EPA To Limit Science Used To Write Public Health Rules (nytimes.com) 273
An anonymous reader quotes a report from The New York Times: The Trump administration is preparing to significantly limit the scientific and medical research that the government can use to determine public health regulations, overriding protests from scientists and physicians who say the new rule would undermine the scientific underpinnings of government policymaking. A new draft of the Environmental Protection Agency proposal, titled Strengthening Transparency in Regulatory Science, would require that scientists disclose all of their raw data, including confidential medical records, before the agency could consider an academic study's conclusions. E.P.A. officials called the plan a step toward transparency and said the disclosure of raw data would allow conclusions to be verified independently.
The measure would make it more difficult to enact new clean air and water rules because many studies detailing the links between pollution and disease rely on personal health information gathered under confidentiality agreements. And, unlike a version of the proposal that surfaced in early 2018, this one could apply retroactively to public health regulations already in place. [...] [The draft] shows that the administration intends to widen its scope, not narrow it. The previous version of the regulation would have applied only to a certain type of research, "dose-response" studies in which levels of toxicity are studied in animals or humans. The new proposal would require access to the raw data for virtually every study that the E.P.A. considers. "E.P.A. is proposing a broader applicability," the new regulation states, saying that open data should not be limited to certain types of studies. Most significantly, the new proposal would apply retroactively. A separate internal E.P.A. memo viewed by The New York Times shows that the agency had considered, but ultimately rejected, an option that might have allowed foundational studies like Harvard's Six Cities study to continue to be used. Harvard's Six Cities study is a 1993 project that "definitively linked polluted air to premature deaths" and is "currently the foundation of the nation's air-quality laws," the report says.
"When gathering data for their research, known as the Six Cities study, scientists signed confidentiality agreements to track the private medical and occupational histories of more than 22,000 people in six cities. They combined that personal data with home air-quality data to study the link between chronic exposure to air pollution and mortality. But the fossil fuel industry and some Republican lawmakers have long criticized the analysis and a similar study by the American Cancer Society, saying the underlying data sets of both were never made public, preventing independent analysis of the conclusions."
The measure would make it more difficult to enact new clean air and water rules because many studies detailing the links between pollution and disease rely on personal health information gathered under confidentiality agreements. And, unlike a version of the proposal that surfaced in early 2018, this one could apply retroactively to public health regulations already in place. [...] [The draft] shows that the administration intends to widen its scope, not narrow it. The previous version of the regulation would have applied only to a certain type of research, "dose-response" studies in which levels of toxicity are studied in animals or humans. The new proposal would require access to the raw data for virtually every study that the E.P.A. considers. "E.P.A. is proposing a broader applicability," the new regulation states, saying that open data should not be limited to certain types of studies. Most significantly, the new proposal would apply retroactively. A separate internal E.P.A. memo viewed by The New York Times shows that the agency had considered, but ultimately rejected, an option that might have allowed foundational studies like Harvard's Six Cities study to continue to be used. Harvard's Six Cities study is a 1993 project that "definitively linked polluted air to premature deaths" and is "currently the foundation of the nation's air-quality laws," the report says.
"When gathering data for their research, known as the Six Cities study, scientists signed confidentiality agreements to track the private medical and occupational histories of more than 22,000 people in six cities. They combined that personal data with home air-quality data to study the link between chronic exposure to air pollution and mortality. But the fossil fuel industry and some Republican lawmakers have long criticized the analysis and a similar study by the American Cancer Society, saying the underlying data sets of both were never made public, preventing independent analysis of the conclusions."
Bureaucracy and red tape (Score:5, Insightful)
Eh, who needs to be able to breathe and drink clean water anyway, right?
Re: Bureaucracy and red tape (Score:3)
On the flip side, industry studies showing thereâ(TM)s no problem with their products must also give up all the data if they want their studies to influence policy.
Re:Bureaucracy and red tape (Score:5, Informative)
The scientific methods currently used by the EPA are meaningful and open according to scientific community norms. The methods and results are published, other teams repeat the studies, or approach the topic from a completely different perspective, and get the same results. So no, it is not even more crappy and scary than corporate greed stifiling scientists through bureaucracy with conflicting requirements.
Re:Bureaucracy and red tape (Score:4, Informative)
Tell me, have you ever had any medical treatment at all? Because if you have, at least one important aspect of it, and normally many more, will have been based on research where underlying data is not published. That is standard in medical science, and always has been. I mean, literally the whole fucking concept of a double-blind RCT is that even the researchers and the patients don't know who gets the treatment and who gets the placebo.
Re: (Score:3)
I can't tell what's worse: the fact that the EPA / Trump administration is planning on doing this, or the fact that some people on Slashdot are stupid and/or venal enough to argue it's a good thing. The options as I see it:
1. They're dumb
2. They're not dumb and are trying to convince others who they think are dumb enough to believe the dumb reasoning
3. They're not dumb, they're not trying to convince others, and instead their objective is to make everything feel like it's up for grabs. Wiping their pus-infe
Re: (Score:3)
Re:Bureaucracy and red tape (Score:4, Interesting)
Sounds like a way to combat tyranny of "scientists" who try to change entire segments of society based on their belief
The oft-quoted Eisenhower "Military-Industrial Complex" speech, is only quoted w.r.t defense spending... however he also talks about the "Scientific-Technological Elite" [yale.edu] in the same speech.
An excerpt:
Today, the solitary inventor, tinkering in his shop, has been overshadowed by task forces of scientists in laboratories and testing fields. In the same fashion, the free university, historically the fountainhead of free ideas and scientific discovery, has experienced a revolution in the conduct of research. Partly because of the huge costs involved, a government contract becomes virtually a substitute for intellectual curiosity. For every old blackboard there are now hundreds of new electronic computers.
The prospect of domination of the nation's scholars by Federal employment, project allocations, and the power of money is ever present and is gravely to be regarded. Yet, in holding scientific research and discovery in respect, as we should, we must also be alert to the equal and opposite danger that public policy could itself become the captive of a scientific technological elite.
I suspect that those that oft-quote Eisenhower for his wisdom will now disagree completely.
Re:Bureaucracy and red tape (Score:4, Informative)
Re: (Score:3)
time-code 9:55 if you are in a hurry
Nobody brings up his warning about that (Score:3)
If I warn you a train and a car are coming at you, and the car doesn't come but the train does, you move out of the way of the train and forget about the car.
Re: (Score:2)
Oh and also as for your NY Post example? Trying to build a case for your 'argument', such as it is, based on one news story is pretty weak.
Re: (Score:3)
Plus: skeptics are skeptical about parts of science that can reasonably be agreed to be not yet settled. Deniers try to make out that parts of science that are no longer up for debate, are indeed so. They make grandiose and vague claims about "science is never settled" but somehow they're never talking about the science of, say, whether lenses can refract light, or whether lithium is less dense than lead, or whether apoptosis exists, or whether electrons have an electrical charge. They'll happily accept all
Re: (Score:3)
Since when is providing the data used to support your conclusions considered "bureaucratic red tape"?
The data isn't the red tape. Obtaining permission to use identifiable data from patients is what creates red tape.
Are there any peer-reviewed journals that publish 'science' that doesn't allow researchers access to the underlying data?
Erm. You have that bass-ackwards. Can you find a single medical journal that routinely provides identifiable information about patients? No, because it is absolutely fucking obvious to nearly everyone -- though apparently not you -- that there is zero good reason to identify a set of patients when publishing a clinical study, and many reasons why that is a foul, unethical and dangerous i
This will be fun (Score:2, Troll)
I predict there will be significant overlap between the people who wharrgarble about gubbermint identifiers [zerohedge.com] and the mark of the beast and the people who will praise this attempt to violate medical privacy.
This is a long-term trend on the right. (Score:5, Interesting)
This has been happening for a LONG time with right wing legislatures in the US.
Heck, way back in 2007, I went to a literal skeptics convention - The Amazing Meeting, and that ended up being the theme of the year. Hosted by the now-defunct James Randi organization, which was largely libertarian, but with all political leanings among members- the focus of that year was on the confluence of politics and science.
Aside from the usual folks of that era, with folks like the Mythbusters and Penn from Penn and Teller, there were lots of elder statesmen (and women) scientists - with a focus on both what goes wrong with science as a process.
The overall message was that while science is still being our best way to know things, folks everywhere still fool themselves, including in the most rigorous scientific circles.
But more than that, there was the definite trend that right-wing politicians in the US at least have long had a history of undercutting all aspects of science that they could get ahold of. All forms of advisement, all forms of systematic shared knowledge, basically everything science-related conservatives of recent eras have been able to eliminate they have - and not just for religious reasons.
This wasn't coming from a place of hatred of conservatism there at all - but it was a valid and maximally skeptical set of facts that repeated themselves over and over, in each administration.
Something about modern conservatism just has a target drawn on all aspects of science, and it's really a weird expression of an agenda that shouldn't see any benefit to that - but does so at every opportunity.
Ryan Fenton
They bent the skeptic community (Score:3)
He did several skeptic videos that reached millions of views and briefly made a good living off it. Then YouTube changed the algorithm and the money dried up. The right wing, seeing a large, valuable community of angry young guys, started throwing money at them. They'd buy advertising on their channels, fly them out to events, fund their patreons, etc, etc. They still lost their YouTube adverti
Re: (Score:2, Interesting)
Re:This is a long-term trend on the right. (Score:5, Insightful)
The social conservatives don't want any inconvenient facts challenging the mythology around which they base their worldview, so there goes huge chunks of biology, cosmology, ecology, and critical thinking in general. Then the business side of things has a problem with science when pesky regulations cut into their short term profit by focusing on long term health/environmental consequences.
Modern conservatives have set themselves up to, for ideological or financial reasons, oppose science. And that's not to say there's never any problems from left-ish sources, but by and large it is the conservatives who have aligned themselves firmly against reality.
Re: (Score:2)
It's depressing, I don't fucking get it. What is up with that shit on the right?
Why can't there be a pro-gun, pro- (sane) single payer system, pro reasonable but not ridiculous taxation, pro-science, pro small but not too small government, pro-choice, anti-war party? Instead we get fucking screeching lunatics on both sides.
Re:This is a long-term trend on the right. (Score:5, Insightful)
Simple version: It's the voting system. You have to vote only for one of 2 parties here, otherwise, the system breaks down. That one-or-another system makes room for ALL KINDS of shenanigans historically, and that's coming to a head even more now.
Most any other voting system, you can get more parties - and those parties will compete to represent the voter much better. No 2+year-long election cycles, no benefit to district rigging (gerrymandering), no electoral college nonsense.
Ranked choice voting, proportional representation systems, all kinds of systems could work - but it's also been classically difficult for voting politicians with power to see the system that gave them that power as flawed in any possible way.
But man... it would be much better to have a liberal party, a conservative party, a pro-science party, and so on. With something like Ranked choice, I could vote for the science party first, then the liberal party next, then just not even list the conservative. Then the science party could vote in coalition with the liberals - but pressure the conservatives to change their stances to get cooperation over time.
It would still be flawed, being politics and all - but nothing like the absurdity we've fallen to with someone like Trump coming out as leader of one of the two major parties.
In a sane culture, that would spell the direct end of this absurd political structure system we have now.
Fortunately, the upcoming generations do seem a lot more sane than the current ones - so at least there's the possibility of improvement after this horrorshow example of how bad our system is now.
Ryan Fenton
Re: (Score:2)
What you're describing isn't Republicans, Democrats, Libertarians or Liberals or anything of the sort.
The party you describe would be called simply Realists. And there are dozens of you. DOZENS, I say.
Re: (Score:2)
Why can't there be a pro-gun, pro- (sane) single payer system, pro reasonable but not ridiculous taxation, pro-science, pro small but not too small government, pro-choice, anti-war party? Instead we get fucking screeching lunatics on both sides.
Bernie? Not exactly for "small government" but that's mostly incompatible with single payer and other social safety nets anyway that you're probably also implying.
anonymized raw data (Score:5, Insightful)
But what I am guessing is that the GOP , the anti science party, want to force traceable data as they are fully aware that would breach privacy , thus they can pollute a lot and shit on the environment, with none of those pesky studies stopping them. Is there any reason any sane person does vote for the republican in the US beside wanting to stick one to the democrats ? I mean
Re:anonymized raw data (Score:4, Informative)
Usually anonymized raw data is available. What is not available is a way to track back to a person. Isn't this the case for those study that they have raw data?
There's many case studies where that's not true, but for larger policy-making studies like this the data is generally de-identified. This means that all obvious identifiers are removed but even something relatively innocent like "Male, 28, broken arm, car accident" can be relatively identifying if you know a 28 year old male who broke his arm in a car accident, because he was in a cast and told everyone he'd been in a car accident. Maybe in this particular study he's the only one with that particular combination. If you then add one more column saying whether he broke one or both bones in his arm you're now spilling medical information he may or may not have told people.
That is why so very little medical data is open to the public, at least here in Norway. You're granted permission to research on the condition that you don't attempt any reverse identification and you don't publish the line item data. Aggregations, correlations, pretty much any kind of summary is fine they're generally truly anonymous but the actual data material would have to go through a very tedious scrubbing process to avoid identifying any outliers and if you did that'd probably tweak the results slightly meaning anyone trying to reproduce your numbers couldn't get any exact matches. It's not just waving a magic wand and saying the data is now anonymous.
Re: (Score:2)
These large scale studies are largely allowed because researchers agree to anonymize the data. Subsequent studies that build on top of those studies have even less access to the raw data. To me it’s largely a way to put up as many obstacles as possible. If every single entity had to get formal consent with every single person providing medical data and that they could be traced, subpoenaed, public shamed, etc. they would be far less likely to allow any data collection.
Also setting up data collection
Re: (Score:2)
Re: (Score:3)
A good idea, but... (Score:2)
That's a good idea, but you need to extend it to include permission to introduce any drug or other chemical into the environment. And no grandfathering.
Re: (Score:2)
Re: (Score:2)
I wonder if Alibaba sells them. I bet they do.
You just missed the big sale [slashdot.org]
But if you hurry, you can still get some before stock runs out. [slashdot.org]
pre-school level thinking (Score:5, Insightful)
More seriously, listen up republican states. If you know what's good for you, you'll push back on this crap hard. You have any idea how much the Flint situtation cost the state? 400 million big ones. Some from state tax money, some from fed tax money. Pollution creates hard dollar costs that need to be covered eventually. That comes out of you're pocket. It's far, far, FAR cheaper to keep reasonable regulations in place than to clean up a toxic dumpster fire after it happens.
We've known this for a long time, but as a civilization it seems like the past 5 years have been a collective rejection of a bunch of wisdom that we paid a hard price to gain. It happens occasionally. The next decade or two, make some popcorn and watch as we re-learn some very, VERY painful lessons. There are gonna be more Flints (due to stupid regulatory rolllback) and wars (due to idiotic foreign policy) and and recessions (because conservative tax cut's and protectionism don't help the economy). And in the end, it will all have been avoidable it we had just listened to the people who we KNEW were smart but as a group we just didn't feel like being told what to do at the moment.
Not a bad regulation (Score:2)
As much as I wanted to hate anything that enables watering down of science in policymaking, this is really only about data transparency and scientists being forced to make their data public. That doesn't sound like a bad thing, and it's surprising that the US would allow policy to be informed on a scientific basis without evidence provided for scrutiny.
The data would of course need to be anonymized, but all good scientists should be doing this already - it's hard to see how any ethics board would approve of
Revolt against technocracy (Score:3)
This is interesting. I'm starting to see this whole populism wave as a revolt against technocracy. In a technocracy, policy is basically made by scientists and computer models. Read Vonnegut's Player Piano.
That book ends in a revolution that destroys everything the scientists have built. Perhaps we need to face the fact that people don't like their choices to be taken away from them by science. In global warming discussion that means a clear line between the domain of science (what is) and the domain of the people (what we want to do about it).
Pretend there's no choice in either and you end up with Trump, or worse...
Re: (Score:2)
The book ends in a failed revolution: The people rise up and destroy the machines that automated most of the population into unemployment - and yet as soon as the destruction is over they set about rebuilding things exactly as they were before.
Re: (Score:2)
"as soon as the destruction is over they set about rebuilding things exactly as they were before."
And so it goes...
Not only that (Score:2)
It looks like they also are removing the 'E' and the 'P' from EPA:
Could someone explain to me why (Score:2)
Could please someone explain to me why we are not doing all we can to have clean are & clean water worldwide?
Money (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
Re:Having to show your work (Score:5, Insightful)
Why don't you start by uploading your complete medical history to Github and making it public?
Re:Can you say ... (Score:4, Insightful)
Can you say de-anonymization?
If it's anonymized, the same people complaining before will still complain that they don't have the 'true data' and still spread FUD anyway.
Re: Can you say ... (Score:2)
Anonymized data is used all the time in health research.
Re: (Score:2)
Or how about you stop stigmatising people based on medical needs. Make every bloody medical record public.
Re: (Score:3)
If it's anonymized, the same people complaining before will still complain that they don't have the 'true data' and still spread FUD anyway.
Which is why you have the anonymization overseen by trustworthy and qualified people acceptable to (the reasonable among) the people complaining now.
Trustworthy and qualified people like the original scientists that the deniers were already complaining about?
Now you are back to square one.
(Even you admit it won't help with the 'unreasonable people' anyway. And would solve nothing.)
That deserves a better response (Score:2)
There is a difference between a denier and a skeptic.
Anyone who can't understand the distinction is likely a denier.
You said:
Which is why you have the anonymization overseen by trustworthy and qualified people acceptable to (the reasonable among) the people complaining now.
The reasonable people you mention could be skeptics.
The unreasonable people would be the deniers.
If someone is not satisfied, even with the trustworthy and qualified people. Then that person is not interested in science or facts and is clearly just a denier. Not worth the effort of trying to educate with facts or evidence or science, as nothing is likely to change their beliefs.
Re: Can you say ... (Score:5, Insightful)
To many people, sadly, science is just their new religion. They even have the term 'denier' to label those they consider heretics.
No it isn't. Science is the process of explaining natural phenomena using observation, data, logic and reason. I don't give intelligent design advocates, anti-vaccination believers or climate deniers the time of day nor am I willing to grant them equal time for the same reason I don't give flat-earthers the time of day. I can send a satellite into orbit and document using radar mapping that the earth is an oblate spheroid. I can document the sharp increase in global temperatures over the last few hundred years and show how it correlates with the industrial explosion that began in the 1760s and that it is directly related to carbon de-sequestration. I can observe evolution happening in microbes in a petri dish. Nobody gets taken seriously in science unless they can back up what they are saying with solid reason and data. There is no truthiness in science. Scientists are the only ones who are not afraid to say that they just can't explain something. It is only religionists who claim to be able to explain everything with bronze and iron age scrolls which at least earns them a tiny bit of respect since they at least have data, even if it is mainly valuable for historical and cultural research. Climate deniers and anti vaccine activists have literally nothing other than arguments they found after sticking their hand up their rectum and rummaging around in there for a while.
Re: (Score:2)
I can document the sharp increase in global temperatures over the last few hundred years and show how it correlates with the industrial explosion that began in the 1760s and that it is directly related to carbon de-sequestration
Sigh.
Correlation != Causation Temperature Records that have been "adjusted", Interpolated, Estimated, Guesstimated. Failed models Fail predictions.
Climate Science isn't any more Science than is all those other things you cite.
See that’s what I was talking about. You shoved your hand up your rectum and that is what you pulled out.
Re: (Score:2)
Actually it's quite the opposite. Science is based solely on testing and observing hypothesis with evidence. Religion is based on belief.
Deniers are literally the religion here.
Re: (Score:3)
Anonymisation is ineffective for these studies, because the data gathered is sufficiently detailed to allow re-identification of many research subjects.
Re: (Score:3)
Can we get this same openness from the corporations that want to sell us their goods. If we want this kind of openness for government regulations shouldn't we also demand this kind of openness for the products companies want to sell us.
Maybe big pharma should be forced to disclose all of their data for drug research before they get approval. If we can't duplicate their research that shows that the drug is safe how can we be sure it is safe?
Re: (Score:2)
Re:Can you say ... (Score:5, Informative)
Can you say "anonymization"? I KNEW you could!
Except that it's prohibited by the rules. Any studies involving clinical data will have to include the names of the patients in studies. Which is against HIPAA unless patients explicitly grant the permission to use their data in research.
Basically, RIP EPA.
Re: (Score:3)
Re:Can you say ... (Score:5, Informative)
Can you say "anonymization"? I KNEW you could!
I can say "supercalifragilisticexpealidocious", too... and that's just about as meaningful.
In a previous job, I wrote code explicitly to anonymize medical records for data use. My job was exactly what's described as being necessary here. Unfortunately, it turns out it's an incredibly hard problem, to the point that my company eventually gave up on trying to fully-anonymize the data and just secured everything to meet HIPAA, instead.
Ultimately, the problem with anonymization is that you're trying to make an individual able to blend in to a larger "bucket" of the population, while still making them unique enough to count as a data point for analysis, rather than a statistical approximation. That should be easily avoided by simply prohibiting small buckets, but humans aren't evenly-distributed enough for that.
For example, let's first consider simply dividing population by geography, so a given medical record can't be identified any more precisely than a given region. That's immediately problematic, as certain medical conditions might only have a few dozen cases in a county. More common conditions might let you reduce your region size down as small as a city* while keeping an acceptably-large collection of people that might be identified, but simply knowing case numbers isn't terribly useful if a researcher can't rule out other factors. They have to account for ethnicity, gender, religion, previous medical history, and other such factors, or they can't draw any meaningful conclusions that might draw policy. That causes more difficulty, as the combination of factors causes exponential increase in the chance of identifying someone. An Italian Jew living in Nashville with lung cancer, diabetes, and lead poisoning? There aren't too many of those around, but it's important to note how their medical history is or isn't different from someone in Boston sharing all other traits.
Let's be optimistic, though... Let's say you can somehow find an appropriate set of filters such that the group never reaches a small quantity. Thanks to the magic of statistics (that I've long-since forgotten how to do), that can be converted into an inverse filter set that most likely does not contain a given individual, which means you can infer that individual's personal information as well.
In short, it's damned near hopeless. To have any chance of getting meaningful results, you have to include enough variables in the data to show they don't matter... but that's also enough variables in the data to identify participants, no matter how you try to slice and dice it.
* Note: There's a city in Montana with one single resident. Any filter including city is not anonymous. There are zip codes that just barely cross state lines, so a mix of (state + zip code) is not anonymous. There are some rare conditions that have only a handful of cases in the United States, or even the world, so even filtering on a single medical item isn't necessarily anonymous.
Re:Can you say ... (Score:5, Insightful)
Thanks for taking the time to explain this. Unfortunately, I think it's very likely that the people agitating for the new rules see the inability to create truly anonymised records as a feature, not a bug.
Re:Can you say ... (Score:4, Insightful)
Re: Can you say ... (Score:3)
the example i use is "in a car accident in x city in y month"
nothing obviously identifiable there.
unless you know the person you are looking for was in a car accident in x city in y month and its then incredibly easy to extract the rest of their linked records, like were they diagnosed with hiv.
Re: (Score:3)
In what possible way is this a non sequitur? It's "what's sauce for the goose is sauce for the gander". It's a perfectly logical thread: AHuxley has said patients should give up anonymity for research purposes and whoever57 responded by saying "you first".
Re:Having to show your work (Score:5, Interesting)
Well the point of this...
would require that scientists disclose all of their raw data, including confidential medical records, before the agency could consider an academic study's conclusions
is to massively delay disclosure of data in a timely manner. Getting the needed permission to publish open records can take between ten to fifteen years to complete and stay this side of that law.
Publish in full, let everyone in the USA see the raw data and the science will stand
And that's fine, but there are privacy laws that regulate that. So you have to involve lawyers, you have to involve regulators, you have to have paper trails, you need auditors, you need clearinghouse staff, verification, independent review of privacy, and so on. It's not just a switch you flip and go "welp there you go, now you have all the data!" Medical records are very regulated and bring with it a load of red tape.
If the US science is correct, US and international peer review would be really welcomed
The data can be verified without needing to know names and other privacy related information. Peer review can be done without adding this additional red tape.
Re:Having to show your work (Score:5, Informative)
Private health records shouldn't even be collected for double blind studies, you should collect deidentified records from the beginning
Cool. However, that's not how it works. Doctor's aren't seeing patients and going "Well gee, I wonder what study I can get these people in?" Typically researchers have to ask hospitals for "ABC" criteria and because of liability regulation, they have to go, "Yeah here's the data, but you can't identify us." Typical data collection for a study happens post-facto, long after the patient has left the hospital. Tracking them down to get permission is a privacy violation, asking them to be open to whatever questions might come down the road is, again, a privacy violation, having the data tied back to a discoverable hospital that hasn't gone through a proper vetting for information collection is, you guessed it, a privacy violation. The ideal data that is anonymized just doesn't exist in vast quantities to make a study with.
if you want to make science-based rules, show the data
Again, you can show the data and not have to tie back to a hospital or an actual patient's name. You just don't need this additional red tape.
Anyone can p-hack a result in a paper
And usually those studies don't play out, lead to anything useful, or anything else. There are indeed people who go their entire life doing initials all their life. They don't usually do well in the field. But there again, paper IT admins exists too. Name me a field and I'm pretty sure we can all find an example of people faking it till they make it. But outside of the sensational news story they cause from time to time, p-hack papers don't lead to anything of any use. In fact they usually lead to the team lead being looked down on.
you can still do that but people will scrutinize more when you have to show work
People are stupid. People will not do that. Stop thinking that people are some sort of oversight on whatever. If you need any proof that people don't give a flying fuck, just look at how the US government has been running for like, IDK, the last six decades. People just want to be fed interesting things.
Re: (Score:2)
I do this for a living. You enroll them in a study at the doctors office in case of clinical research and get the consent right there. In most cases you pay them for extra visits. What you're describing post-facto is large demographic health studies like treatment effectiveness etc and then you use something like I2B2 to filter out identifiers.
It's really hard to use data collected post-facto, the primary problem being identifiable information left in nurses or doctors notes, hence why it's almost never don
Re:Having to show your work (Score:5, Informative)
Alas, the 'data' word might mean EVERY bit of the incoming data; for a bubble-chamber experiment,
that's a few petabytes of images. The rule is an attempt to install a roadblock.
Any attempt at using science research institutes a kind of purity-test.
When I was a researcher, we had to publish all our results, that's what the grant was for; but, all of the raw data? No way!
Finding meaningful items inside the data would be possible for the researchers, but rarely for outsiders
because there are so many trivial variables that are dealt with in so many uninteresting (not published) ways.
Research IS repeatable, by setting another team onto a similar project; but, the rhetoric
behind this proposal completely ignores that kind of repeatability, insinuating instead
that there is an unhealthy secrecy being implemented. There isn't.
It's just that we pull a few pearls out of oysters, and we do NOT display the many oyster shells.
Re: (Score:2)
So what you are saying is, "Trust Us".
I see.
Re: (Score:2)
It's a good thing that there hasn't been a rash of examples of de-anonymization .... oh wait. [simplicable.com]
Re:Having to show your work (Score:5, Informative)
The world and the rest of the USA gets to see the records used after anonymization.
Did you actually read the damn document [nyt.com] then?
In addition, EPA is requesting comment on a variation of this option that would apply only to data and models generated after the effective data of the final rule for this rulemaking. Another alternate option to proposed 30.5 would provide for tiered access to data and models that have confidential business information (CBI), proprietary data, or Personally Identifiable Information (PII) that cannot be anonymized and require that all other data and models be made publicly available if they are to be used as pivotal regulatory science.
Here let me highlight the concerning part. would provide for tiered access to data and models that have confidential business information (CBI), proprietary data, or Personally Identifiable Information (PII) that cannot be anonymized. Just so you understand, that's legal speak for "don't anonymize the data". Having a tiered plan means that there will be some that have access to data that cannot be anonymized which means to stay this side of legal, means mounds of lawyers and paperwork.
I sure am glad you want the data to be anonymous, because this rule change is the opposite of that want. So clearly you must not be for it?
Since when does good science need to hide from independent review, peer review, another chapter, a book, "independent review of privacy"?
Clearly everything this law is about is complete lost on you. It's either you aren't understand what 40 CFR 30 is talking about or you just don't care.
Re: Having to show your work (Score:2)
pretty sure it doesnt. "cannot be anonymised" is the name of the person who collected the data. published the paper or funded the research. Not the anonymised ids on any medical records they collected.
Re:Having to show your work (Score:5, Insightful)
Here's a medical record:
Name: XXXXXXXX XXXXX XXXXXX
Age: 32
Sex: M
Disease: Stage 4 lung cancer
That did not take 15 years to XXX the name.
To find out if it relates to pollution you'd also have to give your adress, and in order to take into account other known risk factors you'd need to give your employment etc, etc and suddenly, you're not so anonymous.
Re: (Score:3)
Here's a medical record:
Name: XXXXXXXX XXXXX XXXXXX
Age: 32
Sex: M
Disease: Stage 4 lung cancer
That did not take 15 years to XXX the name.
Thanks for providing a completely useless medical record which no one can gain any meaningful information from what so ever. Your "medical record" is not data, it's incomplete garbage thrown out as not being relevant to any possible study.
Re: (Score:2)
"would require that scientists disclose all of their raw data, including confidential medical records,"
They shouldn't have personally identifiable (i.e. "confidential") records in the first place, that's simply not needed. And providing raw data is perfectly reasonable, doubly so if they want to be taken seriously.
Re: Having to show your work (Score:2)
It provides for de-identification, which is common in health research.
Re: (Score:3)
No more using "confidentiality agreements" and "confidential medical records" to hide the raw data from later world wide peer review...
While we're at it, let's hear what all those executives in secret meetings with the head of the EPA, the White House, and the rest of this regime, have to say. No more claiming it's "privileged" communication or needed to hear "both sides" so policy can be set.
If what these people have to say is relevant and important, then let us hear it so we can make a decision whether it
Re: (Score:2, Troll)
We're already doing that. Google has maps of traffic and air quality via cell phones. It's not like getting data is hard to do, making a qualified result is another thing completely.
Previous administrations militarized the EPA and FDA for political or personal gain, because it has broad extrajudiciary powers, minimizing such regulation is good. You want the raw data if Trump is going to make a rule right?
Re: (Score:2)
Says the 'Anonymous Coward', who is presumably
unaware that his signature has blunted his point.
Mainly, when I look at people, I see that they're wearing
CLOTHES. Don't you? Of course people can and do
hide all sorts of things. Let he who is without a 'thing to hide'
support the silly statement, since Anonymous Coward hasn't.
Re: (Score:2)
"People shouldn't have anything to hide."
Said the Anonymous Coward.
It's not about what people have to hide, it's about what they have to lose.
Re: Having to show your work (Score:2)
Hehe, there are over 1,000 chemical compounds in coffee. But numbers are scary out of context.
Junk Scientists and Lawyers (Score:3, Insightful)
They want the raw data so that junk scientists, low grade hacks paid to lie, can re-interpret the data and then fight out that interpretation in court over 5 to 10 years, whilst polluting like the insane. You Seppos are in deep shite and make no mistake.
Re: (Score:2)
Tell that to psychologists [wikipedia.org], some of whom have MDs (... which is why this isn't off-topic).
Re:Junk Scientists and Lawyers (Score:5, Insightful)
Re: (Score:2, Informative)
Deleting the patient names and other PII from the raw data is not fucking rocket science.
Smarter people than you(a large number and growing by the look of it) know that it's not that simple.
Re: (Score:2)
Re: (Score:2)
Re: (Score:2)
high-paid hacks paid to lie, who initially interpreted the data?
Who's paying them to lie, if not the government ?
Re:Junk Scientists and Lawyers (Score:5, Interesting)
Sounds good, but that's totally bogus. Eyewitness accounts of a volcanic eruption can't be 'repeated', but DO
qualify as scientific observation. Why wouldn't you allow that?
The proposed rule isn't a precaution, it's a paralytic agent.
The suspicion of 'junk scientists' is best investigated by real scientists. Publication is a
good requirement, but show-me-all-the-raw-data is the way IBM fought
an antitrust suit some decades ago: they declared that all the experience of all their executives was
admissable as data underlying a business decision...
A few traincar loads of documents was read into evidence... thirty million pages
http://www.historyofinformation.com/detail.php?id=923 [historyofinformation.com]
Re: (Score:2)
Re: (Score:2)
Re: You know how deniers work (Score:2)
Re: (Score:2)
Keep the mod points coming folks (Score:3, Interesting)
Polls show 62% of Trump voters will support him no matter what. His lawyers literally argued in court [politico.com] Trump can shoot a man and it's legal. Trump diverted funds from military bases for his wall. That's a much, much bigger deal than the media or that corporate sell out Nancy Pelosi made it out to be. The Constitution is crystal clear, the House of Representatives controls the purse strings. Trump ignored them and got away it it. And that's before we talk ab
One last thing (Score:2, Interesting)
Trump's DOJ is going to have the SCOTUS strike down the Affordable Care Act, and with it the protections for pre-existing conditions. I've got friends and family who only have healthcare today because of that. If it goes away people I know will die. And if he gets reelected that will happen.
He's also planning on means testing Medicare & Social Security.
Re: (Score:2, Insightful)
Trump's DOJ is going to have the SCOTUS strike down the Affordable Care Act, and with it the protections for pre-existing conditions.
The people of my State have had pre-existing condition protections long before the Affordable Care Act .. hell, even before the Massachusetts version of the ACA.
...
Are you mad about it? Are you mad about the repeated failure of your own States government to enact such laws? Are you mad at yourself for not lobbying for such a law?
You say you are bitter and angry, but it only seems reserved for the most-removed governance problem in your life. Your vote matters far more in your State than in the Federal
Re: (Score:2, Insightful)
The list of States with these onerous 12-month exclusion periods written into law are:
California, Idaho, Kentucky, Maine, Michigan, New Jersey, New York, Vermont; Colorado*, Connecticut*, Maryland*, Mississippi*, Montana*, North Carolina*, North Dakota*, Ohio*, Pennsylvania*, Rhode Island*, South Dakota*, Utah*, Virginia*, West V
They'll pick you off (Score:2)
They were slowed down by their defeat with the ACA. They will be back. They _never_ stop. There's literally trillions of dollars out there that could be theirs.
State governments can't stand up to modern mega corporations. Hell, they knew that in the 1700s. [wikipedia.org]
Re: (Score:2)
Dictator Trump is a symptom of a deeper problem - a fundamentally flawed political culture, driven overwhelmingly by partisan concerns, and a public easily manipulated by appealing to base tribalism.
It's not culture, it's mechanics (Score:2)
It's not cultural, pro working class policies poll in the 70s and up. It's systemic.