Slashdot Mirror

Unfortunately the 21st Century Cures Act will encourage the FDA to approve treatments based not on randomized, controlled trials, but on weaker evidence, such as observational studies, case studies, anecdotes, and testimony by patient groups financed by the drug companies.

Here's an article that rounds up some of the other articles about it. http://www.healthnewsreview.or...

Sometimes you read a medical case history and the doctor says, "There are no randomized, controlled trials to demonstrate effectiveness, so it is treated empirically." Because of the 21st Century Cures Act, you'll now see this more and more. We won't know what the effective treatments are.

I realize that the Daily Mail has issues.
https://www.theguardian.com/co... After all, we get Private Eye here in the states.

The big problem with this story is that they're hyping a vaccine that is still in Phase I clinical trials. Yeah, doctors are trying to find a vaccine for the common cold. Doctors have been doing that for 100 years. What's new about this one?

Other than that, it's a somewhat disorganized collection of interesting and maybe even useful information about the common cold. She went to experts and they explained their work and what they thought were the important issues. She spent a day at Imperial College London, let them teach her about the common cold, and wrote a story about it. She could do worse.

If you want to understand how to evaluate (or write) a medical story, Health News Reviews has a great checklist. http://www.healthnewsreview.or...

You also mentioned something about a media source claiming syringes require "extensive medical training" or something... I call BS. Again, diabetics deal with this all the time. There are some precautions, but most are similar to EpiPens, and the additional warnings can easily be explained in a few minutes. You also may want to check into the credentials of that medical professional -- I've seen some media quotes in stories in the past few days saying similar, but it turns out they work for allergy societies that get a huge amount of support from the manufacturer of EpiPens, which at a minimum presents a significant conflict of interest.

That "media source" was Consumer Reports. I have checked into their credentials. Their medical reviewers are probably more qualified in each of the specialties than some of the reviewers in the second-string peer-review journals. And they take no money from industry.

Diabetics do inject themselves with insulin, however there are differences between them and people with anaphylactic reactions so you can't equate the two. The most obvious difference is that insulin-dependent diabetics inject regularly, several times a day, so they're used to the equipment and familiar with it. People with anaphylactic reactions might a reaction once in their lives, once a year, or once every few years (according to a friend of mine who did have an anaphylactic reaction to bee antigen in a doctor's office), so they can forget how to use it.

You want to say that it makes no difference. I don't accept that. In a matter of life or death, you need better evidence than your own personal feeling. You seem to know enough about medicine to be able to look up articles on PubMed, but I'm certain that you're not a medical doctor or medical student. The standard of evidence for pharmaecuticals is a lot more rigorous than, say, the flavors and fragrances industry. I'd rather follow the advice of an MD.

It's not good enough to say that diabetic injections are sort of like epinephrine injections, so if it works for diabetes it seems like it should work for epinephrine. The only thing that will tell you what kind of problems come up when people use manual epinephrine injections is a well-designed study of people who use manual epinephrine injections, preferably with a comparison group of people who use the EpiPen. But that would be hard to do, because an anaphylactic shock is such a rare event.

And contrary to what you say, there is nothing in those studies that addresses the claim that "people won't fill them correctly or they'll lose time in doing all that for people inexperienced with them." Those were just lab studies of 2 narrow issues -- stability and sterility.

If you want to understand the design of medical studies, you could read the NEJM, BMJ, and JAMA Internal Medicine (my preferences) over the last few years. If you want to get a summary of what it's all about, you can look in http://www.healthnewsreview.or...

http://www.consumerreports.org...
Can You Get a Cheaper EpiPen?
You could save about $400 per two-pack with generic Adrenaclick and still protect against life-threatening allergy attacks
By Ginger Skinner
August 11, 2016

The DIY Syringe Method

To further cut costs, some have turned to using manual syringes and buying vials of epinephrine to fill them. The drug costs a few dollars per vial. But experts caution that switching to a do-it-yourself syringe is more complicated and can result in getting too much or little epinephrine. What’s more, you’ll need to be trained by a doctor or pharmacist on how to inject the drug quickly and accurately before attempting to try it during an emergency.

And because there are different concentrations on the market, getting the proper dose is critical, especially for children.

You beat me to it.

Citation needed.

http://www.healthnewsreview.or...

"Frequent fish consumption was associated with a 50% reduction in the relative risk of dying from a heart attack." Her editor's reaction? Slash. Too wordy, too passive. The editor's rewrite? "Women who ate fish five times a week cut their risk of dying later from a heart attack by half."

  https://www.elsevier.com/conne...

Science journalists should be better than "average schlubs". They owe it to their readers to do at least a few minutes of fact checking before publishing.

Gary Schwitzer, who runs the Health News Review blog, gave a checklist that makes it easier.

http://www.healthnewsreview.or...

Tips for Understanding Studies

        Mixed messages about statistical significance
        Surrogate markers may not tell the whole story
        Be careful with composites
        Does The Language Fit The Evidence? - Association Versus Causation
        Odds ratios
        Resources for Reporting on Costs of Medical Interventions
        "Off-label" Drug Use and Marketing
        7 Words (and more) You Shouldn't Use in Medical News
        News from Scientific Meetings
        Absolute vs. Relative Risk
        Number Needed to Treat
        Commercialism
        Single Source Stories
        FDA Approval Not Guaranteed
        Phases of Drug Trials
        Medical Devices
        Animal & Lab Studies

Well, had you read the article .../quote>

Had you read the article? The article to read was not in a newspaper, but the original publication http://papers.ssrn.com/sol3/pa... , in the Social Science Research Network.

1. Association is not causation.

2. It hasn't been published in a peer-reviewed journal. It was presented at a conference, and uploaded to an archive. There were no reviewers to point out any obvious flaws that would be obvious to a specialist in the field that a newspaper reporter might not notice (assuming he had indeed read the entire study).

3. They are economists, not scientists. Scientists in psychology, psychiatry, and other related fields have been doing studies like this for generations. Economists sometimes make big mistakes in science. They make mistakes in conceptualizing problems and figuring out how to apply their data to the real world. They don't know how to correct for confounding factors, like, were the students who had access to cannabis similar to the students who didn't? Maybe foreign students were better students than native students.

Most significantly, scientific (particularly medical) studies have a higher standard of evidence than economic studies. They make their authors prove more details, link by link, in the chain of argument. For example, this study assumes that the native students actually did smoke more cannabis than the foreign students. How do they know that? Did they do blood tests? (No.) They surveyed students on their cannabis consumption, discussed starting on p. 22, but they didn't use the survey to answer the obvious question, which is, did the native students actually smoke more cannabis than the foreign students? What were the percentages? Were they statistically significant? From my first read of the paper, I don't see that they got that data. Maybe the law had no effect. If somebody can find it in there, let me know.

In a medical drug study, researchers might assign one group of patients a drug for treating, say, AIDS. Sometimes the drug has uncomfortable side effects and the patients don't take it (without telling the doctor). They'll count pills to see whether patients are taking the drug. If it's important enough, they'll take blood tests.

4. As a science journalist, I can say that, in my professional opinion, the author of this news story is incompetent. It doesn't follow the generally-accepted best practices for medical journalism http://www.healthnewsreview.or... or any professional journalism, unless you want to use the definition that anybody who calls himself a journalist is one.

The howling mistake that the author, Jamie Doward, made is that he didn't get comment from a knowledgeable source with a different view http://www.healthnewsreview.or... If I was his editor, I would tell him to go back and get a comment. If he didn't routinely get a second comment for a controversial story like this, I would fire him. If I were teaching him in a journalism course, I would give him an F, until he gets the lesson (if ever).

There have been many studies comparing marijuana smokers and non-smokers, and they've found small effects on cognitive performance (sometimes in both directions), but never anything as dramatic as this. This requires an explanation from some of the researchers who have been studying this question for a long time.

There's a reason for this. Science journalists don't have to be smart. They just have to know their limitations. If I write a story about something as controversial as whether cannabis lowers your college grades, and interview one guy, I may not know enough to understand it. So I call up somebody else, who knows more than me, and may disagree with the first source

Well, had you read the article .../quote>

Had you read the article? The article to read was not in a newspaper, but the original publication http://papers.ssrn.com/sol3/pa... , in the Social Science Research Network.

1. Association is not causation.

2. It hasn't been published in a peer-reviewed journal. It was presented at a conference, and uploaded to an archive. There were no reviewers to point out any obvious flaws that would be obvious to a specialist in the field that a newspaper reporter might not notice (assuming he had indeed read the entire study).

3. They are economists, not scientists. Scientists in psychology, psychiatry, and other related fields have been doing studies like this for generations. Economists sometimes make big mistakes in science. They make mistakes in conceptualizing problems and figuring out how to apply their data to the real world. They don't know how to correct for confounding factors, like, were the students who had access to cannabis similar to the students who didn't? Maybe foreign students were better students than native students.

Most significantly, scientific (particularly medical) studies have a higher standard of evidence than economic studies. They make their authors prove more details, link by link, in the chain of argument. For example, this study assumes that the native students actually did smoke more cannabis than the foreign students. How do they know that? Did they do blood tests? (No.) They surveyed students on their cannabis consumption, discussed starting on p. 22, but they didn't use the survey to answer the obvious question, which is, did the native students actually smoke more cannabis than the foreign students? What were the percentages? Were they statistically significant? From my first read of the paper, I don't see that they got that data. Maybe the law had no effect. If somebody can find it in there, let me know.

In a medical drug study, researchers might assign one group of patients a drug for treating, say, AIDS. Sometimes the drug has uncomfortable side effects and the patients don't take it (without telling the doctor). They'll count pills to see whether patients are taking the drug. If it's important enough, they'll take blood tests.

4. As a science journalist, I can say that, in my professional opinion, the author of this news story is incompetent. It doesn't follow the generally-accepted best practices for medical journalism http://www.healthnewsreview.or... or any professional journalism, unless you want to use the definition that anybody who calls himself a journalist is one.

The howling mistake that the author, Jamie Doward, made is that he didn't get comment from a knowledgeable source with a different view http://www.healthnewsreview.or... If I was his editor, I would tell him to go back and get a comment. If he didn't routinely get a second comment for a controversial story like this, I would fire him. If I were teaching him in a journalism course, I would give him an F, until he gets the lesson (if ever).

There have been many studies comparing marijuana smokers and non-smokers, and they've found small effects on cognitive performance (sometimes in both directions), but never anything as dramatic as this. This requires an explanation from some of the researchers who have been studying this question for a long time.

There's a reason for this. Science journalists don't have to be smart. They just have to know their limitations. If I write a story about something as controversial as whether cannabis lowers your college grades, and interview one guy, I may not know enough to understand it. So I call up somebody else, who knows more than me, and may disagree with the first source

Those 5 questions should be asked of pretty much every scientific study done, no matter what the field

That's the way to write a good science story. Whatever the story is about, you explain the basic questions that the reader should ask.

I despair about ever having people know the difference between association and causation. Nevertheless, it's worth repeating.

Gary Schwitzer's web site http://www.healthnewsreview.or... has more detail http://www.healthnewsreview.or...

Those 5 questions should be asked of pretty much every scientific study done, no matter what the field

That's the way to write a good science story. Whatever the story is about, you explain the basic questions that the reader should ask.

I despair about ever having people know the difference between association and causation. Nevertheless, it's worth repeating.

Gary Schwitzer's web site http://www.healthnewsreview.or... has more detail http://www.healthnewsreview.or...

Even "may help" still suggests a causal link, where none has been demonstrated. Such language can confuse a lot of people and can be avoided but it takes some effort that space-saving and lazy journalists and headline writers may not wish to exert.
See http://www.healthnewsreview.org/2012/03/please-read-our-primer-on-observational-studies/