Guidelines for Authors of Books and Papers on Complementary Medicine

A consultation document produced by the Research Council for Complementary Medicine

1) Knowledge Claims
A knowledge claim is a statement about the world. Examples might include:"Massage improves lymph flow" or "The effects of acupuncture vary dependingon how the needle is manipulated.

1.1) Where does your knowledge claim come from?
When making a knowledge claim, it is important to state its origin, inother words, the reasons why you believe it to be true. For example, youmight write: "it has been demonstrated that ....." or "early reportssuggest that..." or "practitioners have long held that....". These phrasesallow readers to assess how much faith they should place in a claim.Not every knowledge claim needs to be preceded by a reference to itsorigin. An author might choose to make an explanatory statement at thebeginning of a whole section (eg. "the following reflects my own personalexperience of treating patients with multiple sclerosis.") Moreover, it isnot necessary to justify a statement if it could not reasonably be doubted:discussions of elementary anatomy or basic aspects of practice("acupuncture involves the insertion of needles through the skin") wouldfall into this category. However, when in doubt, do state why you believesomething to be true. Many books and papers on complementary medicine areflawed because their authors avoided explaining why they have madeparticular claims.

1.2) Justifying a knowledge claim.
There are two general ways of supporting a knowledge claim.  You can eitherrefer to personal experience or quote original scientific research. Theimportant point is not that one is necessarily better than the other, butthat the reader should be able to distinguish between the two. Some authorshave attempted to disguise the fact that their beliefs are based onpersonal experience or collective knowledge by using a thin veil ofscientific sounding language. The general rule to follow when justifying aknowledge claim is: give the reader enough information to be able to decidefor him or herself; don't obscure the origin of your claim.

1.2.1) Using Personal Experience.
It is perfectly valid to justify astatement by referring to personal experience. You can either refer toindividual experience (your own or that of a colleague) or collectiveexperience (traditional knowledge or that of a particular school ofthought.) There are three general principles to follow if you wish to usepersonal experience to justify a claim.Describe the individuals concerned. If you refer to individualexperience, give a brief description of the background, training,experience, clinical style and clinical setting of the practitionerconcerned. This gives a context so that readers can decide whether thepractitioner's experience is applicable to their own patients.Use general statements with care. You should avoid stating that "mostpractitioners believe that ...." or "modern homoeopaths treat by ..."unless you can give evidence that this is indeed the case. You might dothis by referring to text books or to the teaching practices of majorschools. Remember that complementary therapies tend to be practised in anumber of different ways, each sometimes associated with a particularschool of thought. Take care not to assume that all practitioners practisein a certain way or that they hold a certain belief just because you andyour colleagues do.Be careful if you refer to tradition. Some authors justify their ownpractices by defining them as "traditional" or "classical". This causesthree problems. Firstly, authors often fail to provide evidence that atherapy has been practised in a certain way during a particular historicalperiod. Secondly, it is rare that any therapy has ever been practised injust one manner: there are normally a number of different schools ofthought. Thirdly, therapies rarely remain static, they change and progressover time. So it is important to define what you mean by "traditional" or"classical", if you decide to use these terms.

1.2.2) Using Scientific Research.
There are a number of general principles to follow if you wish to use scientific research to justify a knowledge claim.Use references. Avoid using phrases such as "research has indicatedthat....."; "a clinical trial demonstrated...." or "it is now known that..." without giving appropriate references to a book or a journal article. Citing references allows readers to locate and read the work youare using to support your claim. This allows them to decided for themselveswhether your claim is correct and to find any further information thatmight interest them. The publisher or journal you are writing for will havetheir own guidelines as to the form of references you should use.Avoid "secondary sourcing." In general, don't reference an article unlessyou have read the original for yourself. If you quote some research becauseanother author has done so ("secondary sourcing") you are relying on thatauthor's interpretation of the data. There are numerous examples whereauthors have quoted each other's inaccurate data without checking theoriginal paper. This has lead to incorrect information spreading like avirus through the literature. If you do use secondary sourcing, make this clear (eg. "Brown has reported Smith's study of infant massage .......")Don't rely on abstracts. Abstracts - short summaries of a research paperavailable from databases - are often misleading. Authors may claimimportant results in the abstract but close reading of a paper might revealan entirely different story.Use references sparingly. Some authors seem to think that references area bit like steel support rods in a concrete foundation: the more you have,the more secure you are. Large numbers of references may be used to give atext a scientific "feel", regardless of whether the references actuallyshed any light on the subject in hand. Only use a reference if it isgenuinely helpful to do so. Avoid referencing statements such as "sleep is a mental restorative" or "pain is an unpleasant sensation". Use references specifically. One common mistake is to use "shopping lists", where a vague, general claim (eg. "the effects of mental state on physical health are well-known") is followed by a long list of references. Scientific papers address specific questions (eg. "Does mental attitude following mastectomy affect survival in breast cancer?") so you should try to be specific when you link a reference to a knowledge claim. Describe the original research. If a knowledge claim is an important partof your argument, you should spell out what happened in the research you cite. Take a critical stance: you should briefly describe the research method and results and say a few words about any methodological flaws. Don't reference authorities. Some authors have supported knowledge claims by referring to text books written by some well-known expert. Generally, these books merely repeat the claim. If a reference is not to original research, but merely someone's opinion, this should be made clear in the text (eg. "Brown has claimed that" or "according to Brown.") Read research critically. Many research studies contain methodological flaws and cannot be used as the basis for a knowledge claim. Two errors of interpretation are particularly common in reading complementary medicine research. Firstly, it is often inadvisable to make general statements about efficacy of a therapy as a whole on the basis of trials of particulartreatment techniques with particular groups of patients.  Secondly, it isimportant to evaluate all the research evidence on a topic. Some authorsselect just the one of two papers which support their point of view andfail to mention other studies. Make sure that you have all possible databefore starting a book or paper. See Further Reading for more information on locating and reading scientific research.

Only use "weasel words" with extreme care. Words such as "may", "might","could" and so on fail to give a reader useful information. Take a statement such as "high doses of lavender may cause headaches". This could mean either that high dose lavender definitely causes headaches in some patients or that we don't know whether lavender causes headaches but that this might be the case. Weasel words also allow an author to promote apoint of view without being held accountable for it.

Think carefully about causal inferences. Phrases such as "has been linked to" or "has been associated with" can sometimes be used inaccurately to suggest that one thing causes another.

Take care if you cite laboratory-based research. It can be difficult to extrapolate from experiments on animals or human volunteers. For example,just because a certain technique or substance reduces artificially induced inflammation in a rat, this does not imply that it will be effective for patients with arthritis. Similarly, a substance active against bacteria or tumour cells in the laboratory will not necessarily retain this property ina patient's body.

Take care if you quote research from certain journals. The most reliable sources of evidence are peer-reviewed journals: these submit research articles to critical appraisal by researchers and those with knowledge of afield. Also be aware of the possible effects of "publication bias". There is evidence, for instance, that journals from China, Russia or the former USSR do not publish research which finds a therapy to be ineffective. It is relatively easy to "raid" the Chinese or Russian literature for a study supporting the use of complementary therapies in almost any condition. You might want to take this into consideration before citing Chinese or Russian research.

Don't pick and choose your evidence. Many authors only cite papers which supports their views. For them, research is used merely to buttress anargument, a method to vindicate prior beliefs. It is important to includeand evaluate all studies on a particular subject. Research should be used to examine our beliefs critically: it should not just be a rhetorical device.

1.3) Be Specific When You Make a Knowledge Claim.
It is important to avoid making statements about the clinical effects oftreatment which are too general to be of value. Try to be specific in thefollowing ways.

-- Patients. Does the effect occur in everyone, or just in certain patientgroups - those with particular medical conditions or symptom complexes?

-- Technique. Does the effect occur with all applications of a techniques(eg. massage) or only if particular techniques are applied (eg. vigorouskneading)? Similarly, does the effect occur at all typically used doses ordo specific doses need to be used?

-- Clinical value. How big is the effect? What is its clinical value?

-- Frequency. Does the effect always occur or only in a certain proportion of patients?

-- Duration. For how long does the effect last?

2) Concepts
Many of the concepts found in books and papers on complementary medicine are used rather carelessly. Ideas such as holism, energy or "natural" medicine are often taken to have simple and obvious meanings. Yet these concepts are extremely slippery and open to differing interpretations, especially if they are used with insufficient care. Listed below are anumber of common mistakes:

-- Holism. Holistic medicine is a rather vague notion and you should attempta definition if you intend to use it extensively. You would be wise toavoid talking of holistic therapies. Many would argue that holism stems from the way in which health care is practised, not from what particular therapy is used.

-- Nature and the natural. The concept of nature is one which, at firstsight, seems relatively obvious and straightforward. On closer examination,however, "nature" is rather more difficult to pin down. Different societies in different historical periods have held different views of nature and itis difficult to see how some of these could be true and others false. Becareful is saying, for example, that acupuncture is more natural than surgery, or that an essential oil is natural, whereas an antibiotic is not.

-- Metaphors. It is possible that much of what is posited in complementarymedicine is metaphorical. For example, it might be argued that the"meridians" in acupuncture are a useful fiction for linking up real acupuncture points, just as the Greenwich meridian is a useful fiction for navigating the seas. Some authors discuss meridians as if they were just as much part of the physical world as an arm, a leg or an underwater rock. Though this is a perfectly fair position to take, it is helpful if authorsare aware of the possible problem of "confusing the map for the territory"and to state clearly why they believe something to be real or metaphoricaland if so, why.

-- Conventional medicine. There are instances where complementary medicineauthors have made rather simplistic assumptions and criticisms ofconventional medicine. For example, doctors are said to treat only the diseased body part, to suppress symptoms instead of searching for causes and to ignore subjective changes in favour of objective measurement. It is worth remembering that the term "conventional medicine" incorporates a wide variety of practices (for example, nursing, physiotherapy, clinicalpsychology, hospice care). As such, it is misleading to think of conventional medicine merely as "drugs and surgery". Conventional medicine also covers a wide variety of settings. Some of these, such as hospice care, are models of holistic care and do not involve an aggressive,invasive, curative approach to disease. The very diversity of conventional medicine makes resistant to simplistic generalisations. Finally, it isworth mentioning that commenting on the alleged limitations of conventional medicine is often a rather weak method of arguing for the value ofalternatives.

-- Complementary medicine. Some authors talk about complementary medicine asif it were a single entity. It is misleading to do so because complementary medicine is a catch-all phrase to describe a number of very different therapies. If you wish to ascribe a certain practice or set of beliefs to "complementary therapists" remember that this includes everyone fromacupuncturists to yoga teachers, chiropractors and massage practitioners to crystal healers and iridologists. Also remember that many conventionally trained health professionals practise complementary therapies. Moreover,the diversity of complementary medicine is such that it may be misleading even to talk about one therapy as if it constituted a single set of practices and beliefs. Acupuncture is a good example of a diverse therapy:it may be practised based on the theory of the five elements, or the eight principles, or trigger points, or Western diagnoses ("formula acupuncture"). Needles may be inserted shallowly (as in Japan) or deeply(as in China) and may be manipulated or left in place. Some therapists practise acupuncture as a stand-alone therapy, others integrate with herbs and other traditional Chinese therapies. This limits generalisations about acupuncture as a whole.

4) Scholarship
our principles of good scholarship are as follows:

-- Thoroughness and attention to detail. Make sure that every word is speltcorrectly, that the language is consistent throughout the text, that every reference is complete and correct and that any diagrams areself-explanatory and 100% accurate. When you have completed a section, ask yourself: have I really included all necessary and sufficient information?

-- Reflect on the validity of every point you make. Each and every time youmake a claim, or state an argument, take a moment to ask yourself: "Is this really true? Are there circumstances where it is not true?  In what ways might I have over-generalised or oversimplified?" Then ask: "Does the conclusion follow reasonably from the premises?" Finally, ask: "How can Ibe sure that my argument holds? What is the evidence? And if I am not sure,how should I communicate this to the reader?" One useful trick is to play"devil's advocate": imagine yourself as someone who disagrees with the point being made. How would you argue against it, having read the text? What might be an alternative explanation or interpretation?

-- Reflect on the limitations of your argument. You might wish to comment on your methodology (eg. "it may be that the traditional texts do not reflect accurately reflect how acupuncture was practised at the time") or examine opposing arguments or points of view. It is generally wise to be cautious.

-- Disinterestedness. Traditionally, it has been held that scholars should retain a suitable distance from their work. In the same way that a judge should hold no opinion on the guilt or innocence of the accused, but work to ensure that the truth emerges from the court hearing, so it was thought that the scholar should refrain from promoting a particular viewpoint andallow the facts to speak for themselves. Though this is now thought to be an inaccurate and undesirable ideal of scholarship, distrust of undue advocacy remains. It can sometimes be a useful discipline to write as though you had no personal stake in what readers believed following their reading your work, just as the judge has no personal stake in whether theaccused is sent to jail or set free.On the other hand, one thing scholarship should not be is:

-- Sounding academic for the sake of it. Some authors seem to think that theuse of long words, dense and lengthy prose and obscure jargon can by itself make a work scholarly. Such techniques are the equivalent of using too many references: they mistake form (what the argument looks like) for content (is the argument of value?). As a general guideline: don't use a long or unusual word where a short or common word will do; avoid or explain jargon; keep to one idea per sentence; only include something if it is genuinely useful to do so; remember that words can obscure; keep simplicity and clarity as your aim.

5) Further Information.
Critical Appraisal
Some useful papers on critical appraisal include:
Guyatt GH et al. Users' Guides to the Medical Literature. II. How to use an article about therapy or prevention. Journal of the American Medical Association  1994 ;271(1):59-63
Vickers AJ. Critical Appraisal: How to Read a Clinical Research Paper.Complementary Therapies in Medicine 1995;3:158-66
Mays N, Pope C. Rigour and Qualitative Research. British Medical Journal1995;311:109-112
Evans SJW. Good Surveys Guide. British Medical Journal 1991;302:302-3
Locating Research and Information on Complementary Medicine
The Research Council for Complementary Medicine holds CISCOM, an electronicdatabase of research references on complementary medicine. They also produce an information pack, Searching for Published Information in Complementary Medicine: a resource guide. Contact: RCCM, 60 Great OrmondStreet, London WC1N 3JF. Tel:0171 833 8897..