Alternative medicine: research and integrity

Edzard Ernst, MD, PhD


Edzard Ernst: I will move straight away into my talk. This is what I'm going to talk about. I'm going to give a little bit on background. I'm going to talk about research: research activity, research questions, some findings. Then, most importantly, in integrity with the subheadings, methodology, researchers, and published papers, and eventually come to some conclusions. I studied medicine in in Germany, my first post was in in the homeopathy Hospital, then I became a normal physician, conducted quite a few years of basic research, became a specialist, and became chair of rehabilitation medicine in Vienna, and eventually came to Exeter, where I was chair of complementary medicine. There, I built up a team of about 20 co-workers. We published plenty of papers. I got some rewards, and since about 10 years ago, I'm retired and am an Emeritus professor of Exeter University. For those who want to know more about me, I've published a memoir which nature wants to call a crying call for medical Ethics, which is very nice. About the terminology related to alternative medicine, In Exeter we called it complementary medicine. Before that people called it Finch Medicine. Some people called it holistic medicine, integrated medicine and natural medicine, and nowadays I call it so-called alternative medicine, and SCAM for short.

Why do I do this? Well if a therapy doesn't work, it cannot be an alternative to anything, and if a therapy does work, It belongs to evidence based medicine, therefore I think that SCAM is really the better term. I also think so because I once published a book called SCAM. How do we define it? Interestingly enough, there's not one generally accepted definition of alternative medicine. For the purpose of this talk I define it as an umbrella term of a diverse range of therapeutic, and we often forget that diagnostic modalities which have little in common other than being excluded from mainstream medicine. And it goes from A to Z. Somebody counted over 400 different modalities under this umbrella.

More background. Scam is popular. That's something we all know. That shows you some actual figures, and it shows you that actually in the UK it is not that popular. About 20% of the general population use it. If you look at patient populations, this can go up to nearly 100%. Interestingly Germany tries to dominates the world yet again, all be it the alternative one. The highest figures in Europe are achieved by Germans, who use it at a 40% rate (in general population) 

More background. Consumers pay for it out of it from their own pocket, at least in the UK. The press has an insatiable interest. Consumers are bombarded with misinformation. Doctors normally have very little interest, and many people religiously believe in it, and VIPs promote it such as Dr. Oz, who's just running for and to become a Congressman in America. We have Mr. Holt, who believes in homeopathy. Prince Charles, who believes in anything alternative. Olivia Newton John, who just died of cancer, which she partly treated with alternative medicine. Arnold Schwarzenegger who promotes chiropractic. Gwyneth Paltrow, who earns quite a bit of money selling quackery. Luc Montagnier, where it's not quite clear whether he actually promoted it, but supported homeopathy. Linus Pauling, who also invented molecular medicine.

Research activity. People think there is no research, that's quite wrong. I've recently looked up these modalities (Acupuncture, Chiropractic, Herbal medicine, Homeopathy, TCM) and found almost 40,000 publications on acupuncture, 9,000 on chiropractic, etc, etc.  What is particularly impressive is the amount of TCM - traditional Chinese medicine, and I'm coming back to that one because I found this really quite concerning. If you think these figures are high,  I just want to remind you that they are actually relatively small. These are the figures for the subject of pharmacology. 6 million in pharmacology, 9,000 in chiropractic. There you see quite a big difference. 

And of course these are not clinical trials. Most of these publications are data free even and Medline is not very good at identifying clinical trials but my estimates are that we only have about 2,000 clinical trials of acupuncture, 500 of chiropractic. 5,000 about herbal medicine and 500 of homeopathy. But these are my estimates, and they may not at all be accurate.

Research questions. A huge variety of research questions really stem from what actually is SCAM, Do SCAM users with cancer live longer than non-users with cancer. Particularly with sociologists, take great pleasure in researching all sorts of questions which I don't find particularly relevant. 

Next slide shows you the research question that I focused on in those 30 odd years I was involved in the subject. Does it work about it? And by it I mean the specific modality and does it work for certain conditions, of course. So that short question is already a few thousand or even more specific research questions. Is it safe and crucially do its benefits outweighs its risks? These were the research questions that occupied our life in Exeter.

Some of the findings. Surprisingly, some even though they are very popular, some SCAMs have not even a single clinical trial to back them up. Classical example is SchÜssler salts. Not very popular in English-speaking countries, but hugely popular in German-speaking countries, with a lot of money being made on it, and not a single trial as far as I know. Other examples of this category is  Colonic irrigation, Laetrile and Shiatsu. 

Then we have some SCAMs which have been documented to cause significant harm. Examples for that category are spinal manipulation, coffee enemas, chelation therapy, some herbal treatments, and acupuncture. These clearly put the patient at risk. And here I show you a publication from way back in 1996. A large survey on over 4,000 British SCAM users, and we found, these are the black percentage figures, that about 16% reported adverse effects after spinal manipulation, about 13% after acupuncture, 10% after homeopathy and about 8% after herbal remedies. The figures in red are the actual figures as we know them today, almost 25 years after that. Every second person experiences adverse effects after spinal manipulation. With acupuncture our rough estimates was quite correct, it's about 10%. With homeopathy you see 20%, and don't ask me why something that doesn't contain anything can cause adverse-effects, that's a lecture in itself, and was with herbal remedies it really depends on whichever remedy you are talking about? I think the most interesting bit here is chiropractic. As I said, 50% of patients experience adverse effects after spinal manipulation, and this is not even disputed by chiropractors. They say that's normal, because things have to get a little bit worse before they get better, and we're talking about increase of pain, referred pain, which can be quite acute and quite strong and can be strong enough to decrease quality of life. In addition, we have very severe complications, including death. 500 cases have been reported. I need to mention here that there's no reporting system, so we're looking at the tip of a much bigger iceberg and chiropractors hotly dispute that, because obviously that is not good for business.

Next slide shows you what usually or what often happens with these complications. That's the upper spine and the skull, and you see the vertical artery which with certain chiropractic movements is overstretched and then essentially can burst in which case you have a stroke, in which case you can also die 

These were the direct risks. I think the indirect risks are much more important, and much neglected. I've mentioned the question whether uses of SCAN who also have a cancer live longer than non-users? The answer is no in fact, they have a shorter life expectancy. This is an American publication, and they showed that 5 years survival is reduced by 50% if you use also some form of SCAM which is surprising because it's not the direct harm of SCAM they also speculate, but it's the indirect harm of not taking mainstream oncology seriously enough to get cured. 

And some scams are effective, and only these should be considered for integration into routine care. Next slide shows you the range of SCAMs that we are talking about. A few herbal remedies are well supported by clinical evidence. Some dietary supplements. Oil pulling, I'm not going to explain what this is. Alexander technique. Massage, Pilates, laughter therapy, mindfulness, and progressive muscle relaxation. This is from a book that's about to be published, where I review 202 different modalities. You must admit that this is not a long list after reviewing 202 different modalities, but some SCAMs are backed up by reasonably sound evidence. A not quite new meta-analysis I’m showing you here with 30 randomized clinical trials, showing that St. John's wort is superior to placebo and equally effective as conventional antidepressants or equally ineffectively, one might say. And in in my book, St. John's wort is actually better because the side effects, if you use it correctly, are much less than with conventional antidepressants 

This is the last category of SCAMs. They have been tested extensively, and have failed to demonstrate that they do more good than harm. Unsurprisingly that includes homeopathy, it includes chiropractic mainly because of the harm side of the equation that I've explained to you. Osteopathy same roughly, and reflexology, simply because it's not doing much harm but it's also not doing any good.

So now we're coming to the to the main topic, namely, the integrity part of my talk. How we might define integrity – “the use of honest and verifiable methods in proposing, performing and reporting research”. It involves honesty, trust, fairness, respect, responsibility, courage, and as you are critical, I hope, you might well ask the “What about the integrity of my good self”.  It has been called into question by this guy, King Charles, who wrote to my Vice Chancellor, or his first private secretary, alleging that I have breached confidence in in the documents that they were producing. This resulted in an official investigation at the end of which I was pronounced innocent, but my department during those 13 months that it lasted. My department was destroyed, and I went in to early retirement, so my integrity has been called into question, but later on I was clear of the allegation.

Back to integrity of research into SCAM and just a few methodological issues. If we want to know whether treatment works or not we obviously, at least in in my book, want to know whether it's better than placebo.  And there we have a sizable problem in so-called alternative medicine because for some treatments it's truly difficult to find a placebo. Placebo effects might actually be quite important, or might be more important in SCAM than in conventional medicine. If we think of mind body therapies, chiropractic manipulation, and acupuncture, we don't have an obvious placebo treatment which we can use in clinical trials in order to blind patients. The next slide shows you what we have done in Exeter. To solve one of these programs, namely the acupuncture one, we've developed a placebo needle, and you see the schematic drawing and a picture of the actual device and the device is glued to the skin and if you tap on it with the real acupuncture needle, it penetrates the skin, and with the placebo needle it just implodes in itself like a stage dagger, and it doesn't penetrate the skin. Interestingly the clinical trials that have been conducted with such needles tend to be negative. They tend to show no difference between placebo needle and real acupuncture needle.

The researcher is a tricky business in so-called alternative medicine. What do you think about a researcher who spends his life researching his pets therapy, but manages to go through life not publishing a single negative result? He must be very special. In my view, he's so special that on my blog, I created a whole of fame for these guys, The alternative Medicine Hall of Fame. And it's getting more and more crowded, and so it's a work in progress. And the names that you see in red are people from the UK. For me personally particularly remarkable is Adrian White. Adrian was a co-worker in my department until 2005, and under my guidance and direction, and I forgot to say that he was an enthusiastic acupuncturist and doctor so it wasn't quite easy to teach him objective research where you have to obviously leave behind your sympathy for any particular treatment, but we managed to publish together quite a few trials, and without exception they produced negative results. You can see that on the left hand side of the slides. Then we had a major fallout. He left my department, went to another department, where he was allowed to do research unsupervised, or mostly unsupervised, and he managed to publish in the following years nothing but positive results. I'm unaware of any better example to demonstrate bias, in just one single person, and the most powerful effect bias can have in producing results in SCAM research. And on an even bigger level, what do we think of a whole country that’s very active in clinical research and in SCAM research but hardly ever publishes negative results? I've told you how and I've showed you how active China is these days in producing results of TCM in particular, and of course I’m talking about China.

Here's something that we published quite a while ago in the BMJ, which disclosed that practically all RCTs of traditional Chinese medicine that we found, and we are talking about just under 3,000 RCTs in the Chinese literature conducted by Chinese people, produced positive results, and that was the first time I stumbled over this really amazing finding, and later on it was confirmed by several other groups. I think at least half a dozen other groups confirm the basic finding that when it comes to traditional Chinese medicine China very rarely, almost never, produces negative results. Which means either everything works, and we know how unlikely that is. Or there's something wrong with Chinese SCAM research. 

A few published papers to give you some actual examples of what I mean. Here we have an exceptionally poor paper, an observation study that's was aimed, and that's a quote too, “to examine the effectiveness of acupuncture”. Well observational studies, as you probably know, cannot establish that and on top of it was a tiny observation study, and the conclusion was acupuncture was found to be effective for the reduction and relief of symptoms. This would be a totally unimportant study if that's wasn't happening so often in SCAM research. The wrong design to answer the research question. A tiny study, and a totally overblown conclusion. “Thanks to Yoga, I now gently, stretch conclusions instead of jumping to them.” That seems to be the obvious remark to make here 

And here we have the hierarchy of evidence, and of course we were talking of very low down in this hierarchy as an observation study, and I think everybody knows that observation studies are not to be relied upon when we want to test for an effectiveness of a treatment. What about randomized clinical trials? I think we have the same phenomenon of unreliability in many of the clinical, even randomized, clinical trials of SCAM. This one is from Vienna. Professor Michael Frass wanted to evaluate whether homeopathy influenced the outcome of cancer patients, and he recruited 410 cancer patients. Randomized them into 2 parallel groups. One group received homeopathy, the other one standard care plus homeopathy, and the endpoint was global health status and subjective well-being. The results showed patients “improved significantly when homeopathic treatment is administered”. What is wrong? It's not as clearly visible as with the observation study that I showed you previously. What is wrong with that study design? Well, to put it really simply, 2 amounts of money are always more than one amount of money, and that's the trial design which I've been banging on for quite a while. The A+B vs B design. It has to produce a positive result, even if A is a placebo, because we can expect a placebo effect. But the trial design looks rigorous. It is randomized, and it is, etc. So it has some of the features of rigorous clinical trials, but it's really never, or hardly ever, fails to produce a positive result. I can say this with some confidence, because we looked at this systematically a while ago, and with acupuncture we found 13 trials that followed this design, 12 of which showed a significant positive result and one showing a positive trend, so that's what we would expect according to the 2 amounts of money versus one amount of money.

What about systematic reviews? They are supposed to be the Rolls Royce of evidence-based medicine, producing the most reliable evidence. Well, not necessarily so in SCAM. Here we have a systematic review, published in a very respected journal, and the review was acupuncture for chronic low back pain. A meta-analysis of 25 RCTs concluded that acupuncture may have favourable effects on self-reported pain and function limitation. That all looked very convincing, except if you know that the largest trial with 1162 patients wasn't included for some reason. Well, maybe because it was negative. And anyway, if you do include it, which we later did, the overall results turns out to be no longer positive. And that's not just confined to acupuncture. It happens with other subjects as well. Here's an example of homeopathy. Individualized homeopathy was submitted to a systematic review. 32 RCTs were included, and the conclusion was medicine prescribed in individualized homeopathy may have small specific treatment effects. That's a very surprising result of a meta-analysis because homeopathic remedies, individualized or not, usually don't contain anything, so how can they be better than placebo. To understand the result you need to know that 2 high quality RCTs were not included. The 2 were both negative, and if you include them in the meta-analysis, the results turn out to be negative.

So we have doubts about the integrity in SCAM research, but we should ask what effect this lack of integrity might have? I believe that the scientific community will take scam research less and less seriously. Thus even credible research will be ignored, and some valuable modalities which one might find under this umbrella of SCAM, might not get the recognition that it deserves.

I'm coming to my conclusions. I will keep this very simple just for conclusions. Alternative medicine might not be the best name; so-called alternative medicine, I think, is the much better name, and I tried to explain why this is so. I've shown you that research activity is relatively low, so some people might be impressed by the numbers of studies of acupuncture for instance, but relative to research activity in other areas, it is really minute. I’ve shown you that research findings are mixed. By no means, everything is negative, but much of it is negative. By no means everything is safe. For most modalities we cannot honestly say that the benefits outweigh their risks. Finally, I personally, after researching this area for about 30 years, have serious doubts about the integrity of the research and more delicately the researcher. And that's about it.


Klara Rombauts: So at the anti-cancer funds we often get questions from cancer patients who have heard about, now, it's very popular, cell therapies, and they are being sold at extreme prices, and it seems that there's no regulatory thing that can be done. And also as Professor Ernst pointed out in his presentation, It's a chiropractic, there's also that which has been there, but it seems that on a regulatory level nothing can be done. So how could Europe play a role in this, what could be done in order to stop these practices. 

Edzard Ernst: Regulation is really a difficult subject for me. To put it bluntly, even the best regulation of nonsense will result in nonsense. That's for me the main problem that we have with regulation, and alternative practitioners have a schizophrenic attitude towards regulation. On the one hand, they love to be regulated because it gives them some veneer of respectability. On the other hand, if you regulate alternative practitioners properly, by which I mean regulate them such that they have to abide by the rules of evidence-based medicine, you regulate them away, because then a homeopath cannot possibly practice homeopathy anymore, to use just one example. Having said all this, obviously one needs some form of regulation to protect consumers and patients from the worst excesses of quackery. I wish I knew how to do that. Quite honestly, I don’t.

Susan Bewley: In a way, it's the opposite question which is: to show willing, and integrity, and benefit, how come we can't get those remedies that have been shown to work, into practice in the NHS, partly because they're useful, and partly because they really show up all the rest that don't work in the sense that you know, here you are, complementary medicine. Here's a little small box of tricks that work, and all those that don't. Why cant we get that small list onto NICE. 

Edzard Ernst: I often wonder myself, but NICE is totally reluctant to even touch it (alternative medicine). I was unofficially told by somebody, very high up in NICE, that our present king prevents that. I don't know whether this is true. On a more cynical note, somebody said: “In medicine we make progress, funeral by funeral”. 

Leeza Osipenko: Oh, I guess my opinion would be just from being within NICE is that it cannot actually be in the scope of technology appraisals because of price. This is not an expensive medication which is comparable to current oncology therapies, where significant budget decisions need to be made. Another question, and part of guidelines, but here I think that it's much better for NICE to stay completely clear, because as soon as they touch one alternative medicine there will be inquiries. Why not the next? So It's out of remit I suppose.

Edzard Ernst: Yeah, if we talk about guidelines, it’s slightly different. I was talking of specific evaluation i.e. of NICE evaluating homeopathy for instance.

Leeza Osipenko: Yeah and the really huge thing is budget impact because NICE, actually has  not historically evaluated absolutely everything coming through. So, for example, Haemophilia, HIV drugs were never in the remit, all for different reasons, and there are many interventions that, just by scope, don't make it into technology appraisals, homeopathy being one of them. There are many reasons, many of them you've already mentioned.

David Colquhoun: I was a bit surprised about your slightly optimistic view of acupuncture. It seems to me the biases involved are so big, that there's really no good evidence that works, but NICE has evaluated it, and it's come up with completely different contradictory conclusions according to who was on the guidance group. I was on the radio who recommended it for non-specific, low back pain, or something like that. Then another guidance group came along, which had more pain specialists, and they said they do not recommend that acupuncture. And then, just recently, some other guidance group for some other sort of chronic pain, did recommend this. They are just hopelessly inconsistent. 

Edzard Ernst: I'm not aware that I was very positive about acupuncture but maybe it's my desire not to be outright negative that makes me positive about something, at least. And anyway, I entirely agree with you. The guidelines sometimes included and excluded acupuncture. It might well be connected to which experts they recruit. If they recruit anybody from hall of fame then the result is likely to be different from if they recruit a real scientist 

David Colquhoun: Yes, that was very much the case I think.

Francois Maignon: I wanted to capture your thoughts on something you didn’t pick up on in the talk and that was on under registration and under reporting of clinical trial results. I was involved in the review on the efficacy reporting trials In France. When we were writing the report we discovered that less than 10% of the clinical trials conducted in homeopathy were actually registered, or reported some results. And this is also not at all reflected in the systematic reviews. Do you have any ideas or thoughts on this this misconduct in the field of SCAMS. 

Edzard Ernst: Yes, well, I've shown you my whole of fame, of people who have only published positive results, and I've shown you the example of China, a whole country only publishing positive results in one specific area. How can this be? In in the case of China, I’ve often asked Chinese co-workers (over the years we had quite a few Chinese co-workers), why this might be, and the best explanation they could give me was that in China, it is quite unthinkable that a relatively junior researcher would do something that contradicts the opinion of his or her peers. it's not possible. So, consciously or unconsciously, studies are being designed to produce positive results, but that wasn’t quite your question. Negative results remain unpublished. That clearly is the case, and then methods to at least try to get a handle on this problem of publication bias. But also negative results are being tortured until they can turn out to be positive. There are many methods in statistics, David knows that much better than I do in how to achieve that. Lastly, there's the most delicate question of fraud. Fraudulent research happens, sadly, all time; not just in SCAM, but everywhere and perhaps in SCAM it is more prevalent, not publishing a study for which people have paid for, and patients have given their time, and perhaps even more towards, is in my view, fraud. The mechanism to avoid this or minimize is registering clinical trials. So people are working on minimizing all these adverse outcomes, but the world is not perfect, and we are clearly far from totally eradicating fraud in particular.

Leeza Osipenko: What do you think is the key contributor to the popularity of SCAM? Is this financial profits? Because it's very well financed by the companies who want to push their either homeopathic products or some other products to the market. Is this the failure of traditional mainstream medicine, interventional medicine? Or is this really poor education of general population in general who don’t want to question, “well my grandma use that so let’s just use that”.

Edzard Ernst: It’s all of that and much more. There's clearly not one single reason for the popularity of SCAM and if you ask different people, you get different answers. Somebody who's essentially healthy, and probably quite wealthy, has totally different reasons than somebody who's clinging to the last straw, and who feels that mainstream medicine cannot help anymore. So there is a vast array of different reasons, and they need to be weighted differently for different populations. One reason that we haven't mentioned yet, which I think is quite important. Our journalists. The sort of rubbish that that is being published about certain treatments and VIPs like Gwennyth Paltrow and King Charles do their fair share in misleading the public, and I think that that is much more important than generally recognized.