N. Thornton wrote:
Hi Kevin.
I think this should clear up some things.
All one need do is collect the data:
1. realise this is an effective way to learn successful methods
2. attract the people
3. assess the claimed results
4. Apply statistics to discover what is actually working.
Then we will know what works.
But you can't apply meaningful statistics to back of the envelope
studies.
I'm not suggesting a back of envelope study. The data already exists,
and only needs collecting and assessing. The individuals have mostly
been assessed sufficiently within the medical system to know they have
found a real result. Applying statistics to such collected data does
indeed reveal a clear pattern. Lets give you an example.
method A 3 positive results out of 112 attempts
method B 30,000 positive results out of 50,000 attempts
method C 71 positive results out of 400 attempts
method D 4 positive results out of 6 attempts
method E 60 positive results out of 900 attempts
Very simple stuff, it shows clearly which method should be researched
in more depth, and which one might warrant more data collection. This
elementary statistics can indeed be applied.
Correlations by themselves have little value. Correlations have to be
shown to be *causative*. Its why the tobacco industry could claim that
the was no proof that smoking causes cancer. I agree that you can use
this as a starting point for further investigations, but by themselves
have little value.
One needs controlled studies, double-blind tests etc...
etc....Anything else is, in reality pretty much useless.
This is a standard misperception of our time.
Not at all. This has evolved precisely because it was recognised that
the prior methods had seriously failings.
Collecting the info I
propose would indeed show who has recoverd from what and how, and it
will show up any substantial trends. Such data is not 100% reliable,
but 75% reliability is more than enough of a basis on which to do a
more thorough study on a method. I'll explain this more further down.
One has no real idea how reliable it is at all. There are too many
places where the results could be totally meaningless.
lots of
doctors and nurses have come across the odd one who achieved a
remarkable result with their condition.
Which could have been luck, i.e. nothing to do with any purported
cure,
just fixed itself on its own. Happens all the time.
Of course. But when you get the data and find tens of thousands of
results from a particular method, way above the luck level with all
other approaches, that is statistically significant data, not just
luck. That is what we're looking for.
I agree, if you see strong correlations, then this should warrant
further study.
Yet the NHS is failing to
collate and assess such data, and study known successes to learn more
techniques it can apply.
One offs are not success.
So if you have a condition and you find a way to successfully treat
it, and have used the treatment lets say 10 times in your life, and it
has resulted in rapid favourable outcomes each time, you're saying
that's not a success? It is by definition a success.
Thats not what I'm saying. I meant one off by not working on all
individuals, only a few individuals.
For example, my twin brother suffered with severe eczema for around 25
years, from birth. Continually scratching. It was horrendous. His skin
was a mess. Nothing helped. His life was utter misery from it. One day
he decided to go completely vegan, that's is no animal products
whatsoever. After around 3 months, he noticed that he was no longer
scratching. He was cured, and has been ok now for the last 20 years. On
occasions he has inadvertently taken animal products, in all cases he
now has a severe allergic reaction confirming that food was the problem.
So, this method worked for him, will it for others?
This can
*only* come about by controlled trials. This is really a no-brainier.
That's what you've been taught, and I understand why. In reality you
dont need controlled trials to get 75% confidence.
This makes no sense. Without controlled experiments you have no idea
what you've got. Its 101 science.
You can get that by
pooling multiple results of uncontrolled trials. You'll never get
100%, but with some things you can get enough confidence to warrant
doing a more in depth investigation.
The data could all be Swiss cheese. The literature is filled with
erroneous "studies".
Now, tell me, was the treatment oral magnesium suphate discovered by
double blind controlled trials? I think not.
Ho hum...
Yet it works, and was
discovered to work.
You can pick any one off the cuff example you like, but it means nothing
to the principle of doing good science.
Obviously it is not only controlled trials that
can produce useful data, you only have to understand statistics to see
that, rather than believe what you're taught blindly.
You doing that condescending one-upmanship again. The scientific method
is not something one learns "blindly". This is exactly what the method
is designed to eliminate. Its a process to ensure that what you think
you have, is what you have. When peoples welfare is at stake, its even
more relevant then in the inanimate sciences.
There are very good reasons why dbc trials are generally insisted on,
but to discount _all_ others is an (understandable) mistake... I could
get into all that in much more depth, but its a whole thread on its
own.
I don't discount non controlled studies out of hand. If they indicate
significant promise, than that is an argument to do the studies
correctly.
Instead the NHS refuses to learn what it
doesn't know.
Confirmation of claimed cures involve large amounts of money.
The system already has that data. Its on peoples medical records.
Those people have already been assessed by their docs/specialists and
found to be better. It costs little.
Doubt it somehow. Things usually mushroom out of hand. There is simply
too many claimed cures to pursue any. Waiting lists are way over a year
for many treatments in the UK NHS.
What does need changing is a specific form is required for this, to
make this data clear and mass-harvestable.
Oh dear...another form...I don't think that you have really thought this
through.
At the moment sometimes the
relevant info is written on the records, sometimes its not, and almost
never is it reported to anyone who could make use of it.
The system don't have it.
Wrong again.
1. It takes no fortune to run a small data collection point and issue
the forms.
I don't agree. Its extra staff. Every expenditure requires good support
for it.
2. It is actually a very cost effective way to do research to find new
treatments, much more so than paying for drugs that cost billions to
develop. The system does have the money to pay for that, it just needs
to allocate a little of that money more wisely.
But you cant determine what will actually pay out in the end. You
assuming that you choose correctly in what your studying. Most won't pan
out. Its the way life works. There are 1000's of different medical
conditions that all need cured, selecting one, e.g. such as depression,
is a choice that not many will take.
When the survey results achieve a successful treatment, it will start
to save the system money. It is simply better value than billion pound
drugs.
Arguable, depression is a non issue for the NHS. They can simply send
depressed people away. Maybe it costs society money in the long run, but
that will have little impact on NHS policy.
Its like a river bed, among all the dirt there is gold
and diamonds, and those valuable things are simply not being made
any
use of.
I think you bit out on this one.
I know, but I'll wait until I see you've comprehended the idea before
taking on board such judgements. I hope from this reply you will be a
bit clearer on the concept I propose. Only when you have got what I'm
actually proposing will you be in a position to make a genuinely
informed judgement. So far that has not been possible - think about
that. Soon or now it will be.
You proposal is really a dreaming proposal. Nice idea, but will never
work. It don't take into account the way real life works. Every one is
competing for limited resources, with there own pet agendas. I don't
think you understand just what it takes to get an organisation to go and
spend money, especially one like the NHS.
The risks of doing something wrong simply does not allow for this
approach.
That doesnt stop todays researchers. Research projects get it wrong
all the time, its expected, its OK.
I did not suggest this. Its the people that implement/recommend the cure
that has the liability. If something goes wrong, people get sued. Its
that simple.
One only has two say the words "law" "sue" to put most people in the
picture.
No, you've just missed it. Cite me a case of a researcher being sued
for honestly producing data that turns out to not pan out long term.
Its an erroneous criticism.
No, you missed it. Why would you think I was referring to the data
collectors?
With what I'm proposing it is clear upfront to all that some of the
leads this generates will pan out and some wont. There wont be any
surprises or broken promises when something's found not to work later,
we know that will happen.
Unfortunately, only a few % will pan out. its the way its always been.
I'm thinking you'll be rather clearer by now on just what I'm
proposing.
Indeed. Wishful dreaming...
Kevin Aylward
http://www.anasoft.co.uk
SuperSpice, a very affordable Mixed-Mode
Windows Simulator with Schematic Capture,
Waveform Display, FFT's and Filter Design.