Download Lectures on Biostatistics (1971).
Corrected and searchable version of Google books edition

Download review of Lectures on Biostatistics (THES, 1973).

Latest Tweets

Jump to follow-up

The King’s Fund recently published Assessing complementary practice Building consensus on appropriate research methods [or download pdf].

Report title

It is described as being the “Report of an independent advisory group”. I guess everyone knows by now that an “expert report” can be produced to back any view whatsoever simply by choosing the right “experts”, so the first things one does is to see who wrote it.  Here they are.

  • Chair: Professor Dame Carol Black
  • Harry Cayton, Chief Executive, Council for Healthcare Regulatory Excellence
  • Professor Adrian Eddleston, then Vice-Chairman, The King’s Fund
  • Professor George Lewith, Professor of Health Research, Complementary and Integrated Medicine Research Unit, University of Southampton
  • Professor Stephen Holgate, MRC Clinical Professor of Immunopharmacology, University of Southampton
  • Professor Richard Lilford, Head of School of Health and Population Sciences, University of Birmingham

We see at once two of the best known apologists for alternative medicine, George Lewith (who has appeared here more than once) and Stephen Holgate

Harry Cayton is CEO of Council for Healthcare Regulatory Excellence (CHRE) which must be one of the most useless box-ticking quangos in existence. It was the CHRE that praised the General Chiropractic Council (GCC) for the quality of its work.  That is the same GCC that is at present trying to cope with 600 or so complaints about the people it is supposed to regulate (not to mention a vast number of complaints to Trading Standards Offices).  The GCC must be the prime example of the folly of giving government endorsement to things that don’t work. But the CHRE were not smart enough to spot that little problem.  No doubt Mr Cayton did good work for the Alzheimer’s Society.  His advocacy of patient’s choice may have helped me personally.  But it isn’t obvious to me that he is the least qualified to express an opinion on research methods in anything whatsoever. According to the Guardian he is “BA in English and linguistics from the University of Ulster; diploma in anthropology from the University of Durham; B Phil in philosophy of education from the University of Newcastle.”

Adrian Eddlestone is a retired Professor of Medicine. He has been in academic administration since 1983. His sympathy for alternative medicine is demonstrated by the fact that he is also Chair of the General Osteopathic Council, yet another “regulator” that has done nothing to protect the public
from false health claims (and which may, soon, find itself in the same sort of trouble as the GCC).

Richard Lilford is the only member of the group who has no bias towards alternative medicine and also the only member with expertise in clinical research methods  His credentials look impressive, and his publications show how he is the ideal person for this job. I rather liked also his article Stop meddling and let us get on.. He has written about the harm done by postmodernism and relativism, the fellow-travellers of alternative medicine.

Most damning of all, Lewith, Eddlestone and Holgate (along with Cyril Chantler, chair of the King’s Fund, and homeopaths, spiritual healers and Karol Sikora) are Foundation Fellows of the Prince of Wales Foundation for Magic Medicine, an organisation that is at the forefront of spreading medical misinformation.

I shall refer here to ‘alternative medicine’ rather than ‘complementary medicine’ which is used in the report. It is not right to refer to a treatment as ‘complementary’ until such time as it has been shown to work. The term ‘complementary’ is a euphemism that, like ‘integrative’, is standard among alternative medicine advocates whose greatest wish is to gain respectability.

The Report

Kings Fund logo

The recommendations

On page 10 we find a summary of the conclusions.

The report identifies five areas of consensus, which together set a framework for moving forward. These are:

  • the primary importance of controlled trials to assess clinical and cost effectiveness.
  • the importance of understanding how an intervention works
  • the value of placebo or non-specific effects
  • the need for investment and collaboration in creating a sound evidence base
  • the potential for whole-system evaluation to guide decision-making and subsequent research.

The first recommendation is just great. The rest sound to me like the usual excuses for incorporating ineffective treatments into medical practice. Notice the implicit assumption in the fourth point
that spending money on research will establish “a sound evidence base". There is a precedent, but it is ignored. A huge omission from the report is that it fails to mention anywhere that a lot of research has already been done.

Much research has already been done (and failed)

The report fails to mention at all the single most important fact in this area. The US National Institutes of Health has spent over a billion dollars on research on alternative medicines, over a period
of more than 10 years. It has failed to come up with any effective treatments whatsoever. See, for example Why the National Center for Complementary and Alternative Medicine (NCCAM) Should Be Defunded;   Should there be more alternative research?;   Integrative baloney @ Yale, and most recently, $2.5B Spent, No Alternative Med Cures found. .

Why did the committee think this irrelevant? I can’t imagine. You guess.

The report says

“This report outlines areas of potential consensus to guide research funders, researchers, commissioners and complementary practitioners in developing and applying a robust evidence base for complementary practice.”

As happens so often, there is implicit in this sentence the assumption that if you spend enough money evidence will emerge. That is precisely contrary to the experence in the USA where spending a billion dollars produced nothing beyond showing that a lot of things we already thought didn’t work were indeed ineffective.

And inevitably, and tragically, NICE’s biggest mistake is invoked.

“It is noteworthy that the evidence is now sufficiently robust for NICE to include acupuncture as a treatment for low back pain.” [p ]

Did the advisory group not read the evidence used (and misinterpeted) by NICE? It seems not. Did the advisory group not read the outcome of NIH-funded studies on acupuncture as summarised by Barker Bausell in his book, Snake Oil Science? Apparently not. It’s hard to know because the report has no references.

George Lewith is quoted [p. 15] as saying “to starve the system of more knowledge means we will continue to make bad decisions”. No doubt he’d like more money for research, but if a billion dollars
in the USA gets no useful result, is Lewith really likely to do better?

The usual weasel words of the alternative medicine industry are there in abundance

“First, complementary practice often encompasses an intervention (physical treatment or manipulation) as well as the context for that intervention. Context in this setting means both the physical setting for the delivery of care and the therapeutic relationship between practitioner and patient.” [p. 12]

Yes, but ALL medicine involves the context of the treatment. This is no different whether the medicine is alternative or real. The context (or placebo) effect comes as an extra bonus with any sort of treatment.

“We need to acknowledge that much of complementary practice seeks to integrate the positive aspects of placebo and that it needs to be viewed as an integral part of the treatment rather than an aspect that should be isolated and discounted.” [p. 13]

This is interesting. It comes very close (here and elsewhere) to admitting that all you get is a placebo effect, and that this doesn’t matter. This contradicts directly the first recommendation of the House of Lords report (2000).. Both the House of Lords report on Complementary and Alternative Medicine, and the Government’s response to it, state clearly

“. . . we recommend that three important questions should be addressed in the following order”. (1) does the treatment offer therapeutic benefits greater than placebo? (2)  is the treatment safe? (3) how does it compare, in medical outcome and cost-effectiveness, with other forms of treatment?.

The crunch comes when the report gets to what we should pay for.

“Should we be prepared to pay for the so-called placebo effect?

The view of the advisory group is that it is appropriate to pay for true placebo (rather than regression to the mean or temporal effects).” [p 24]

Perhaps so, but there is very little discussion of the emormous ethical questions:that this opinion raises: 

  • How much is one allowed to lie to patients in order to elicit a placebo effect?
  • Is is OK if the practitioner believes it is a placebo but gives it anyway?
  • Is it OK if the pratitioner believes that it is not a placebo when actually it is?
  • Is it OK for practitioners to go degrees taught by people who believe that it is not a placebo when actually it is?

The report fails to face frankly these dilemmas.  The present rather absurd position in which it is considered unethical for a medical practitioner to give a patient a bottle of pink water, but
perfectly acceptable to refer them to a homeopath. There is no sign either of taking into account the cultural poison that is spread by telling people about yin, yang and meridians and such like preposterous made-up mumbo jumbo.  That is part of the cost of endorsing placebos. And just when one thought that believing things because you wished they were true was going out of fashion

Once again we hear a lot about the alleged difficulties posed by research on alternative medicine. These alleged difficulties are, in my view, mostly no more than excuses. There isn’t the slightest
difficulty in testing things like herbal medicine or homeopathy, in a way that preserves all the ‘context’ and the ways of working of homeopaths and herbalists. Anyone who reads the Guardian knows
how to do that.

In the case of acupuncture, great ingenuity has gone into divising controls. The sham and the ‘real’ acupuncture always come out the same. In a non-blind comparison between acupuncture and no acupuncture the latter usually does a bit worse, but the effects are small and transient and entirely compatible with the view that it is a theatrical placebo.

Despite these shortcomings, some of the conclusions [p. 22] are reasonable.

“The public needs more robust evidence to make informed decisions about the use of complementary practice.

Commissioners of public health care need more robust evidence on which to base decisions about expenditure of public money on complementary practice.”

What the report fails to do is to follow this with the obvious conclusion that such evidence is largely missing and that until such time as it is forthcoming there should be no question of the NHS paying for alternative treatments.

Neither should there be any question of giving them official government recognition in the form of ‘statutory regulation’. The folly of doing that is illustrated graphically by the case of chiropractic which is now in deep crisis after inspection of its claims in the wake of the Simon Singh defamation case. Osteopathy will, I expect, suffer the same fate soon.

In the summary on p.12 we see a classical case of the tension

Controlled trials of effectiveness and cost-effectiveness are of primary importance

We recognise that it is the assessment of effectiveness that is of primary importance in reaching a judgement of different practices. Producing robust evidence that something works in practice – that it is effective – should not be held up by the inevitably partial findings and challenged interpretations arising from inquiries into how the intervention works.

The headline sounds impeccable, but directly below it we see a clear statement that we should use treatments before we know whether they work.  “Effectiveness”, in the jargon of the alternative medicine business, simply means that uncontrolled trials are good enough. The bit about “how it works” is another very common red herring raised by alternative medicine people. Anyone who knows anything about pharmacology that knowledge about how any drug works is incomplete and often turns out to be wrong. That doesn’t matter a damn if it performs well in good double-blind randomised controlled trials.

One gets the impression that the whole thing would have been a lot worse without the dose of reality injected by Richard Lilford. He is quoted as a saying

“All the problems that you find in complementary medicine you will encounter in some other kind of treatment … when we stop and think about it… how different is it to any branch of health care – the answer to emerge from our debates is that it may only be a matter of degree.” [p. 17]

I take that to mean that alternative medicine poses problems that are no different from other sorts of treatment. They should be subjected to exactly the same criteria. If they fail (as is usually the case) they should be rejected.  That is exactly right.  The report was intended to produce consensus, but throughout the report, there is a scarcely hidden tension between believers on one side, and Richard Lilford’s impeccable logic on the other.

Who are the King’s Fund?

The King’s Fund is an organisation that states its aims thus.

“The King’s Fund creates and develops ideas that help shape policy, transform services and bring about behaviour change which improve health care.”

It bills this report on its home page as “New research methods needed to build evidence for the effectiveness of popular complementary therapies”. But in fact the report doesn’t really recommend ‘new research methods’ at all, just that the treatments pass the same tests as any other treatment. And note the term ‘build evidence’.  It carries the suggestion that the evidence will be positive.   Experience in the USA (and to a smaller extent in the UK) suggests that every time some good research is done, the effect is not to ‘build evidence’ but for the evidence to crumble further

If the advice is followed, and the results are largely negative, as has already happened in the USA, the Department of Health would look pretty silly if it had insisted on degrees and on statutory regulation.

The King’s Fund chairman is Sir Cyril Chantler and its Chief Executive is Niall Dickson.  It produces reports, some of which are better than this one. I know it’s hard to take seriously an organisation that wants to “share its vision” withyou, but they are trying.

“The King’s Fund was formed in 1897 as an initiative of the then Prince of Wales to allow for the collection and distribution of funds in support of the hospitals of London. Its initial purpose was to raise money for London’s voluntary hospitals,”

It seems to me that the King’s Fund is far too much too influenced by the present Prince of Wales. He is, no doubt, well-meaning but he has become a major source of medical misinformation and his influence in the Department of Health is deeply unconstitutional.  I was really surprised to see thet Cyril Chantler spoke at the 2009 conference of the Prince of Wales Foundation for Integrated Health, despite having a preview of the sort of make-believe being propagated by other speakers. His talk there struck me as evading all the essential points. Warm, woolly but in the end, a danger to patients. Not only did he uncritically fall for the spin on the word “integrated”, but he also fell for the idea that “statutory regulation” will safeguard patients.

Revelation of what is actually taught on degrees in these subjects shows very clearly that they endanger the public.

But the official mind doesn’t seem ever to look that far. It is happy ticking boxes and writing vacuous managerialese. It lacks curiosity.


The British Medical Journal published today an editorial which also recommends rebranding of ‘pragmatic’ trials.  No surprise there, because the editorial is written by Hugh MacPherson, senior research fellow, David Peters, professor of integrated healthcare and Catherine Zollman, general practitioner. I find it a liitle odd that the BMJ says “Competing Interests: none. David Peters interest is obvious from his job description. It is less obvious that Hugh MacPherson is an acupuncture enthusiast who publishes mostly in alternative medicine journals. He has written a book with the extraordinary title “Acupuncture Research, Strategies for Establishing an Evidence Base”. The title seems to assume that the evidence base will materialise eventually despite a great deal of work that suggests it won’t. Catherine Zollman is a GP who is into homeopathy as well as acupuncture. All three authors were speakers at the Prince of Wales conference, described at Prince of Wales Foundation for magic medicine: spin on the meaning of ‘integrated’.

The comments that follow the editorial start with an excellent contribution from James Matthew May. His distinction between ‘caring’ and ‘curing’ clarifies beautifully the muddled thinking of the editorial.

Then a comment from DC, If your treatments can’t pass the test, the test must be wrong. It concludes

“At some point a stop has to be put to this continual special pleading. The financial crisis (caused by a quite different group of people who were equally prone to wishful thinking) seems quite a good time to start.”

Print Friendly

19 Responses to King’s Fund reports on alternative medicine: little consensus and less progress

  • I expect better.

    It would have been more satisfying if they’d asked you to chair the whole thing.

  • Superb!

    It would be fine if someone was suggesting that some new, promising, plausible therapy be tested, but the ones they are promoting have been debunked so many times.

    If they choose not to believe the proper evidence then let them fund their own research, but stop conning the public!

  • Stephen Holgate is pro-woo??

  • Great post!

    Nevertheless it is a dire state of affairs with all these reports and regulatory bodies. The people who run them are either biased or have a vested interest, and have no plans to tease out the truth from the crap.

    You would think the billions of dollars spent in the US would have put this to rest by now, but for every 1 good piece of research the quacks churn out a crap one. Then they all focus and quote from the crap one.

  • David, thanks very much for this excellent analysis.

    A few comments on the report’s statement: “The view of the advisory group is that it is appropriate to pay for true placebo (rather than regression to the mean or temporal effects)”

    The concept of avoiding “regression to the mean or temporal effects” is mentioned on every third page on average. I wondered if this was because they had only just discovered that a controlled trial controls for regression to the mean and temporal effects, of if they wanted readers to forget that there are other biases to be concerned about.

    I declare up front that I believe (perhaps controversially) that the placebo effect is under-researched and underused in conventional health care.

    I also believe that some people include bias effects in the concept of placebo effect. I think that this view is not helpful to the scientific examination of placebo effects, as I hope the following discussion shows.

    The report advocates the use of pragmatic trials (or as they term them “effectiveness trials”) for alternative medicine interventions. Such trials are usually open, and the lack of blinding can lead to serious contamination of the results by biases. This likelihood is ignored in the report. This convenient oversight is severely problematic.

    Outcome measurements from an open pragmatic trial comparing a test treatment with usual best care reflect the difference in:
    (i) specific effects
    (ii) placebo effects
    (iii) bias effects

    Let us see how this works for acupunture for back pain where the outcome measurements are patient-assessed pain and disability.

    (i) the specific effect of acupuncture has been shown to be clinically unimportant. The specific effects of usual best care are small. So the results reflect the net effects of placebo and bias.

    (ii) the placebo effects of acupuncture in this sort of trial are unknown, but are likely to be positive in the test group (who are getting extra attention by therapists who believe in their treatment). The placebo effects in the control group are likely to be negative as they will feel that they are not getting the extra attention and possibly effective treatment being given to the test group.

    (iii) the size of bias effects is unknown, but it is likely that cognitive biases such “see what you want to see” and “tell them what you think they want to hear” will be positive in the treatment group and negative in the control group.

    The outcome measurement will not, as the report assumes, just reflect the specific and positive placebo effects. It will reflect any negative placebo effects in the control group, and the cognitive biases in both groups. Results from such trials can therefore not provide reliable evidence unless they are supported by evidence from other sources that negative placebo effects in the control group and cognitive biases in both groups are clinically unimportant.

  • Today there was an editorial in the BMJ along the same lines as the King’s Fund report, that is also rebranding pragmatic trials: http://www.bmj.com/cgi/content/extract/339/sep01_2/b3335

  • The BMJ editorial Twaza linked sounds terribly reasonable, but – as s/he alludes to – is a bit of a Trojan horse. The conclusion Twaza aptly summarises is perhaps unsurprising when one notes that the three authors are the founder of the Northern College of Acupuncture (Hugh MacPherson), David Peters of the University of Westminster (a familiar figure to readers of DC’s blog) and Catherine Zollman, another long-time enthusiast for CAM. Incidentally, Hugh MacPherson’s list of publications reveals a string of articles co-authored with George Lewith (see DC articles passim.. And all three authors are – believe it or not – Foundation Fellows of the Prince’s Foundation for Integrated Health (now where have we just been hearing about them?).

    Sometimes it does feel like a very small – or perhaps a better word would be “cosy” – CAM world.

    While the BMJ editorial is behind a paywall, the online comments thread is free access. There is already a common sense response there from James Matthew May, which gently but astutely points out some of the sleight-of-hand inherent in the way CAM people approach this area. Here’s hoping the print BMJ runs his letter in full.

  • PS I see DC and Edzard Ernst have now weighed in over at the BMJ comments thread.

  • Dr Aust, I can’t see what you say you see! The link you gave is to a response by James Matthew May.


  • All three responses are there when I click it – try refreshing the page or clearing the browser cache.

  • Dr Aust, I see the responses now, thanks; the browser must have refreshed its cache.

  • I don’t have access to the full BMJ editorial. Does it at least acknowledge the possibility that this “evidence gap” may be there simply because the treatments don’t work?

    In many cases, of course, it may well be that there isn’t an “evidence gap”. What needs to be dealt with with is a gap between belief and reality.

  • Yes, but ALL medicine involves the context of the treatment. This is no different whether the medicine is alternative or real. The context (or placebo) effect comes as an extra bonus with any sort of treatment.

    Excellent point – and one worth emphasising.

    The corollary of this is that if we fund alternative medicine on the grounds that the placebo effect is stronger than in conventional medicine, than we will have arrived at that state through the consistent and systematic undermining of conventional medicine, thus weakening the strength of the placebo effect there. It won’t have happened because somehow the placebo effect was magically better to begin with in the alternative branch.

    This, of course, is a major reason why claiming that the placebo effect is good grounds for assigning public funds for research into alternative medicines is highly unethical.

  • Mojo

    You have put your finger on the big problem for both the Editorial and the King’s Fund report. The possibility that the treatments might not work is barely envisaged, despite the fact that the US experience suggests that would be the outcome.

    There is everywhere an implicit assumption that if you pour enough money into it, (positive) evidence will eventually emerge.

  • Well, do enough well designed trials and positive evidence will emerge. The problem is in dealing with the other 95% of the evidence. ;)

  • May I refer readers to my latest blog entry: http://www.orthobase.org. This touches on this subject.

Leave a Reply