Evcidence-based medicine is highly compromised

[Perspicuo]

Ben Goldacre: What doctors don't know about the drugs they prescribe
http://www.ted.com/talks/ben_goldac...hey_prescribe/transcript?language=en#t-492066




_______________________________
As are my typographic skills: "evcidence" (Evidence-based medicine is highly compromised).
Siht happnes

 

[T.G.G. Moogly]

Thanks. That was a great read. Scary too.

 

[barbos]

Yes this is scary but it is an old news too.
Soft sciences like biology are filled with statistically incompetent or even outright fraudulent "studies".
People need publications on their resumes and they literally hunt for flukes to publish and they get away with it.
Hard sciences like physics are different because there is not much variety in topics of research and every positive result will be tried to replicate. In biology every god damn molecule can be tried on these poor lab rats and subsequently published, which will not be even tried to replicated because everyone knows that replication will be a failure and waste of time. Somewhat similar situation in hot fields like material science, they too publish a lot of crap which never gets replicated.

 

[untermensche]

Yes this is scary but it is an old news too.
Soft sciences like biology are filled with statistically incompetent or even outright fraudulent "studies".
People need publications on their resumes and they literally hunt for flukes to publish and they get away with it.
Hard sciences like physics are different because there is not much variety in topics of research and every positive result will be tried to replicate. In biology every god damn molecule can be tried on these poor lab rats and subsequently published, which will not be even tried to replicated because everyone knows that replication will be a failure and waste of time. Somewhat similar situation in hot fields like material science, they too publish a lot of crap which never gets replicated.

There are many drugs that must produce predictable results or they won't be prescribed.

If a doctor prescribes a blood pressure medication they are going to measure the effects of that drug for a long period of time.

And all cancer drugs undergo constant testing in trials. Many cancer patients enter trials where drugs are compared and this is happening all over the world.

If a drug is not producing significant results compared to placebo then this is discovered and those drugs are not recommended in guidelines that are constantly being updated.

It is possible a cancer drug could falsely gain approval, but once approved it's effectiveness will be tested.

 

[fromderinside]

Yes. Its a problem doctors face in their professions, convergence between profit motive and keeping up with research, usually results in doctors depending on salespersons interpretations of current research as they consider buying drugs.

At another portal between discovery of potential drugs and exploiting them for market there arises a conflict between scientific motive and profit motive at the basic research level and again at the development of drug level.

Government tends to treat the medical process similarly to the military acquisition process which is pretty well defined. There are going to be abuses but the scientific aspects of drug research. However, the development processes are pretty well controlled and designed to thwart these until they reach the marketplace as prescriptions.

All the above being said I thought this tract from innovation.org might set a frame upon which discussions about what may be threatening about doctor knowledge of drugs and their development. "Drug Discovery and Development:
UNDERSTANDING THE R&D PROCESS" http://www.phrma.org/sites/default/files/pdf/rd_brochure_022307.pdf

 

[untermensche]

Yes. Its a problem doctors face in their professions, convergence between profit motive and keeping up with research, usually results in doctors depending on salespersons interpretations of current research as they consider buying drugs.

That can be the case in general practice, and it is a real problem, but Oncologists tend to closely follow guidelines.

http://www.nccn.org/professionals/physician_gls/f_guidelines.asp

 

[ronburgundy]

Yes this is scary but it is an old news too.
Soft sciences like biology are filled with statistically incompetent or even outright fraudulent "studies".
People need publications on their resumes and they literally hunt for flukes to publish and they get away with it.
Hard sciences like physics are different because there is not much variety in topics of research and every positive result will be tried to replicate. In biology every god damn molecule can be tried on these poor lab rats and subsequently published, which will not be even tried to replicated because everyone knows that replication will be a failure and waste of time. Somewhat similar situation in hot fields like material science, they too publish a lot of crap which never gets replicated.

The level of fraud in medicine has much more to do with incompatibility between private profit motive and honest science than it does the "softness" of the science.
Studies have shown that medicine and pharmacology (where profit motive is high) are the most fraud-ridden areas of science, whereas plenty of other areas of biology that do more basic (and not directly profitable) research do not suffer from such fraud levels.

You are correct, that in general, the life and social sciences are being harmed by lack of replication. There is a great need to incentivize replication and to de-incentivize "positive" null-hypothesis-rejecting findings. This requires changing the nature of publication processes to publish based upon the quality of methods and not upon whether the results are "interesting" or "ground-breaking", along with journals dedicated to replications unlike most current journals that reject failed replications both upon grounds of lackof novelty and bogus notions that null results are uninterpretable. We also need promotion and tenure processes that give less weight to acquiring external grants which too often are contingent of achieving or having achieved in the past "profitable" findings (broadly defined).
Some of this has started to change due partly to the unlimited publishing space of online journals.

 

[Underseer]

There are huge problems with the implementation of the scientific method in medical research. This is a topic of much discussion among medical researchers. Hopefully, they'll be able to address at least some of the problems.

Worse, the people who really need to read and understand the research (doctors) are not researchers themselves, so they'll always make mistakes in interpreting the research.

What's surprising about all of this is that as flawed as the implementation of science is in medical research, look at what a profound impact it's had over the last couple of centuries. Even bad science is better than what we had before.

 

[Speakpigeon]

There are huge problems with the implementation of the scientific method in medical research. This is a topic of much discussion among medical researchers. Hopefully, they'll be able to address at least some of the problems.

Worse, the people who really need to read and understand the research (doctors) are not researchers themselves, so they'll always make mistakes in interpreting the research.

What's surprising about all of this is that as flawed as the implementation of science is in medical research, look at what a profound impact it's had over the last couple of centuries. Even bad science is better than what we had before.

Progress is due first to hygiene, social programmes and good habits rather than medicine per se and then probably to specific campaign like vaccination where the science is probably better with a broader base in terms of the population tested. Medicine may work in niche markets like cancers but too many people definitely take too many drugs.

More progress would be achieved today in developed countries if we could somehow improve diet and the situation of substance abuse, including alcohol. Civilisation has come to a point where people seem to be bent on killing themselves through bad habits and medecine pretends it can patch up the consequences on our health with costly prescription drugs, i.e. more drugs. The medical profession is barely healthy enough to cure the diseased.
EB

 

[ronburgundy]

Worse, the people who really need to read and understand the research (doctors) are not researchers themselves, so they'll always make mistakes in interpreting the research.

This is a critical point. First there is the problems with the research itself (much of the problems being deliberately fraud fueled by profit motive), then even when the research is sound, it is not implemented soundly by doctors, most of whom are largely illiterate when it comes to understanding science. 90% of med school is memorization of facts and routinizing established procedures. Very little of it is about learning how to reason from evidence or interpret scientific findings.
On top of that, doctors choices are manipulated by drug and tech companies who give them kick backs and do shit like hire hot young females to hock their wares, meaning that what your doctors pushes on you has as much to do with whether he got a hard-on for the sales girl than whether the science shows its the best option.

 

[barbos]

Yes this is scary but it is an old news too.
Soft sciences like biology are filled with statistically incompetent or even outright fraudulent "studies".
People need publications on their resumes and they literally hunt for flukes to publish and they get away with it.
Hard sciences like physics are different because there is not much variety in topics of research and every positive result will be tried to replicate. In biology every god damn molecule can be tried on these poor lab rats and subsequently published, which will not be even tried to replicated because everyone knows that replication will be a failure and waste of time. Somewhat similar situation in hot fields like material science, they too publish a lot of crap which never gets replicated.

There are many drugs that must produce predictable results or they won't be prescribed.

If a doctor prescribes a blood pressure medication they are going to measure the effects of that drug for a long period of time.

And all cancer drugs undergo constant testing in trials. Many cancer patients enter trials where drugs are compared and this is happening all over the world.

If a drug is not producing significant results compared to placebo then this is discovered and those drugs are not recommended in guidelines that are constantly being updated.

It is possible a cancer drug could falsely gain approval, but once approved it's effectiveness will be tested.

I was talking about academia, not drug companies. As for them (drug companies) there were quite a few cases where dangerous/useless drugs were knowingly (by a drug company) approved.

 

[barbos]

Yes this is scary but it is an old news too.
Soft sciences like biology are filled with statistically incompetent or even outright fraudulent "studies".
People need publications on their resumes and they literally hunt for flukes to publish and they get away with it.
Hard sciences like physics are different because there is not much variety in topics of research and every positive result will be tried to replicate. In biology every god damn molecule can be tried on these poor lab rats and subsequently published, which will not be even tried to replicated because everyone knows that replication will be a failure and waste of time. Somewhat similar situation in hot fields like material science, they too publish a lot of crap which never gets replicated.

The level of fraud in medicine has much more to do with incompatibility between private profit motive and honest science than it does the "softness" of the science.
Studies have shown that medicine and pharmacology (where profit motive is high) are the most fraud-ridden areas of science, whereas plenty of other areas of biology that do more basic (and not directly profitable) research do not suffer from such fraud levels.

You are correct, that in general, the life and social sciences are being harmed by lack of replication. There is a great need to incentivize replication and to de-incentivize "positive" null-hypothesis-rejecting findings. This requires changing the nature of publication processes to publish based upon the quality of methods and not upon whether the results are "interesting" or "ground-breaking", along with journals dedicated to replications unlike most current journals that reject failed replications both upon grounds of lackof novelty and bogus notions that null results are uninterpretable. We also need promotion and tenure processes that give less weight to acquiring external grants which too often are contingent of achieving or having achieved in the past "profitable" findings (broadly defined).
Some of this has started to change due partly to the unlimited publishing space of online journals.

Good luck with changing the system

 

[barbos]

Speaking of bad relatively hard science.
Today's example of it:
http://www.sciencemag.org/content/early/2015/01/21/science.aaa3035
http://www.bbc.com/news/uk-scotland-glasgow-west-30944584
These people are either clueless idiots or very skillful frauds. I suspect the latter, their wording suggests that. But hey, they got publicity and publication.
Even though their "experiment" is nothing but a dishonest yet skillful misinterpretation.

 

[T.G.G. Moogly]

Speaking of bad relatively hard science.
Today's example of it:
http://www.sciencemag.org/content/early/2015/01/21/science.aaa3035
http://www.bbc.com/news/uk-scotland-glasgow-west-30944584
These people are either clueless idiots or very skillful frauds. I suspect the latter, their wording suggests that. But hey, they got publicity and publication.
Even though their "experiment" is nothing but a dishonest yet skillful misinterpretation.

They "observed" straight line distance speed. But the photons are traveling on their own straight lines. That would be my response.

 

[Perspicuo]

Speaking of bad relatively hard science.
Today's example of it:
http://www.sciencemag.org/content/early/2015/01/21/science.aaa3035
http://www.bbc.com/news/uk-scotland-glasgow-west-30944584
These people are either clueless idiots or very skillful frauds. I suspect the latter, their wording suggests that. But hey, they got publicity and publication.
Even though their "experiment" is nothing but a dishonest yet skillful misinterpretation.

We're talking about science not journalism.

It appears to me your example is off topic.

 

[barbos]

Speaking of bad relatively hard science.
Today's example of it:
http://www.sciencemag.org/content/early/2015/01/21/science.aaa3035
http://www.bbc.com/news/uk-scotland-glasgow-west-30944584
These people are either clueless idiots or very skillful frauds. I suspect the latter, their wording suggests that. But hey, they got publicity and publication.
Even though their "experiment" is nothing but a dishonest yet skillful misinterpretation.

We're talking about science not journalism.

It appears to me your example is off topic.

Don't blame journalists here, these "scientists" were pretty clear in their claims.
Yes. it's a bit off topic, but the problem we are talking about is more general and affect all sciences, not just medicine.

 

[barbos]

Speaking of bad relatively hard science.
Today's example of it:
http://www.sciencemag.org/content/early/2015/01/21/science.aaa3035
http://www.bbc.com/news/uk-scotland-glasgow-west-30944584
These people are either clueless idiots or very skillful frauds. I suspect the latter, their wording suggests that. But hey, they got publicity and publication.
Even though their "experiment" is nothing but a dishonest yet skillful misinterpretation.

They "observed" straight line distance speed. But the photons are traveling on their own straight lines. That would be my response.

It was not a straight line, but yes, two photons had very different paths.

 

[ronburgundy]

We're talking about science not journalism.

It appears to me your example is off topic.

Don't blame journalists here, these "scientists" were pretty clear in their claims.
Yes. it's a bit off topic, but the problem we are talking about is more general and affect all sciences, not just medicine.

Overreaching claims is one thing, outright fraud is another. Both occur in all sciences to a degree, but the degree matters alot. Studies tracking fraud across disciplines shows that medical research (especially pharma) has by far the most fraud. Fraud needs motive, and direct person profit is a larger motive for it than any other professional academic gains. Thus, there is more fraud in applied than basic science, particularly when its being applied to create claims that will sell a commercial product, which is what 95% of "medical" research amounts to.

 

[untermensche]

There are many drugs that must produce predictable results or they won't be prescribed.

If a doctor prescribes a blood pressure medication they are going to measure the effects of that drug for a long period of time.

And all cancer drugs undergo constant testing in trials. Many cancer patients enter trials where drugs are compared and this is happening all over the world.

If a drug is not producing significant results compared to placebo then this is discovered and those drugs are not recommended in guidelines that are constantly being updated.

It is possible a cancer drug could falsely gain approval, but once approved it's effectiveness will be tested.

I was talking about academia, not drug companies. As for them (drug companies) there were quite a few cases where dangerous/useless drugs were knowingly (by a drug company) approved.

I know what you were talking about.

But it doesn't carry over to drug research.

Ultimately drug research never ends. Even after a drug is approved it is constantly undergoing review.

That is how drugs like Vioxx were found to be dangerous and taken off the US market.

And this is much more the case for cancer drugs (the topic of the OP). They are almost exclusively prescribed based on guidelines.

If a cancer drug is not effective in post approval studies it doesn't matter what the manufacturer says, the drug won't be used.

 

[Underseer]

Here's a video version of the same thing in the OP:

http://www.ted.com/talks/ben_goldacre_what_doctors_don_t_know_about_the_drugs_they_prescribe