing there are concerns? How can they and/or the manufacturer trace the causes without documentation?

> What should FDA do if upon said random testing there are concerns?

Destroy or return the entire batch, depending on the level of concern.

> How can they and/or the manufacturer trace the causes without documentation?

That shouldn't be the FDA's concern. The FDA's concern should begin and end with the quality of the actual drugs that are entering the US supply chain. The drugs are things in themselves and can be tested as such -- without regard to where they came from or exactly how they were made.

A random sample by definition isn’t the entire batch, so how would you define ‘batch’ without documentation? The FDA would force recall on the entire product entering the country? Until how many samples are cleared?

Pretty simple.

Most cheap generic drugs are imported by the containerload, so simply create an arbitrary "batch" from each instance of importation. If the manufacturer is decent and has its own batch number, then the label could read

"Manufacturer batch: XXXX FDA Import of Record batch: YYYY."

Drugs that are imported in more specialized ways -- e.g. by small-volume refrigerated air freight -- are, as a general rule, far more expensive and batches are correspondingly smaller. The above paradigm would still work, where each importation instance is, for the FDA's purposes, a batch.

Testing fees, which shouldn't amount to more than a few thousand bucks per import, can be the responsibility of the importer -- so as to dissuade importers from receiving small batch after small batch.

So you treat each subsequent container load as a new batch without a potential common cause, like manufacturing defects? Testing every single batch for any possible harmful defects?

Every import, which could be multiple container loads.

If there's a manufacturing defect, the manufacturer -- after losing a couple of batches or having them returned to sender, and having paid for multiple transit and testing costs -- should wise up.

In any case, this would absolutely protect the US pharmaceutical supply chain without pushing unnecessary and excessively onerous requirements onto manufacturers. How foreign facilities manufacture drugs should not be the FDA's concern; drugs are chemical systems and chemicals are easily amenable to testing as things in themselves.

So instead of manufacturers bearing the cost of quality control during manufacturing you expect every single import to be tested for all potential contamination? How would that work exactly? And without documentation of quality assurance, how would any contamination be traced by the manufacturer?

> How would that work?

It's not difficult. There are known quantitative and qualitative testing methods for finished drug products, and these usually cost anywhere from a couple hundred to a few thousand dollars. In other words: It's typically very cheap.

So say a company imports 1M doses, and say you're an FDA inspector. At the port of entry, you take a statistically valid sample from that batch of doses, and you subject them to those tests. (This could be at an FDA/Customs lab or a third party lab -- in any case, the importer pays.) If there's a problem, you either test additional samples, or dispose of the batch summarily with a note to the importer as to what you've found.

Testing every import shipment, instead of stationing FDA inspectors overseas, would do more to improve the quality and availability of drugs, and more to reduce costs.

As things stand right now, a lot of drug companies don't want anything to do with the USA simply because they don't want the FDA getting into their business, inspecting their facilities, etc. It is a great, and wholly unwarranted and unnecessary, burden.

> how would any contamination be traced by the manufacturer?

That's not the FDA's problem. A batch of drugs is a thing in itself. A metformin capsule is a metformin capsule regardless of who made it and how.

But of course there would be import records and any halfway competent manufacturer would be able to tie a shipment to a production run.

Absolutely not. Historically, time and time again the pharmaceutical industry has proven it cannot be trusted and that regulatory oversight is required.