The FDA fails to adequately screen or monitor products | Hacker News Reader

The FDA fails to adequately screen or monitor products

nytimes.com

74 pointsjs236 comments

There's a few things that are problematic about this opinion piece.

First off, many of the items being discussed are related to surgical treatments. It is very hard to impossible to do a double blind study of a surgical thing. What are you going to do? Open a bunch of people up and only do the actual treatment on half of them? I've heard of this being done in very rare situations, but the ethics of this are pretty edgy.

Secondly, there appears to be a mischaracterization of women being disproportionately affected by things. This is a pretty complex topic, but lets balance out the offered information by pointing out that prostate cancer (men) kills more people than breast cancer (women) and receives a lot less funding

https://www.telegraph.co.uk/news/2018/02/02/prostate-cancer-...

My personal experience with FDA regulation has been that they are being asked to do an impossible task, and not doing a great job of it. People close to me have been hugely (negatively) impacted by FDA decisions because they are going for a one size fits all solution for everything, and because they somehow fancy that they know more than doctors involved in a particular situation.

Doctors are biased too, that's the problem. Often times, knowing a lot about something will bias you even more because you fail to acknowledge or accept what you do NOT know or what conflicts with what you THINK you know.

Everybody's biased, and I've certainly met my share of biased doctors. If you have a specific complaint, each doctor brings their toolbag of golden hammers that they know about and rarely has the ability to look beyond it. It's one of the challenges of being a patient to wade through that.

The FDA isn't really going to be very helpful with this, and they labor under the additional burden of being much more subject to politics. Sure, they may stop the obvious quackery, but remember that general anesthesia used to be considered quackery, as did prescribing antibiotics for stomach ulcers.

What really bothers me about this is a) The editorial board, when writing such an important piece, is not identified on the actual article other than 'the editorial board'. Sure you can find out who is on the board but you don't know what parts they played in the write up. I hate this convention.

b) Chain of information custody issue. They cite this organization with a link to the page but there is nothing easily apparent on that page that would lead you to know how legit the actual underlying data (which they are quoting) is. At least on a quick check.

https://www.icij.org/investigations/implant-files/

It's almost as if it's a 'well we are the NY Times the paper of record so you can trust that we have thoroughly vetted all of this as well as how the actual data was compiled'.

This is not to say there isn't an issue. But it's easy for a writer to sit back and take pot shots (and find something to support that pot shot) w/o the scope of the problem other than 'and big number of people are impacted'.

The FDA regulates (or fails to regulate) many different classes of products (drugs, biologics, devices, human cell and tissue products, supplements, food, tobacco and nicotine products, etc). FDA's authority and its effectiveness is extremely variable across these classes.

My sense is that the traditional drug and biologic drug parts still work reasonably well, but the biggest problems lately are coming from devices and from human cell and tissue products (and with products that are being pushed into those categories). A lot of this is due to the ways the laws have been written (due to lobbying and legislative capture), but some of it is undoubtedly a result of regulatory capture and incompetence.

Another example of a government agency that has been captured and lost the will to govern

Congress has gradually eroded its authority. For example FDA can no longer do its own science but has to rely on the companies to do so. It has been stripped of any authority over so-called "natural" supplements so they can make unproven medical claims (e.g. "our homeopathic remedy will cure your cold") thanks to Oren Hatch being paid reelection dollars by Herbalife. Etc.

Don't blame the agency, blame congress.

They want a:

> global medical device registry where patient outcomes for all medical devices are openly monitored

A global open database of people’s most sensitive medical information, what could go wrong.

You can measure outcomes without specifics of who was impacted. The doctor actually as a bulwark in this case and only provides the data specific to the devices as a report. I.e. we performed 175 hip replacement surgeries using x device. During that time there was a median issue rate of x percent. This is within x range of standard. This would have the added benefit of evaluating the medical professionals against similar circumstances especially if you factors in removing outliers that have a higher risk profile.

Jesus! As someone who works in the Pharma industry that’s dam near impossible. We’d love to get outcomes data for our product post-approval. Getting that data is damn near impossible as physicians have zero incentive to provide it, even when something goes wrong.

Yeah, the current binary model of "proven safe" is utterly broken.

First, the product is considered 100% unsafe and criminal to use until going through some defined set of steps.

Then it is considered 100% safe forever, and no further knowledge needs to be collected.

Any sane system based on gathering knowledge and saving lives would be very different.

No one in the industry calls drugs 100% unsafe or 100% safe. That’s why there is a long list of side effects on every drug commercial.

A risk/benefit profile is what drugs have.

I'm talking about the legal/regulatory framework.

Even there, 100% is an overstatement, there are some minor and slow feedback mechanisms.

Oh, nice to see it's not just us. We build diagnostic devices, and getting some performance data (like how frequently our device provides inconclusive readings) out of some hospitals networks is a pain.

So provide an incentive. If Pharma has an obligation to gather that data, they can be required to incentivize physicians.

Then physicians will send fake data because they can't track many patients but still want money or just don't want to put the effort into tracking. Since they can't send back individual tracking info trying to filter out the false data will be nearly impossible.

Money distorts incentives in ways that are not always desired.

That's a mess. How do you measure patient outcomes of, say, using a radiology software package (which is typically designated a Class 2 Medical Device)?

By the way, every medical device will have a registry number that you can certainly use to browse for things like recall notices.

I think that gets to the heart of one of the problems in the current system, which is that Class 2 is hilariously broad and the distinction between Class 2 and Class 3 is also little odd... For example, I actually had no idea that metal in plastic hip implants were Class 2. I think the fact that metal in plastic hip implants are Class 2, but metal on metal were shifted to Class 3 points to something being really off in how the system is implemented.

Like I understand that metal in plastic has a long history of use and that in practice that means that they likely have lower risk (in the sense that there are probably less unknown unknowns and known unknowns). But it still feels a little weird, since it doesn't automatically reduce the risk of any specific product/implementation.

(For context, compression socks designed to prevent blood pooling in the legs are also Class 2... but all other medical compression socks are Class 1).

Medical device regulation has a clause where if your product is close enough to a product that is already approved there is a shorter regulatory process, that could apply here. There's a lot of grandfathering of medical devices due to those rules.

Oh yes I know (I work at a company making diagnostic medical devices).

I feel like that mechanism makes sense for somethings, but certainly feels kinda sketchy for other things.

Today must be an even-numbered day at the NYTimes.

On odd-numbered days we scold the FDA for demanding too much data and investigation, and so taking too long to let products onto the market.

It can be both - The FDA is inefficient when it does test a product causing the costs to increase, and also doesn't screen every product as well as it should.

It's also has separate legal frameworks for drugs and medical devices.

FDA doesn't test, drug companies do and then provide data. That's why they can cheat and lie as much as they do.

It should not be surprising that different writers have different priorities and opinions.

Would it not be more suspicious if the paper's content were monotone and homogeneous?

Yes, but today is an odd-numbered day at HN, so we scold the media for indecisiveness. Otherwise, it's for their consistent bias.

It should not be surprising that different commenters have different priorities and opinions.

Would it not be more suspicious if Hacker News' content were monotone and homogeneous?

Yes, but today is an odd-numbered day at HN, so we become meta like reddit. Otherwise we scold them for their memes.

We really need numerical estimates, even bad ones, of the number of lives saved versus cost for different levels of regulation in order to talk sensibly about all of this.

The CDC does an enormous amount of this already.

They research problems, then research possible effects of regulation, and then they advocate for that regulation. They then measure the effects. One example is smokefree regulations[1] limiting people's exposure to tobacco.

There are many nonprofits and other governmental organizations that are also researching these things heavily and providing estimates.

Just because you haven't seen the numbers doesn't mean no one is crunching and publishing them.

1. https://www.cdc.gov/tobacco/data_statistics/fact_sheets/seco...

Here you go

http://www.fdareview.org/issues/theory-evidence-and-examples...

These all seem to be pre-2000?

Thank you! There is a cost to process and bureaucracy. It is a balance between innovation and protecting the public. Like security and marketing, there is no end to how much regulation you can have because you can always regulate more. You can add more rules, more laws, more regulators, more auditors, etc. etc. etc.

It makes sense to rush new medical devices and drugs to market when there's no better option. For example the Reagan FDA removed the review process for new AIDs drugs during the 1980s epidemic. But the FDA should have scrutinized vaginal mesh more closely given that there are already so many effective forms of birth control.

People also need to remember that you can't pick and choose regulations. I feel like many readers of this article will be rightfully appalled, but conversely also want to see increased medical use of psychedelics and marijuana despite a similar lack of research on long-term effects.

Transvaginal mesh is for pelvic organ prolapse, not some sort of birth control.

> the Reagan FDA removed the review process for new AIDs drugs during the 1980s epidemic

Just a quick historical correction: Reagan administration did no such thing. The FDA would not release potentially helpful drugs that were untested or unproven even as 100,000s of gay men lay dying in such a horrific way. Gay activists begged and pleaded while Reagan did literally nothing to help. He didn’t even mention the words gay or AIDS until 1986. The outbreak was in full bloom by 1983. Act Up was a grassroots organization designed in part to pressure the government to allow the use of drugs with limited track record. There was a modicum of success as AZT was approved in 1987 following Act Up protests on wall street were over 100 people were arrested

I’m curious where you heard that Reagan fast tracked AZT and similar. Sadly, I wouldn’t be surprised if it was in some ‘legit’ history books.

Source: “On March 24, 1987, 250 ACT UP members demonstrated at Wall Street and Broadway to demand greater access to experimental AIDS drugs and for a coordinated national policy to fight the disease.”

https://en.m.wikipedia.org/wiki/ACT_UP

Reply on news.ycombinator.com