Magnetic bracelets and British Medical Journal

I just read in my local Danish newspaper Politiken that the British Medical Journal has published a positive test of magnetic bracelets against arthritis. The conclusion is:

Pain from osteoarthritis of the hip and knee decreases when wearing magnetic bracelets. It is uncertain whether this response is due to specific or non-specific (placebo) effects.

Click to expand...

Obviously, they have the usual problems of performing blind tests of magnetic bracelets when the patients can easily ascertain if their bracelet if magnetic or not, but understandably the peddlers of magical bracelets are elated! To be fair, one manufacturer asks if it is not time to conduct animal tests in order to find out once-and-for-all if it is placebo or another effect that is at work.

My local newspaper used the opportunity to tell us that there are other alternative treatments that also have documented effects that science cannot explain, e.g. homoeopathy! And then we are told that magnetic sheets and blankets can be used to shield against earth rays!

I wonder why they did not praise the effect of magnetic bracelets against aliens. At least, I have never heard of someone wearing a magnetic bracelet who was abducted by aliens!

steenkh said:

Obviously, they have the usual problems of performing blind tests of magnetic bracelets when the patients can easily ascertain if their bracelet if magnetic or not...

Click to expand...

Not only do they have the opportunity, they took advantage of it:

...it appears that 54% of the standard (strong) magnet group believed they had the real magnet, while 47% of the dummy group correctly guessed (or worked out) that they had the dummy magnet.

Overall, a highly significant number of the patients correctly determined which group they were in (54% of the standard magnet group thought they had the magnet, while only 38% of the weak group and 16% of the dummy group thought they had a real magnet). Furthermore, the results show that the main reason why they thought so was because they detected the magnetic force (32% of the standard group, 20% of the weak group and 6% of the dummy group).
Source

Click to expand...

It's not double-blind. It's not even single-blind! People knew if they had a magnetic bracelet or not!

steenkh said:

I just read in my local Danish newspaper Politiken that the British Medical Journal has published a positive test of magnetic bracelets against arthritis. The conclusion is:


Obviously, they have the usual problems of performing blind tests of magnetic bracelets when the patients can easily ascertain if their bracelet if magnetic or not, but understandably the peddlers of magical bracelets are elated! To be fair, one manufacturer asks if it is not time to conduct animal tests in order to find out once-and-for-all if it is placebo or another effect that is at work.
......
I wonder why they did not praise the effect of magnetic bracelets against aliens. At least, I have never heard of someone wearing a magnetic bracelet who was abducted by aliens!

Click to expand...

I'm wondering what's going on at the BMJ. I seem to be seeing more loony studies quoted in out local papers. Maybe just my failure to notice when they publish something good or my local paper's commitment to science?

steenkh said:

To be fair, one manufacturer asks if it is not time to conduct animal tests in order to find out once-and-for-all if it is placebo or another effect that is at work.

Click to expand...

Unfortunately, observer bias can be an even bigger problem in animal studies, because the observer has to interpret how the animal has responded rather than experiencing it directly. Such a study would only be valid if nobody who was involved in evaluating the animals' responses had any way of knowing which group they were in. Indeed, I'd be very suspicious of any subjective evaluation and I think that only objective measurements would give valid data.

Also, the law relating to animal experimentation is extremely strict, and it is very probable that any proposal to leave animals suffering from arthritis untreated, either as part of a deliberate placebo group or by "treating" with a possibly ineffective method would receive short shrift. A study I read regarding a herbal remedy for arthritis in the dog was rendered all but useless by the insistance of the licensing authorities that the owners of all the animals in the trial had to give them a recognised pain-killer if they thought they were in pain.

Rolfe.

First, with the majority of research results contradicting this study, there is no reason to think this piddly little study should outweigh other research.

I have to wonder why they even bothered? It is a small sample size, only subjective outcome measured, and poorly blinded. The abstract was suspiciously brief considering many would only read the abstract as the paper isn't available without subscription.

Second, there is an interestingly large number of magnet distributors with responding comments.

Third, no mechanism is even speculated let alone suggested as to why wearing a bracelet would affect one's hip pain.

I ain't buying it.

Re: Re: Magnetic bracelets and British Medical Journal

Rolfe said:

Also, the law relating to animal experimentation is extremely strict, and it is very probable that any proposal to leave animals suffering from arthritis untreated, either as part of a deliberate placebo group or by "treating" with a possibly ineffective method would receive short shrift. A study I read regarding a herbal remedy for arthritis in the dog was rendered all but useless by the insistance of the licensing authorities that the owners of all the animals in the trial had to give them a recognised pain-killer if they thought they were in pain.

Rolfe.

Click to expand...

You could do it with rats. There is a model for localized arthritis which involves injection of complete Freund's adjuvant (CFA). The outcome is change in the animal's reactivity to noxious stimuli. Specifically, prior to injection, the area is exposed to heat, mechanical and/or pressure stimulation. The intensity of the stimulation required before the animals moves away (usually it is the paw) from the stimulation is the baseline. Then inject the CFA. Repeat the pain testing. Usually, the animal will withdraw at much lower levels of stimulation (we interpret this as hyperalgesia). Then, you manipulate your independent variable. Rats with magnetic versus inert paw/leg cuffs (not sure how close it has to be to the painful area) and repeat the pain testing. Change from the hyperalgesic values is the measure of analgesia.

I don't want to get into a whole ethical debate, but to date, it is still ok to have placebo groups on this test.

Re: Re: Re: Magnetic bracelets and British Medical Journal

TruthSeeker said:

I don't want to get into a whole ethical debate, but to date, it is still ok to have placebo groups on this test.

Click to expand...

Yes, if you got it licensed as an experiment, with experimental animals, rather than a clinical trial with patients, it would be possible.

I have some doubt whether the ethical committees that look at the potential benefit to be gained from painful animal experiments would be terribly impressed with this idea though. I have even more doubt that any of the magnet-peddlers will ever actually put together a proposal.

Rolfe.

Re: Re: Re: Re: Magnetic bracelets and British Medical Journal

Rolfe said:

I have some doubt whether the ethical committees that look at the potential benefit to be gained from painful animal experiments would be terribly impressed with this idea though.
Rolfe.

Click to expand...

Yes, this would be the biggest hurdle. It is difficult to argue it's worth unless you did it from a skeptical point of view. That is, these data may be used to provide people who are suffering with clinical evidence regarding the efficacy of these treatments which are widely available but not evidence-based. Not sure.

We don't need animal studies for this research. Just better designed studies.


You can tell your subjects you are evaluating different metals. Tell one group you are testing the copper. Tell a second group you are testing magnets and so on.

You can tell one group you expect results, another you expect no result and say nothing to a third.

You can put the magnet around a wrist, ankle or directly over the hips. Is there a difference?

You can include objective measurements like inflammation, or range of motion.


There are many ways to get around subjects guessing which test group they are in.

Is there a source that has ONLY properly conducted double blinded studies in it?

skeptigirl said:

We don't need animal studies for this research. Just better designed studies.


You can tell your subjects you are evaluating different metals. Tell one group you are testing the copper. Tell a second group you are testing magnets and so on.

You can tell one group you expect results, another you expect no result and say nothing to a third.

You can put the magnet around a wrist, ankle or directly over the hips. Is there a difference?

You can include objective measurements like inflammation, or range of motion.


There are many ways to get around subjects guessing which test group they are in.

Click to expand...

This would actually be a difficult study to get approved. You can't mislead patients and the informed consent form has to tell them their chances of being assigned to any condition (and what the conditions are) and what the expected benefits and side effects of the treatments are.

Eos of the Eons said:

Is there a source that has ONLY properly conducted double blinded studies in it?

Click to expand...

Nope because even if you set out to do this you make a misstake from time to time.

TruthSeeker said:

This would actually be a difficult study to get approved. You can't mislead patients and the informed consent form has to tell them their chances of being assigned to any condition (and what the conditions are) and what the expected benefits and side effects of the treatments are.

Click to expand...

Who is giving this approval? I doubt there are any laws regulating human trials of magnet therapies.

You wouldn't be misleading anyone about the basics of the study. From your interpretation you'd have to tell everyone whether or not they have the placebo.

Ethics is in the eye of the beholder. From this site:
http://www.msu.edu/~biomed/rcr/humsubjects.htm
I couldn't see any laws except the FDA has specific regulations or "guidance documents" for very specific studies like gene therapy. In those cases you would have laws regulating giving persons drugs and invasive therapies that did cover human trials.

geni said:

Nope because even if you set out to do this you make a misstake from time to time.

Click to expand...

That assumes a lot.

A mistake from time to time doesn't mean every time.

From Quackwatch:

http://www.ncbi.nlm.nih.gov/entrez/...ve&db=PubMed&list_uids=10714732&dopt=Abstract No difference when the treatments were applied in the clinic only so the subjects had no chance to go home and "peek".

http://www.japmaonline.org/cgi/content/abstract/87/1/11 No difference when the insole was compared to insole with magnet.

The only studies I've seen with positive results have subjective questionnaires and no description of the blinded procedure in the abstract or a poorly blinded procedure. If someone wants to hunt down a blinded study with clear elimination of placebo effect I'd be happy to read it/them. I haven't seen any.

skeptigirl said:

Who is giving this approval? I doubt there are any laws regulating human trials of magnet therapies.

You wouldn't be misleading anyone about the basics of the study. From your interpretation you'd have to tell everyone whether or not they have the placebo.

Ethics is in the eye of the beholder. From this site:
http://www.msu.edu/~biomed/rcr/humsubjects.htm
I couldn't see any laws except the FDA has specific regulations or "guidance documents" for very specific studies like gene therapy. In those cases you would have laws regulating giving persons drugs and invasive therapies that did cover human trials.

Click to expand...

Well, I can only speak for research in Canada. The approval first comes from the funding source. Of course, if you are funded by the magnet manufacturers, you can get away with more, but then your results are less credible. I wouldn't say research ethics is in the eye of the beholder anymore, thank goodness. My research is approved by the university, hospital and funding council according to TriCouncil Standards (these aren't laws). This article although a bit out of date, reviews the most pertinent guidelines which you will see have a long history with wide acceptance.

You don't have to tell people the condition they are assigned to just what each condition is and the probability of their being assigned to it.

I don't have any of my consent forms on this computer, but from memory (with adjustment for our discussion) it reads something like (I'll provide a direct quote tomorrow, if you are interested)

"This study compares treatment A and B. Treatment A is a magnetic bracelet which may relieve your pain. In previous studies, it has been associated with improved pain. As well, XX% of people report this side effect and XX% report that one. Treatment B is also a bracelet made from different metals. It has been associated with XXXX. We are interested in comparing the effects of A and B on your arthritis. You have a 50% probability of receiving A or B. This will be determined randomly. The research staff will not know whether you have received A or B. Should it be necessary to provide the best medical care, information regarding your assignment to A or B is available from Dr. CCCC.

TruthSeeker said:

Well, I can only speak for research in Canada. ....

Click to expand...

No one is saying to ignore ethics in human trials I merely wanted to point out your claim about how a study could or couldn't be blinded was a bit more dogmatic than reality would have it.

It would be stupid if a study couldn't be properly blinded due to a 'rule' when the risk of harm to the subject was nil. I think you are taking informed consent out of context here.

skeptigirl said:

No one is saying to ignore ethics in human trials I merely wanted to point out your claim about how a study could or couldn't be blinded was a bit more dogmatic than reality would have it.

It would be stupid if a study couldn't be properly blinded due to a 'rule' when the risk of harm to the subject was nil. I think you are taking informed consent out of context here.

Click to expand...

How so? I don't think I understand how I am taking it out of context. We are discussing the information that must be made available to a person with a medical condition (osteoarthritis) that is considering enrolling in a clinical trial which includes a potentially active versus potentially inert treatment. Please clarify if that is not what we are discussing.

My reality is that I run clinical trials. I was just trying to convey the reality I deal with. It is difficult to argue that risk to subjects is nil, especially in a treatment study (one really can't say with 100% confidence that our manipulation would be without risk) and I can't imagine an REB allowing such a statement in a consent form. Even in questionnaire-only studies we include a sentence saying that answering certain questions may be upsetting (for instance on a depression questionnaire) and that patients are free not to answer them for that or any other reason.

TruthSeeker said:

How so? I don't think I understand how I am taking it out of context. We are discussing the information that must be made available to a person with a medical condition (osteoarthritis) that is considering enrolling in a clinical trial which includes a potentially active versus potentially inert treatment. Please clarify if that is not what we are discussing.

My reality is that I run clinical trials. I was just trying to convey the reality I deal with. It is difficult to argue that risk to subjects is nil, especially in a treatment study (one really can't say with 100% confidence that our manipulation would be without risk) and I can't imagine an REB allowing such a statement in a consent form. Even in questionnaire-only studies we include a sentence saying that answering certain questions may be upsetting (for instance on a depression questionnaire) and that patients are free not to answer them for that or any other reason.

Click to expand...

How do you blind any study if you can't disguise what you are testing for? And how do you study the effect suggestion has on outcome is you can't make suggestions?

Go back to my post and tell me which of the possible study designs you find issue with ethically? You said my suggestions wouldn't meet disclosure standards. That's silly. Maybe you could elaborate a little and we can clear up where we might be misunderstanding each other.

I will, but not tonight. It's been a rough one (see my thread in community if you want details)

skeptigirl said:

How do you blind any study if you can't disguise what you are testing for? And how do you study the effect suggestion has on outcome is you can't make suggestions?

Go back to my post and tell me which of the possible study designs you find issue with ethically? You said my suggestions wouldn't meet disclosure standards. That's silly. Maybe you could elaborate a little and we can clear up where we might be misunderstanding each other.

Click to expand...

Thanks for your patience regarding my response.

You can still blind a study and tell patients they are in a blinded study. They don't have to know to which group they've been randomized but they do have to know that the various groups exists. Also, you have to tell them what effects you are expecting. Because you tell everyone this, you are not introducing a systematic bias.

If you wanted to study "suggestion" than you wouldn't have to manipulate the type of bracelet, just the instructions, which of course involves lying to subjects. People of course do study these things but there is a great deal of debate around whether it is ethical to deceive people. Most of the work on pain and placebo and suggestion (like Benedetti's or Price's involves healthy volunteers to minimize the lie and most include at least some chance of getting a real analgesic).

As it is, it is difficult to get approval for a true placebo study. Usually, we compare new treatments to standard care with the promise (if at all possible) that they will receive the active treatment if it proves to be effective. Even then, we have "unblinded" data safety monitoring boards who regularly analyze the data to make sure we aren't missing any adverse outcomes or continuing with placebos after an effect has been shown. The idea of giving an ill patient a placebo and lying to them about it, well...that would be difficult (that is really all I'm saying)

If you wanted to compare the different types of bracelets than you would have a more difficult time justifying deceiving patients. That is, you can't tell one group you are testing one thing and another group that you are testing something else. You have to be pretty straightforward with patients about most aspects of the study. Lots of researchers hate this because it makes things much more complicated.

Perhaps the guidelines for research conduct aren't like this where you work, but it's like that here. Different research cultures maybe. Hope this helps.

TruthSeeker said:

Thanks for your patience regarding my response.

You can still blind a study and tell patients they are in a blinded study. ..............................
Perhaps the guidelines for research conduct aren't like this where you work, but it's like that here. Different research cultures maybe. Hope this helps.

Click to expand...

I don't see that any of the suggested studies I mentioned above couldn't be done within the ethical guidelines you mention here. Again, I ask you to address my suggested study designs specifically. I think you may have read them in haste.

I read your designs both when you posted them and last night when I responded. Perhaps it would be more helpful if you clarified the design with attention to addressing the ethical concerns I've raised. When my REB has a problem with a study, it is on the investigator to demonstrate there is no reason for concern

skeptigirl said:

You can tell your subjects you are evaluating different metals. Tell one group you are testing the copper. Tell a second group you are testing magnets and so on.

You can tell one group you expect results, another you expect no result and say nothing to a third.

You can put the magnet around a wrist, ankle or directly over the hips. Is there a difference?

You can include objective measurements like inflammation, or range of motion.


There are many ways to get around subjects guessing which test group they are in.

Click to expand...

Truthseeker
This would actually be a difficult study to get approved. You can't mislead patients and the informed consent form has to tell them their chances of being assigned to any condition (and what the conditions are) and what the expected benefits and side effects of the treatments are.

Click to expand...

I think we can eliminate the last two as included in your ethics objections.

So starting with using different metals in your bracelets, you include copper, magnetic, and a dummy bracelet. You tell people you are testing the bracelets for their effect on hip pain or whatever you are studying and give them the usual disclosures about the study except: You tell one group no more than that. You tell the second randomized group you are testing the copper bracelet and a third group you are testing the magnetic bracelet. You don't tell any of the groups which bracelet they are getting. Where is your ethics issue?

In the second scenario, you have all the usual consents or whatever, and you set your study up as the original researchers did. But to a third of the group you tell them you expect to see a difference, a third you say you expect to disprove the difference and to the other third you say nothing.
Where is your ethics issue?

I am a bit annoyed that you couldn't substantiate your claim my suggestions could not be incorporated into an ethical study. When I asked you for specifics you gave more generalities. When I asked you to apply your generalities to my specific statements you refused a third time.

I stand by my discussion of possible ways a magnet could be tested and at least some if not all of the placebo effect could be sorted out despite subjects being able to easily discover which group they had been randomized into. Your claim such methods could not pass human subject ethics review is nonsense.

I really don't want to fight with you about this. I tried my best to answer your questions. I'm sorry if you are not satisfied. You haven't really addressed my concerns about witholding information from patients other than to dismiss them as "nonsense". These concerns are not nonsense.

The ironic truth is that as I participate in this thread I am preparing for an inspection of my clinical trials by Health Canada. I am not concerned because we rigidly adhere to their guidelines and I am pretty good at clinical trial ethics (your evaluation notwithstanding). Nonetheless, the preparation is still anxiety provoking situation. Anyway...

skeptigirl said:

I think we can eliminate the last two as included in your ethics objections.

So starting with using different metals in your bracelets, you include copper, magnetic, and a dummy bracelet. You tell people you are testing the bracelets for their effect on hip pain or whatever you are studying and give them the usual disclosures about the study except: You tell one group no more than that. You tell the second randomized group you are testing the copper bracelet and a third group you are testing the magnetic bracelet. You don't tell any of the groups which bracelet they are getting. Where is your ethics issue?


In order to run this study in Canada, you would have to tell the subjects that there are three groups and what each one consists of and their chances of being assigned to each group. If your design includes this (I can't tell if each group is aware of the existence of the others and the probabilities) then I have no problem.

In the second scenario, you have all the usual consents or whatever, and you set your study up as the original researchers did. But to a third of the group you tell them you expect to see a difference, a third you say you expect to disprove the difference and to the other third you say nothing.
Where is your ethics issue?


This one is a bit trickier. It has to do with manipulation of expectation and withholding information about expected outcomes. I think you could do it if the consent form said for 1/3 "previous studies have shown benefits from wearing these bracelets including decreased pain. We are going to compare the analgesia/range of motion (whatever outcome) to assess the size of this difference. " versus for the second 1/3 "the results of previous studies have been inconclusive. We do not expect whatever outcome to differ between the group. (Aside: but then why would anyone consent????)," versus the trickiest of the three. You cannot withold information about anticipated outcomes when you are studying patients and potentially active treatments. Our guidelines require that you tell patients what we already know about a treatment and what outcome (positive and negative) you are expecting. I'm just not sure you could get away with saying nothing about expected outcome. Perhaps others working in a different research culture could comment.




I am a bit annoyed that you couldn't substantiate your claim my suggestions could not be incorporated into an ethical study. When I asked you for specifics you gave more generalities. When I asked you to apply your generalities to my specific statements you refused a third time.

I stand by my discussion of possible ways a magnet could be tested and at least some if not all of the placebo effect could be sorted out despite subjects being able to easily discover which group they had been randomized into. Your claim such methods could not pass human subject ethics review is nonsense.

Click to expand...

Again, I am sorry you think I am spouting "nonsense". I know the guidelines for my research culture. In fact, I have a meeting at 10:00 am which I really must prepare for....

TruthSeeker said:

I really don't want to fight with you about this. I tried my best to answer your questions. I'm sorry if you are not satisfied. You haven't really addressed my concerns about withholding information from patients other than to dismiss them as "nonsense". These concerns are not nonsense.

Click to expand...

What I wasn't satisfied with was your cop out when I asked for specifics. If you didn't have time, that's all that need be said but twice you said I should initiate the discourse about why you were objecting. As the design suggestions included few details your objections suggested the few details I gave were inherently unethical. That is where the nonsense lies as is clear when you finally address things below.

In order to run this study [referring to the three different bracelets] in Canada, you would have to tell the subjects that there are three groups and what each one consists of and their chances of being assigned to each group. If your design includes this (I can't tell if each group is aware of the existence of the others and the probabilities) then I have no problem.

Click to expand...

And so where in my description do you find me saying anything contrary to your additional study design details? As you see, once you think about what I said, you find such a study is indeed ethically possible.

This one [suggesting outcomes to see how the suggestion affects the results] is a bit trickier. It has to do with manipulation of expectation and withholding information about expected outcomes. I think you could do it if the consent form said for 1/3 "previous studies have shown benefits from wearing these bracelets including decreased pain. We are going to compare the analgesia/range of motion (whatever outcome) to assess the size of this difference. " versus for the second 1/3 "the results of previous studies have been inconclusive. We do not expect whatever outcome to differ between the group. (Aside: but then why would anyone consent????)," versus the trickiest of the three. You cannot withhold information about anticipated outcomes when you are studying patients and potentially active treatments. Our guidelines require that you tell patients what we already know about a treatment and what outcome (positive and negative) you are expecting. I'm just not sure you could get away with saying nothing about expected outcome. Perhaps others working in a different research culture could comment.

Click to expand...

Are you claiming you tell your research subjects about the full range of all the research outcomes from every study to date?

Here you are narrowly defining 'expected outcome'. An expected outcome can be stated as very specific or very general and still be correct and ethical. We don't know but want to find out if there is a difference, we want to see if there is an effect, we want to show there is an effect, we want to show there is no effect, we want to see if there is or is not an effect, and so on are many ways to say the same thing yet can influence subjective reports of pain outcomes. Your suggestions about how the study might be designed to control for suggested outcomes influencing outcomes are essentially no different than mine. Yours would also work. Mine were not unethical.

Again, I am sorry you think I am spouting "nonsense". I know the guidelines for my research culture. In fact, I have a meeting at 10:00 am which I really must prepare for....

Click to expand...

I have no doubt you know your guidelines. But you voiced objections to my suggestions of possible study designs which I had described in the briefest possible terms. They weren't inherently unethical, you were just unable to imagine how one could incorporate such elements into an ethical design.