Intended for healthcare professionals

Education And Debate

The role of complementary and alternative medicine

BMJ 2000; 321 doi: (Published 04 November 2000) Cite this as: BMJ 2000;321:1133

High Self-Citation Rates and a raft of Opinions - is this research?


A range of important criticisms can be levelled at this article by
Ernst [1]. One presumes that BMJ readers would wish to reach a sober,
detached and completely balanced view of CAM modalities. This is an
important topic in modern medicine and has been described as posing a
major “challenge to regular medicine” [2]. Ernst’s numerous publications
simply fail to provide such balance. Instead, he displays an apparent
addiction to self-referencing and focuses solely on reductionist studies.
His alleged EBM approach also omits to say ‘observational studies can be
as valid as randomized control trials’ [3]. I would strongly encourage
readers to seek out and obtain more balanced perspectives on CAM,
especially those increasingly numerous studies from social scientists, who
are less obviously partisan and reductionist than the medically qualified,
who habitually give only one side of the story [3].

“[The]...reductionist probing of the molecular causes of
becoming prosaic.” [4]. And, any "attempt to reduce the vast complexity of
human life to a few comparatively simple mechanisms has an undistinguished
history" [5]. "While this surge of biological reductionism is
understandable, it risks throwing the baby out with the bath water" [6].
"Issues that are complex, multidimensional, and grounded in individual
experience lend themselves to study by descriptive and qualitative
methods." [7] Such an as-is approach is central to all CAM modalities, yet
it is signally absent from Ernst's work.

As Dr Heptonstall's has pointed out [8], Dr Ernst certainly does seem
to rely heavily on referencing his own work. This deplorable habit makes
it difficult for others to form a balanced evaluation of a sole researcher
in a new field, to develop respect for, or to place much faith in their
work. It appears to bolster all one’s arguments and opinions on previously
published material; yet, such self-citations are about as valuable as
chaff in the wind and comprise a thinly disguised academic deceit if they
continually rely on one’s own previous publications. I think it is self-
evidently true, in any field, that neutral and balanced research thrives
on ‘cross-pollination’ between different researchers. Anyone who deviates
from that must attract suspicion.

As he rarely if ever cites social science perspectives on CAM, he
could be unaware of them or choose not to discuss or publicise them -
presumably because he disagrees with them, does not understand them or
wishes to deny their validity by ignoring them. Either way, he denies BMJ
readers access to an important slant on this matter. Either or both of
these shortcomings could be purely accidental, of course, but they still
deny readers the full picture. This serious failing casts a shadow over
the real quality of Ernst’s work. How can one confidently place any trust
in such an unbalanced approach?

One thinks of Professor Sir Cyril Burt, for example, and his
pioneering work [in the 30s] on twins, when it was discovered [in the
70s], that he had fiddled his results on a massive scale for decades [9].
Although, clearly, Ernst does not stand accused of any such crime, it does
highlight the danger of being a sole, pioneer in a new field. Any type of
'unbiased scientific evaluation' [1] of his work by others, is at a poorly
developed stage, his own prejudices are almost bound to be mirrored in his
work, and much that is really 'dross', could in the short run be regarded
as 'gold'.

He states:
"Complementary and alternative medicine is largely opinion based." [1]

This is a particularly foolish thing to say, because the rest of his
article comprises merely a raft of his own opinions - 16 of the 24
references [67%] are to his own work [10]. These opinions are therefore
stacked like castles in the air - and supported by what? Opinions are
usually based upon experience and neutral research, not conducted by one
person, but by many. On what basis should the opinion of Ernst be
preferred as deserving greater respect than the collective experience of
CAM therapists or other researchers? It is denigrating, dismissive and
belittling to use this term 'mere opinion'. By portraying therapeutic
efficacy as 'mere placebo', and the views and observations of thousands of
practitioners as 'mere opinion', seems a tad too severe and does not seem
to denote a neutral, sober or even-handed evaluation of CAM in the year
2000. His impatience to denigrate CAM suggests anything but an unbiased

In no sense can this article be regarded as a fair appraisal of 'the
role of CAM' at this point in time, as the title of the paper misleadingly
suggests. Indeed, it is wrongly titled, as it fails even to acknowledge
any role at all for CAM in modern medicine, let alone discusses it. Being
a highly critical and dismissive appraisal, it also claims that 7 out of
11 [77%] CAM modalities have no therapeutic value [1] and cannot be shown
to have any usefulness. Even a child could see that such a claim flies in
the face of patients flocking to these therapies, around the world, in
ever-increasing numbers, and also in the face of the accumulated skills of
CAM practitioners, who he should credit with more brains, as they daily
and yearly accrue many sound observations of the clinical usefulness of
these modalities. What actually makes his opinion so superior to everyone

Why should anyone denounce and dismiss the 'mere opinions' of
millions of patients and thousands of satisfied practitioners, or the
research of social investigators, and select instead the 'opinion' of one
researcher, who mostly references his own publications? Does this denote
someone who has neutrally empathised with his subject of study, or one who
seeks to stand above it, and to systematically demolish its credibility
before the medical profession? On what basis can such conduct inspire
respect for detached and scrupulously neutral work? Stated simply, it does
not. In the final analysis, and to arrive at a sober judgement, one must
try to weigh up who is providing the most balanced and neutral account of
this matter, and why. Clinicians need to place together a range of
approaches to that of Ernst in order to arrive at a wider view of CAM.

Finally, consider the following quotes:
“People turn to alternatives because they have become disillusioned with
conventional medicine which has failed to deliver the promise of
eradicating suffering and providing good health.” [11]

“People turn to alternatives because of dissatisfaction with the
doctor-patient relationship...[they] spend too little time with and have
little respect for their patients.” [11]

“Today’s medicine, it is argued, can best be described as Fordist
medicine, which emphasises quantity rather than quality and produces
alienated and dissatisfied patients.” [11]

“...they believe GPs spend too little time with patients...and
believe that [they] do not listen to what their patients have to say.”

Such valid, patient-centred, and human observations are never even
mentioned by Ernst, let alone discussed. OK, so he offers 'rejection of
science and technology', 'rejection of the establishment' and
'desperation' as reasons. By gazing solely through the ‘conceptual
spectacles’ of reductionist RCTs, systematic reviews and meta-analyses, he
seems content to blithely ignore the other half of medicine - the human

Siapush again:
“It is dissatisfaction with the medical encounter that leads to a
favourable attitude towards alternatives. People turn away from
orthodoxy...because of the way they are treated by doctors.” [11]

“Although studying attitudes is important...the examination of
behaviour proper provides a more valid investigation of the reasons for
the growth of alternative medicine.” [11]

This is valuable material that deserves consideration and that does
comprise "a challenge to regular medicine" [2]. Moreover, what is the
observed behaviour of actual patients? They flock to these CAM modalities,
year on year. Why?
“The widespread use of unconventional therapies...shows that patients are
seeking treatment which conventional scientific medicine cannot provide.”

Yet, Ernst still insists:
“No single determinant of the present popularity of complementary medicine
exists.” [1]

Such a brazenly inaccurate comment beggars belief. He might just as
well come clean with his readers and say: ‘I saw nothing because I did not
even bother to look’. Such a position is just about as banal as that of
Pope Urban VIII, who refused to look through Galileo’s telescope at the
moons of Jupiter ‘because there aren’t any there according to scripture’.
I think that reflects the dismal quality of Ernst’s so-called ‘research’.


[1] Ernst, E, The Role of Complementary and Alternative Medicine, BMJ
2000; 321:1133-1135 (4 November)

[2] BMJ 1994, 309; 1669 [17 December] Medicine and Books, Challenging
Medicine, Ed Jonathan Gabe, David Kelleher, Gareth Williams, review by
Donald L Madison

[3] David L Schriger, One is the Loneliest Number: Be Skeptical of
Evidence Summaries based on limited literature reviews, Annals of
Emergency Medicine, 36.5, November 2000, 517-19

[4] Education and Debate, Clinical academic medicine: a Socratic
dialogue, BMJ 1997, 315, 593-5 [6 September] D G Grahame-Smith

[5] BMJ 1998; 317:1728-1728 (19 December), Sacred cows: to the
abattoir! The promise of the neurosciences, A M Daniels,

[6] BMJ 1996; 313:957-958 (19 October), Editorials, "Is my practice
evidence-based?" T Greenhalgh,

[7] BMJ 1999; 319:1296 [full] (13 November), The impact of
informatics, Universities without walls: evolving paradigms in medical
education, Roderick Neame, Brooke Murphy, Frank Stitt, Mark Rake,

[8] BMJ letter, New Definition Required, John Heptonstall, 9 Nov 2000

[9] Cyril Burt links:

[10] Two further examples - in the article "Complementary medicine:
From quackery to science?" [Editorial], Journal of Laboratory &
Clinical Medicine, 127(3): 244-245, March 1996, 6 out of the 13 references
are to Ernst's own work = 45% self-citation. In the article: "Location
bias in controlled clinical trials of complementary medicine", J of Clin
Epidemiol., 53, 2000, 485-9, of the 37 references, 5 are to Ernst himself
= 14%.

Some Ernst BMJ self-citation rates: 2 in 16 [12.5%] 1 in 3 [33%] 2 in 12 [16.5%] 2 in 3 [67%] 1 in 13 [7.7%] 1 in 3 [33%] 3 in 13 [23%] 1 in 3 [33%] 1 in 1 [100%] 1 in 3 [33%] 2 in 5 [40% 4 in 13 [31%] 2 in 3 [67%]

This sample yields a mean self-citation rate of 38.2% with a range of
7.7 to 100%. Although this is bad enough, if we include four or five of
his regular co-authors [such as White, Pittler, Abbott, Assendelft], then
their pooled mutual and self-citation rates reach an even higher average.
I am sure it would be very easy to show that these self-citation rates are
extremely high if compared to most other BMJ articles or academia in

[11] M Siahpush, Postmodern values, dissatisfaction with conventional
medicine and popularity of alternative therapies, J of Sociol, 34.1, March
1998, 58-70

[12] J J Chan and J E Chan, Medicine for the Millennium: the
challenge of postmodernism, Medical Journal of Australia, 2000, 172, 332-4

Competing interests: No competing interests

09 December 2000
Peter Morrell
Hon Research Associate, History of Medicine
Staffordshire University