View Full Version : How to "spin" the result of a scientific study


Kunga Dorji
07-05-14, 12:11 AM
See:
http://online.liebertpub.com/doi/full/10.1089/cap.2014.0020

Cardiovascular Safety of Stimulants in Children with Attention-Deficit/Hyperactivity Disorder: A Nationwide Prospective Cohort Study

However- as reported in Medscape:

ADHD Meds May Double Cardiovascular Event Risk in Kids

http://www.medscape.com/viewarticle/827747?nlid=60723_2842&src=wnl_edit_dail&uac=127657DR

(anyone can subscribe)

So the hard data:
Rate of all cardiovascular events in the population of Danish Children
84 per 100,000 patient years.
Rate of all cardiovascular events in Danish children with ADHD:
170 per 100,000

Increased risk = 86 events per 100,000 patient years.

So- there is no information as to whether any of these events were life threatening or irreversible, or whether they were even considered significant enough to modify the treatment given.

There is nothing to counter evidence that had failed to show any life threatening events in the stimulant treated group.

There is nothing to compare these very low risks with the much higher risks associated even with penicillin (which is usually prescribed unnecessarily anyhow).

There is no attempt to quantify the number of adverse events per day prevented by being on stimulant medication when it is needed.

This is a wonderful example of "spinning" a minor result to make a headline.

Neurotypicals-- you can't life with them, you cant live without them ---well somebody has to file stuff for us :)

SB_UK
07-05-14, 01:47 AM
Heck Namazu came out as Banazu typing in the dark.
Anyway Namazu described the problem wonderfully.

Highly statistically significant with no clinical significance.

It's the core problem with the application of math in medicine.

Imagine we had a coin which was weighted so that 1 time in a million it would preferentially land on Heads instead of Tails because it's very, very, very slightly weighted - well this could be highly statistically significant given a billion coin tosses ... ... but in reality - the ("clinical") 'bias' is neither here nor there.

We don't want to identify 100 genes which make a tiny contribution to disease in inherited disorders - we want to stop people getting them in the first place. Identifying 100 genes with variants which predispose to disease isn't really of clinical interest - maybe pure academic/scientific interest - but the patient who's suffering might struggle with the idea that massive amounts of disease research money is being funneled into keeping scientists in highly paid jobs - and not into making their lives better.
He has taken a lead a role in the identification of almost 100 risk loci for inflammatory bowel diseases (IBD). I don't know ... ... a neverending story !!

How to end the madness.

To realise that there's a commonality between all disease.

This centres on psych. distress from living in an unequal society and inappropriate and excess food intake alongside lack of exercise in the sun - of course alongside all of the other (downstream) money-related issues ie pollution, family breakdown, addiction ... ... so an entirely environmental basis to ALL common disease.

SB_UK
07-05-14, 02:05 AM
---well somebody has to file stuff for us :)

I don't think much of what we file will ever be returned to.

It's a precedent-centric instead of a future-centric mindset which desires a return to the 'good' old days instead of something considerably better.

Saw this in one research project where tech advanced rendering previous data woefully inadequate.

Only info which I think is worth saving 'd be computer games, music, art, video ... ... but even in this arena - as tech advances occur we will need to regenerate higher resolution / higher fidelity copies for dissemination - as our eyes and ears - senses become acclimatised to 'better'.

Play an exquisitely programmed game now like GTA5 on PC and it's painful to go back a generation - this basic idea is what I've seen in medical research - such an effort to store and access previously generated data - though where tech means that it's better to regenerate the dataset from scratch.

In actual fact though - there's no real point to the accumulation of 'omics data - we need a change in society to overcome disease. There is though a point to the retention of information which entertains (builds an internal sense of the appreciation of quality) - because we become better as a consequence.

Question - lower quality data instrumental in 'training' developing neural networks.
The simple nursery rhyme at the appropriate developmental time-frame ... ... but give it a few years and the nursery rhyme will cause pain to the more educated ear.

A form of education - not stipulated (standard education) - but 'liking' the finer things in life - because of the pleasure in sensory information upload.

Kunga Dorji
07-05-14, 02:28 AM
I don't think much of what we file will ever be returned to.

It's a precedent-centric instead of a future-centric mindset which desires a return to the 'good' old days instead of something considerably better.

Saw this in one research project where tech advanced rendering previous data woefully inadequate.

Only info which I think is worth saving 'd be computer games, music, art, video ... ... but even in this arena - as tech advances occur we will need to regenerate higher resolution / higher fidelity copies for dissemination - as our eyes and ears - senses become acclimatised to 'better'.

Play an exquisitely programmed game now like GTA5 on PC and it's painful to go back a generation - this basic idea is what I've seen in medical research - such an effort to store and access previously generated data - though where tech means that it's better to regenerate the dataset from scratch.

In actual fact though - there's no real point to the accumulation of 'omics data - we need a change in society to overcome disease. There is though a point to the retention of information which entertains (builds an internal sense of the appreciation of quality) - because we become better as a consequence.

Question - lower quality data instrumental in 'training' developing neural networks.
The simple nursery rhyme at the appropriate developmental time-frame ... ... but give it a few years and the nursery rhyme will cause pain to the more educated ear.

A form of education - not stipulated (standard education) - but 'liking' the finer things in life - because of the pleasure in sensory information upload.

Yes-- we cling to the past too much.
Still- the act of filing keeps the neurotypicals gainfully employed.

Abi
07-05-14, 02:35 AM
Thread temporarily closed for staff review.

Abi
07-05-14, 02:43 AM
Moderator Note

Please stay on topic.

Thank you.

(Thread re-opened)

InvitroCanibal
07-05-14, 12:12 PM
Great post! I find this happens more in the public press where studies become misrepresented or misinterpreted. Another good example would be the 1999 study on medication effectiveness vs community involvement and awareness and cognitive behavioral therapies.

The general data showed a 60 percent success rate on meds alone but with cbt and community awareness the rate was 80 percent.

Later on after the study was done they did a follow up study to track the results of the impact treatment had made in the now adult childrens lives. This is where things become mixed up, because the parents were now free to do what they wanted. They never checked to see if parents had in fact continued medicating their children or not, but the conclusion was that the long term efficacy was statistically invalid.

So many times articles came out using this as a "drugs don't work" rally against adhd and the like. It's very dangerous how we interpret studies and bad information is deadly.

Kunga Dorji
07-05-14, 06:49 PM
Great post! I find this happens more in the public press where studies become misrepresented or misinterpreted. Another good example would be the 1999 study on medication effectiveness vs community involvement and awareness and cognitive behavioral therapies.

The general data showed a 60 percent success rate on meds alone but with cbt and community awareness the rate was 80 percent.

Later on after the study was done they did a follow up study to track the results of the impact treatment had made in the now adult childrens lives. This is where things become mixed up, because the parents were now free to do what they wanted. They never checked to see if parents had in fact continued medicating their children or not, but the conclusion was that the long term efficacy was statistically invalid.

So many times articles came out using this as a "drugs don't work" rally against adhd and the like. It's very dangerous how we interpret studies and bad information is deadly.

The anti ADHD camp is very persistent at misrepresenting any minor finding that can be used to bolster its cause.

However, another issue in the interpretation of any study is what is meant by "effectiveness"- some studies show far higher success rates than others and this is lost in vague references to "statistical significance".

Kunga Dorji
07-05-14, 07:10 PM
This chapter Bioelectromagnetic Medicine Chapter 44 Unabridged Version
Cranial Electrotherapy Stimulation for Anxiety, Depression, Insomnia, Cognitive Dysfunction, and Pain: A Review and Meta-Analyses
Daniel L. Kirsch,
Ph.D., D.A.A.P.M., F.A.I.S. and Ray B. Smith, Ph.D.

of this book:
Bioelectromagnetic Medicine
Edited by: Paul J. Rosch, M.D. The American Institute of Stress, Yonkers, New York, U.S.A., and New York Medical College, Valhalla, New York, U.S.A.
and Marko S. Markov, Ph.D. Research International, Buffalo, New York, U.S.A.
Published: April, 2004 by Marcel Dekker, Inc., New York ISBN: 0-8247-4700-3; 850 pages;

provides an excellent analysis of some of the pitfalls in metanalysis- and some of the different things meant by "effectiveness"-- a very tricky word.
It used to be freely available to all visitors to the alpha-stim website but is now password protected for health care providers only.

I have a copy and am happy to forward it if approached by PM.

Carol
07-06-14, 05:14 AM
Upon hearing this spin, my innocence was lost:

Back in the 90's, litigation was brought against Big Tobacco on behalf of 46 states to recoup Medicaid and public health care coverage of medical expenses caused by smoking. Big Tobacco collected and analyzed data, proving undeniably that the lifetime medical expenses of smokers were lower than the lifetime medical expenses of non-smokers.

It's really true. You see, smokers had shorter lives

Kunga Dorji
07-07-14, 09:26 PM
Upon hearing this spin, my innocence was lost:

Back in the 90's, litigation was brought against Big Tobacco on behalf of 46 states to recoup Medicaid and public health care coverage of medical expenses caused by smoking. Big Tobacco collected and analyzed data, proving undeniably that the lifetime medical expenses of smokers were lower than the lifetime medical expenses of non-smokers.

It's really true. You see, smokers had shorter lives

What a classic.
I knew they were corrupt, but I had clearly underestimated them.