About ITP
Childhood ITP
Teenage ITP
Adult ITP
Medical Advisors
Research into ITP
Events
Convention
Seminar
Send for:-
Publications
American Perspective
Downloads
Leaflets and more....
Webwatch
April 2007Some people believe in the oddest things. Those who have read previous Webwatch articles "Does drug X cure ITP?" from 2004 and "Lies, dammed lies and statistics", from 2005 will have detected that I have an interest in trying to avoid this tendency so it will not come as surprise that this Webwatch follows similar ideas.
Whilst looking for a good site about the placebo effect, I came across an interesting site entitled "hidden persuaders" at skepdic.com/hiddenpersuaders.html. It presents some short but fascinating descriptions of all sorts of ideas directly relevant to understanding the problems faced by patients, doctors and researchers.
One such problem is the "regressive fallacy", "the failure to take into account natural and inevitable fluctuations of things when ascribing causes to them". As has been heard during talks at the annual conventions and seminars, some patients seem to find it hard to understand why their platelet counts fluctuate. The answer is that they will fluctuate just as everything else in nature fluctuates. The regressive fallacy is to believe that a rise in platelet count after an event was due to that event and is the stamping ground of the charlatan, "take this ancient remedy and your count will improve". If you do take it and your count improves, they take it as proof. If it fails, they get you to try another remedy. Eventually natural variation provides a rise in the count and bingo, proof the remedy works! When asked why the first failed, they will simply tell you they were fine tuning the treatment to your body.
Another tool of the charlatan is communal reinforcement, "the process by which a claim becomes a strong belief through repeated assertion by members of a community." The relevant point here is the lack of testing, a (false) claim is made many times and it becomes fact. Dr Joseph Goebbels was an world class exponent of this as even a brief look at www.calvin.edu/academic/cas/gpa/goebmain.htm will show. His idea was to tell a big lie very often. After a while many did believe him. The same works for common untested ideas, say them often enough and people will believe. It took years to get rid of the idea that "night airs" caused disease. Similarly, as recently as the mid 19th Century, it was thought Cholera was due to miasma. False or not, even well educated people believed even though no evidence was ever provided to support the idea. As it says on cancerweb.ncl.ac.uk/cgi-bin/omd?miasma+theory, a miasma is "an explanation of the origin of epidemics, based on the false notion that they were caused by air of bad quality, e.g., emanating from rotting vegetation in marshes or swamps.".
The placebo effect is well known and can be very powerful but in the wrong hands, "cures" are claimed from all sorts of daft substances, often very expensive. The placebo effect is an important reason why double blind clinical trials are carried out. As it says on skepdic.com , "The placebo effect is the measurable, observable, or felt improvement in health not attributable to treatment.". The word placebo is Latin for "I will please" (skepdic.com and en.wikipedia.org/wiki/Placebo).
Selection bias is a significant danger in clinic trials. As it says on skepdic.com, this is either "self-selection of individuals to participate in an activity or survey, or as a subject in an experimental study" or worse, "the selection of samples or studies by researchers to support a particular hypothesis." . It is not only medical trials that suffer from this problem. Whenever you hear of a poll such as "57% of shoppers would vote Conservative at the next election" you can see the selection bias. Was it the shoppers out at a convenient time and location for the pollsters or did it include a representative sample of all shoppers? Probably not.
It is always pleasing when a clinical study finds a positive effect of a treatment or drug, but what happens if no effect is found? Does this make the study less important than others? Of course not, it simply shows that work needs to be directed elsewhere, proving something does not work increases knowledge. The section on skepdic.com is clear on this point but also highlights a danger, that of the research that contradicts previously held beliefs. This is where the file-drawer effect becomes apparent, an effect that "refers to the practice of researchers filing away studies with negative outcomes." It is avoided in clinical trials, and in most science, by ensuring that results are checked by other independent teams. Little is accepted into general use until results are confirmed several times. As I write this, a report was broadcast on BBC Radio Four saying that "20 good quality studies" had some interesting results about the effect of diet on breast cancer. The language used implied that 20 studies were just about sufficient for the ideas to become mainstream.
Although believing odd things can be very harmful, sometimes it can be
made entertaining. Derren Brown uses tricks of the mind to entertain.
Both on his website at www.channel4.com/entertainment/tv/microsites/M/mindcontrol
and in his book, Trick Of The Mind, he gives away much of how he works.
He also tries to point out the risks of believing without good cause.
A fascinating example from his book is relevant here, it goes like this:-
Imagine a horrid disease is around. (To avoid confusing this with ITP
or any other disease, I shall call it by a name used at playtime in my
primary school days, The Lurgy.) Luckily, The Lurgy is rare, it occurs
in only 1 in 10,000 of the population. It does not favour male or female,
young or the old. There is a test for The Lurgy that is 99% accurate,
that is to say that 1% of test results are wrong, some will say you have
The Lurgy when you do not and vice versa. For the sake of your peace of
mind, you ask your GP for the test. You have the test and after a while,
you are sitting in the surgery and are shocked to find the results are
positive, it says you have The Lurgy. Oh dear! Your mind focuses on the
idea the test is 99% accurate, so all is lost. Pain, suffering and death
are to follow in short order. But just a minute, what is the probability
you actually have The Lurgy?
You may be surprised to find the outlook is quite good. As Derren Brown
points out, you have forgotten that The Lurgy only strikes 1 in 10,000.
Consider a million people, of these, 100 will have the disease, 99 of
these unfortunate people will have a correct test result, 1 will be wrong.
Of the 1 million, 999,900 will not have the disease, but 9,999 of them
or 1%, will have a false test result. You may be in this group, in fact
there is only about a 1% chance you actually have The Lurgy. This runs
directly counter to intuition but is quite good news!
There are plenty of examples of similar counter intuitive tricks of the mind. One is the Monty Hall effect, well described on en.wikipedia.org/wiki/Monty_hall_problem or and animated version on math.ucsd.edu/~crypto/Monty/monty.html. You have 3 doors in a game show, one hides a prize. You choose one but before you open it, the host, who know where the prize is, opens a different door and you see no prize. You now have 2 doors left, one of which hides the prize. The host now asks if you would like to change your choice. Should you?
Happy surfing.
Howard