Differences in rules of evidence -- a way out ?

An important way in which people here differ is in what they consider to be valid rules of evidence. For example, some will believe and act on the assertions of those posters or integrative medicine professionals whom they have come to trust, while others insist on support from peer-reviewed, double-blind studies in which all the relevant confounding variables are either randomized away or controlled.

In either case, the "truth" uncovered by either strategy is difficult to defend to someone who thinks differently about what constitutes valid evidence. In some cases there may be good reasons for this, as when we learn that in some cases of journal articles, results of failed studies are not reported, or in the "trusted authority approach" conflict of interest in uncovered.

In any case, it is up to each of us to do an "experiment of one" to determine whether the finding, however it is arrived at, applies to us. This approach is often referred to here, and offers a way to respond to conclusions arrived at by what for you are questionable means without trying to change the person's mind about how to arrive at valid conclusions.

... which leads me to my last offering, which is to refer you to an article on applying principles of single-case research to determining whether a supplement is working for you :[www.af-ideas.com]

and you can google "single case research for a lot more.

-- DickI (previously "Dick" -- my old name won't go through)

In my opinion, all too often "The Placebo Effect" is overly used to apply a negative connotation to a treatment or therapy which does not have the support of or is a threat to mainstream medicine, but which may very well have something to offer. Most who use this term know nothing about the thing they are criticizing or poo pooing. The term can also be used in a way that is an insult to the intelligence of a patient and is an attempt to set the doctor in a position of intellectual superiority. After all, how can you make a convincing argument against the placebo effect? That's what's so great about it.
As a person who has worked around and with researchers a long time and who is married to a PhD scientist, it's important to know that many researchers fudge their data and results to make things come out nice and neat and in the way they think it should. Many studies are badly carried out and/or designed. Statistics can be used in ways that are very misleading. Often the person that gets his work published knows someone and he or she does not necessarily have the most compelling paper to publish. It's kind of a old boy buddy system. So, all the so called "Peer Reviewed" research should be looked at with a very critical eye.

Tish

Very well said, Tish! Good for you! -- and good for us. Thank you. Erling

Dick,

I agree using single-case methodology is very appropriate for use in analyzing protocols for an individual.

George

"The term can also be used in a way that is an insult to the intelligence of a patient and is an attempt to set the doctor in a position of intellectual superiority. After all, how can you make a convincing argument against the placebo effect? That's what's so great about it."

Thank you, Tish. I have had all too much contact with medical persons this last calendar year, but was not able to articulate what ails me about their damnably smug attitudes. You have put words to the feeling these persons give me.

PeggyM

Thank you Tish! Those biased study results are coming home to roost almost daily now in the news on drugs up for closer scrutiny after initially receiving FDA approval. Many doctors are saying...if you have to use a drug, to be safe, use one that's been used for several years so that most of the adverse effects are known, and avoid the latest newcomers. Certainly some drugs can be lifesavers but there are many other options that can also be very useful, less expensive and have no adverse side effects. What we need to do is be aware and your post helps create awareness. Jackie

Tish,

Thanks for the inside look at some of the things that go on behind the scenes in the research jungle. More is coming out about researchers doing a bunch of studies until they come up with one that is positive -- ar at least marketable--, while tossing those with negative results: (http://archinte.ama-assn.org/cgi/content/full/169/11/1022).

One answer is to have a registry where the results of all studies in a particular area are inexpensively recorded and thereby reducing the incentives to get positive results at the expense of validity (?). Someone would have to enforce this, however ...

Re the placebo effect:

(Believe it or not), I have long been a fan of the placebo effect.

For example, it would seem to me to be counter-productive for a doctor to devalue the placebo effect of something the patient is doing which seems to be helping, since this effect can add to the curative effect of treatment. That it has a "real" neurological effect has been demonstrated many times [www.physorg.com].

I would think that important question would be how to make optimal use of placebo power (characterized by such terms as hope, expectations of improvement or of a particular treatment effect, enhanced by such things as belief in the expertise of the physician or the influence of advertising ) -- by figuring out how to determine which people are placebo reactors and how to enhance this effect when appropriate.

In other words, this effect is simply another class of independent variable, whose value or lack of it is determined the same way as other variables (?)

(The effects of a treatment whose power is mostly placebo, however, *may* lose it's effectiveness sooner than one whose effects are non-placebo. But this may depend on the individual and other factors.)
.
So, if a treatment is working for you, I certainly wouldn't discard it solely on the basis of research showing that it is no better than a placebo, especially if the placebo effect is substantial, as it often -- surprisingly -- is.

We are now back to the "experiment of one" outcome research, as opposed to research into the process and research whose goal to set public health policy for large numbers of people.

These are very large topics so I will leave it at that.

-- Dick

The Single Case model is useful even for those using traditional meds that have been validated with double blind, placebo controlled studies. Even here, how do you know that you are in the category where the treatment is effective. A Single Case model could help.

The three models that seem to have applicability are:

Classic Reversal
A-B-A

Simplest Single-Case Design
A-B

Alternating-Treatments Design
A-B-C-A-C-A-B-A-C

It is great in theory, but if you find something that appears to work, you are reluctant to go back to baseline (i.e. A-B-A) to "make sure" it is the protocol that is working.

I'm a case in point. In my first four months of afib, my afib load was 57% (i.e. I was in afib 57% of the time). This includes 2.5 months of persistent afib. In the subsequent 5.75 years, my afib load has been 0.057%, and 2/3is of that was in one 20 hour session, the first one after starting my supplement program.

I did do an A-B-A. I'd been on the supplement program for about 2.5 years and in NSR for 2 years. I thought I might be sensitive to some of the supplement fillers, so I stopped all supplements. In a day or two, I was back in afib (for an hour - converted with PIP flecainide, the other arm of the protocol). Anyway, I was understandably reluctant to continue alternating A-B-A and subsequently continued with "B" (i.e. supplements).

I've also not been rigorous enough to statistically analyze my results. The differences were large enough "by inspection" that I didn't feel this necessary (57/0.57 =1,000).

George