The history of research in homeopathy demonstrates the difficulty in proving its effectiveness, though homeopathy continues to be an attractive therapeutic intervention1. Medicines used in homeopathy are substances which have been serially agitated and highly diluted (SAHDs). Thus, following the model of accepted biomedical research on drug action, rigorously designed clinical trials focused on testing the difference between SAHDs and placebo. Many of these demonstrated no statistical difference. Or, where they had, the studies proved difficult to reproduce (ibid). This led to the conclusion that success in homeopathy was due to non-specific or placebo effects2. Such results have spurred researchers on to question the validity of placebo-controlled trials as a valid test of homeopathic practice, given the complexity of both homeopathy as an intervention, and the nature of chronic disease3,4. Designing valid research which can test the effectiveness of homeopathy is currently of the utmost importance if the profession is to survive in the present medical-political climate, in which some of Britain’s most prominent doctors are calling for the NHS to stop funding homeopathic hospitals, on the basis that homeopathy is a ‘bogus’ therapy5.
Additionally, with a greater understanding of complementary and alternative medicines (CAM) in general, there is a potential for improvement in health care delivery6. In a paper published in the BMJ, Mason (2002) called for CAM to be evaluated rigorously, but pointed out that reductionist research methods such as randomised controlled trials (RCTs) fail to take into account the holistic nature of CAM. Evidence of effectiveness could be provided if: 1) well designed RCTs do not misrepresent CAM’s holistic essence, 2) practitioners are recognised as a component and 3) both specific and non-specific outcome measures with long follow up are included7.
Advocating whole systems research (WSR) in complementary and alternative medicine8 (Verhoef et al 2005), signals a movement away from RCTs and a growing interest in qualitative research. The paper by T Thompson and M Weiss (2006) Homeopathy – what are the active ingredients? An exploratory study using the UK Medical Research Council’s framework for the evaluation of complex interventions is an example of whole systems research (WSR) in action9. This paper centres on a critique of Thompson’s prospective case-series study, because it sets an important example as the first to implement WSR in homeopathy. WSR methodology is more appropriate then RCTs in providing real-world validation of the effectiveness of homeopathy. Thompson’s study highlights homeopathy as a complex therapeutic intervention, as it explores which aspects of the intervention are active ingredients of the treatment, and which of these are non-specific or specific to homeopathy. As a formal investigation, both qualitative and quantitative evidence is gathered which provide a higher level of evidence base than a single anecdotal case report10. This study introduces a number of innovative and original devices into homeopathy research, which can be taken up, used and developed by others in the profession, to the advantage of both patients and providers.
RCTs and Homeopathy
As stated earlier, RCT’s have attempted to test the homeopathic remedy, or SAHD, as an active ingredient because RCTs are considered the best method to test for isolated specific effects of a drug11. However, RCTs cannot accurately reflect homeopathy as a system of therapeutic intervention, partly because a basic tenet of homeopathy is that no specific medicine is given for a specific condition as in conventional medicine. The system of homeopathy requires that each individual receive a medicine based on his or her idiosyncratic symptom pattern. Even when individualized homeopathic prescribing was combined with double-blinding and placebo control, the researchers challenged the assumption that the specific effects of the homeopathic medicine can be separated from the non-specific effects of the consultation4. In addition, the authors questioned the legitimacy of placebo use in this context, since the homeopathic practitioners could not evaluate remedy responses in the normal manner. Normally, follow-up prescriptions may or may not change as a result of the homeopath’s assessment of reaction to the remedy. Instead, the prescribers became confused, as it was difficult to assess the meaning of the patients’ responses.
For CAM generally, there is a tension concerning RCT as the research model of choice, which Jonas et al (2002), describe most comprehensively. Because CAM interventions usually involved a combination of components acting synergistically, these authors maintain that RCT’s alone cannot provide enough information:
“Placebo-controlled RCTs are precise instruments designed to dissect a single hypothesis of pharmacological specificity. In CAM interventions the hypothesis is much more diverse.”
Jonas, Lewith, Walach (2002:33)
They recognize that the theoretical models which underpin CAM processes involve stimulating the whole organism’s self-healing response, utilizing specific and non-specific treatment effects. For example RCTs are not suited to measure interactions or therapeutic relationships, since by their nature they cannot be reduced to one active ingredient. Interestingly, they also consider whether pharmacologically specific effects in conventional medicine might also be working synergistically with psychological and other effects of general care12. ‘Synergy’ means combined, enhanced effects: the effect is more than the sum of the parts acting individually13.
A reversed research hierarchy is proposed, where the highest consideration is given to assessment of general effectiveness of the treatment by prospective outcome studies with large numbers over long periods of time. Outcome studies provide high external, real-world validity, by reflecting more accurately day-to-day practice, though they are considered to be of low internal validity as bias is not controlled. These should complement RCTs, with their high internal but low external validity1.
With regard to homeopathy in particular, Walach and Jonas have this to say about placebo-controlled trials:
“Those who still believe in the superiority of homeopathy over placebo in clinical trials should try to replicate one of the promising positive results reported above and book a good therapist for the time after the trial.” – Walach and Jonas (2002:240)
The authors seem to imply that such trials are doomed, and any researcher prepared to attempt it should expect to need assistance for recuperation in the aftermath! An explanation for the irreproducibility of RCTs is offered within the model of complex systems below.
A different paradigm: Whole Systems Thinking
This brief synopsis is derived from Capra, F. (1996) The Web of Life14 :
The science of systems, or systems thinking, emerged in the early 20th century, in the field of biology, and was taken up by Gestalt psychology, ecology and physics. In a major paradigm shift from the mechanistic world view of Galileo, Descartes and Newton, it is holistic in its concern with how parts of systems function in relation to each other and to the whole. In the reductionist mechanistic view, understanding the whole is derived from a break-down of its components; in systems thinking, parts do not exist separately or independently of their interactions and relationships within the whole system. Properties of the whole emerge which are not present at the level of its components, and if the whole is dissected into its parts, those properties will disappear. In biology, a system can exist in a single cell, an organ, an organism, a community, and an ecosystem. Understanding the properties of any of these is contingent upon viewing the networking of processes within and between the systems. In short, the whole is more than the sum of its parts.
This paradigm matches well with the holistic world-view of homeopathy which attends multiple manifestations of disease, and considers symptoms as local expressions of a global disturbance3. Therefore, research questions should correspond. For example, a systems-relevant question would be ‘Does the intervention affect network function?’ (ibid) It follows logically that testing the efficacy of SAHDs as isolated agents in a complex therapeutic intervention is of little value, as long as the other components of the system are disregarded.
An understanding of the nature of complex systems sheds light on the irreproducibility of reductionist research on a whole system: systems function with marked sensitivity to small perturbations within and between its parts. Therefore it is unlikely that conditions are precisely the same in a repeat experiment, or that a precise outcome in one point of space and time will replicate. (ibid)
Whole Systems Research in Homeopathy
The critique that follows was influenced by criteria set out by Greenhalgh (1997)15 on how to assess the methodological quality of published papers, and will examine the study’s originality, background theory, design validity, data validity and results, from a WSR perspective.
The original methodological aspects of Thompson’s study are:
1. Addresses the hypothesis that homeopathy is a complex intervention where different active ingredients may act synergistically as well as individually.
Previous research has focused on efficacy of the remedy as the active agent, ignoring the context of treatment. Real-world validity is high.
2. Chooses the prospective Formal Case Study design.
The FCS is a more rigorous approach which enables a higher level of research validity, by formulating an objective structure of analysis10. This enhances knowledge derived from the study and sets a model for future research.
3. Expands data sources to incorporate a range of qualitative textual data which, when triangulated with quantitative numeric data from outcome measurements, provide a firm basis of reliability.
This is an innovative approach, which integrates theory and knowledge from social science disciplines (psychology, psychotherapy, anthropology) with recognised methods of evaluations within homeopathy and general medicine, so that a more holistic outcome evaluation can be achieved. Qualitative data in the form of patient’s artwork and reports from ‘significant other’ are particularly novel inclusions.
Rather than attempting to fit the RCT model to real clinical practice conditions, as in the study by Weatherley-Jones et al (2004), the natural setting of the clinic is the basis for Thompson’s exploration. In all other respects, including duration of treatment, patients were offered Bristol Homeopathic Hospital standard package of care. FCSs should be used to investigate phenomena within the context of real-life, especially when a phenomenon and context are not clearly separable from each other. ‘Formality’ indicates a replicable systematic approach, superior to anecdotal case-reporting10 .
Though the research was not specifically designed to prove the effectiveness of homeopathy, the degree of success of each case outcome was integral to the analysis of the ‘active ingredients’.
The term ‘active ingredients’ is creatively borrowed from a pharmaceutical perspective, where an ‘ingredient’ has a direct causal effect on the treatment outcome. ‘Ingredient’ can be defined as “a constituent element of anything” or “something that enters as an element into a mixture” 16. A reductionist perspective would test elements in isolation17, whereas WSR perspective examines patterns, interactions and relationships that emerge as elements of a system operating together3. This study design included a sufficiently broad range of qualitative and quantitative data to enable conclusions to be drawn (see below).
Choices were made about recruitment and inclusion for this study. In real-life practice, patients attracted to homeopathy suffer from a wide range of diagnosed illnesses. For this study, three conditions were selected: irritable bowel syndrome (IBS), chronic fatigue syndrome (CFS) and childhood atopic dermatitis (AD) with a degree of severity. The choices are significant and interesting, in that both IBS and CFS are chronic, systemic, debilitating conditions, considered to be linked to psychological factors, and not easily treatable with conventional medicine, since no ’cause’ is recognized18. AD is a common childhood disease of genetic origin and associated with multiple allergies, conventionally treated by avoiding irritants and application of topical steroids (ibid). These conditions fit the definition of ‘chronic’ as persistent and recurring diseases (ibid). According to homeopathic principles, these disorders would be categorized as functional, as opposed to lesional, where severe structural changes are present19. The selection of these conditions for study adds relevance and weight. To demonstrate effectiveness of homeopathy with this range of conditions could signify a considerable contribution to the health of the population.
Thompson has drawn upon theoretical constructs from models of healing from social science disciplines. Perhaps this is because there is no formal theory of homeopathic case-taking, though it is recognised that an individual’s psychological symptoms are often the most individualising. What is beyond the scope of Thompson’s report is an explanation of why these particular concepts are relevant to this study. What is the overlap between psychology and homeopathy, and what, in theory, is unique to homeopathy? Is it even necessary to investigate deep ‘life-world’ experiences, in order to treat homeopathically? An in-depth interview, or case-taking, is the method of transmitting the patient’s unique inner state to the practitioner, and guides the remedy choice. Methods of patient interview vary widely between practitioners, just as patients can describe their symptoms at different levels of detail. This can be observed in homeopathic journals, where some case reports emphasise a list of physical and/or mental symptoms to support the remedy choice, whereas others include deep life-story analyses20,21. Veterinary homeopaths must select a medicine without engaging the patient in any verbal interaction.
One obvious drawback is the small sample size (n=18). However, the comprehensive and intensive nature of the data was not used to prove the effectiveness of homeopathy by statistical significance but to enable reasoned judgements about changes in the health status of its participants, and to evaluate the relative importance of components of the system. Data from a range of qualitative and quantitative sources was triangulated. Triangulation is defined as the use of a variety of data sources or methods to examine a specific phenomenon to produce a more accurate account of the phenomenon under investigation22. As there is no control/non-treatment group, a substantial number of data sources is needed to validate statements.
Quantitative data was gathered from numeric outcome scales, both generic and specific to the individual complaint, and supplied data on patients’ improvement in quality of life, as well as symptom severity. All are recognized as accurate and reliable within general medicine. (see Appendix I)
Qualitative data in the form of exit interview, patient artwork and ‘significant other’ reports enabled expressions of global outcome, i.e. how much better the patient felt. According to WSR, health is an emergent property of a person as a complex living system and cannot be established by fixed end-point measurements alone23.
Thompson was able to identify various ‘ingredients’ of the homeopathic intervention as well as comment on their relative importance, regardless of whether the individual case treatment was successful or not. As reported, one third of cases experienced major global health improvement, one third some, and one third none. This enabled relative evaluations. For example, patient expectation was uniform, so therefore did not correlate with actual improvement.
In order to identify these ‘ingredients’, data from patient interviews was coded separately and blindly by both investigators, and analysed by computer, according to pre-existing and emerging themes. By doing so, a number of key issues were addressed regarding challenges to homeopathy, i.e. that success in homeopathy is due to placebo effect (Shang et al 2005) or the empathic consultation24.
Non-specific ‘ingredients’ would refer to components acting as beneficial, independently of the SAHD given:
- Patient expectation of benefit
This did not correlate with outcome in every case. Patients who did not improve held the same expectations as those who did.
- Patient openness to the mind-body connection
Generally this did appear to correspond to good outcome in most cases. This supports the idea that there are ‘alternative patients’ as well as alternative therapies25.
- Consultational empathy
This appeared to be necessary for good outcome, but with no direct correlation between empathy level and outcome.
The theory of disclosure maintains that by disclosing trauma, the person is enabled to gain self-insight and repair damaged self-perception26. In this study, only one case is cited in which this may have mediated marked health gains.
Results led to the conclusion that deep probing in consultation was not necessary for therapeutic benefit. In two of the cases of major global improvement, no deep ‘life-world’ was revealed.
The two ‘ingredients’ specific to homeopathy are intrinsically linked:
‘Homeopathicity’ is a process whereby therapist matches patient with the most suitable SAHD (remedy). Accuracy of match between patient and remedy is the goal of the consultation and is dependent on practitioner skill. Clarity of this match as correlated with major health changes was analysed and determined retrospectively. All cases with no improvement were categorised as unclear match. Twice the number of improved cases were clear-match than unclear. Interestingly, the homeopathic skill of practitioner was not measured in any other way.
- The remedy
Various phenomena suggested evidence of the remedy as specifically active. One patient receiving more than one remedy could accurately document which was most effective, through both narrative and measurement profiles. This phenomenon contradicts the notion that homeopathic medicines act as placebo.
Other research designs that would contribute to this topic
Having identified the ‘active ingredients’ in a prospective case-series study, retrospective observational studies or large scale observational studies could focus on patient openness to the mind-body connection, patient expectations and practitioner role.
RCTs incorporating a holistic model are also possible, as Frei et al have shown, in a trial of homeopathic treatment of ADHD in which the optimal medication for the patient (child) was determined before the placebo-controlled, cross-over trial began27.
Medical Anthropology may provide interesting models for qualitative studies to investigate the mechanism and meaning of healing in homeopathy.
The paradigm of whole systems research enables homeopathy as a therapeutic intervention to be investigated within the real-life setting of the clinic. Thompson and Weiss have conducted an innovative, ground-breaking and rigorously designed study which goes beyond the limitations of RCTs . They have identified a number of therapeutic components which can form the basis for further investigations, either through observational retrospective or further prospective case-series studies. Many more pragmatic trials are needed to overturn the view that homeopathy is a non-specific placebo effect.
1. Walach, H., Jonas, W.B. (2002a) Ch.14 Homeopathy pp. 229-245, in George Lewith, Wayne B. Jonas and Harald Walach (2002) Clinical Research in Complementary Therapies London: Churchill Livingstone
2. Shang, A., Huwiler-Muntener, K., Nartey, L., Juni, P., Dorig,S., Sterne, J.A., Powsner, D., Egger, M. (2005) Are the clinical effects of homeopathy placebo effects? Comparative study of placebo-controlled trials of homeopathy and allopathy Lancet 366 pp.726-732
3. Bell, I.R., and Kothian, M. (2006) Models for the Study of Whole Systems Integrative Cancer Therapies 5(4) pp.293-307
4. Weatherley-Jones, E., Thompson, E.A., Thomas, K.J. (2004) The placebo-controlled trial as a test of complementary and alternative medicine: observations from research experience of individualised homeopathic treatment Homeopathy 93, pp.186-189
5. Campbell, D., and Fitzgerald, M. (2007) Royals’ favoured hospital at risk as homeopathy backlash gathers pace The Observer Sunday April 8, 2007 located via Internet http://observer.guardian.co.uk/uk_news/story/0,,2052505,00.html)
6. Verhoef, M.J., Casebeer, A.L., Hilsden, R.J. (2002) Assessing Efficacy of Complementary Medicine: Adding Qualitative Research Methods to the “Gold Standard” The Journal of Alternative and Complementary Medicine Vol. 8, number 3, pp.275-281
7. Mason, S. Tovey, P Long, A.F. (2002) Evaluating Complementary Medicine: methodological challenges of randomised controlled trials. BMJ 2002; 325 832-834
8. Verhoef, M.J., Lewith, G., Ritenbaugh, C., Boon, H., Fleishman, S., Leis, A. (2005) Complementary and alternative medicine whole systems research: Beyond identification of inadequacies of the RCT Complementary Therapies in Medicine vol.13 pp. 206-212
9. Thompson, T.D.B., and Weiss, M., (2006) Homeopathy – what are the active ingredients? An exploratory study using the UK Medical Research Council’s framework for the evaluation of complex interventions BMC Complementary and Alternative Medicine vol. 6:37
10. Thompson, T. (2004) Can the caged bird sing? Reflections on the application of qualitative research methods to case study design in homeopathic medicine BMC Medical Research Methodology 4:4
11. Jonas, W.B., Lewith, G., Walach, H. (2002) Ch.1: Balanced Research Strategies for complementary and alternative medicine, pp. 3-27, in George Lewith, Wayne B. Jonas and Harald Walach (2002) Clinical Research in Complementary Therapies London: Churchill Livingstone
12. Walach, H., Jonas, W.B., Lewith, G. (2002b) Ch.2: The role of outcomes research in evaluating complementary and alternative medicine pp. 29-45, in George Lewith, Wayne B. Jonas and Harald Walach (2002) Clinical Research in Complementary Therapies London: Churchill Livingstone
14. Capra, F. (1996) The Web of Life Great Britain, Harper Collins
15. Greenhalgh, T. (1997) How to read a paper: Assessing the methodological quality of published papers BMJ 1997;315:305-308 (2 August)
17. Bar-Yam, Y. (2000) Concepts in Complex Systems, from the website of New England Complex Science Institute http://necsi.org/guide/concepts/reductionism.html
18. Macpherson, G. Ed. (2002) Black’s Medical Dictionary 40th edition A&C Black, London
19. Eizayaga, F. (1991) Treatise on Homoeopathic Medicine Buenos Aires Ediciones Marecel
20. Datta, A.K., (2003) Chronic diarrhea cured with Natrum Sulphuricum Homeopathic Links vol.16 Winter, pp. 215
21. Gujjar, B. (2003) My child will become a criminal Homeopathic Links vol.16 Winter, pp. 210-211
23. Bell, I.R., Caspi, O., Schwartz, G.E.R., Grant, K.L., Gaudet, T.W., Rychener, D., Maizes, V., Weil, A. (2002) Integrative Medicine and Systemic Outcomes Research Arch Intern Med vol.162 pp.133-140
24. Bikker, A.P., Mercer, S.W., Reilly, D. (2005) Empathy is crucial for enablement: A Pilot Prospective Study on the Consultation and Relational Empathy, Patient Enablement, and Health Changes over 12 Months in Patients Going to the Glasgow Homoeopathic Hospital The Journal of Alternative and Complementary Medicine Vol. 11, No. 4 pp.591-600
25. Caspi, O., Koithan, M., Kriddle, M. (2004) Alternative Medicine or “Alternative Patients”: A Qualitative Study of Patient-Oriented Decision-Making Processes With Respect to Alternative and Complementary Medicine Medical Decision Making MO-MO 2004
26. Pennebaker JW, Susman JR. (1988) Disclosure of traumas and psychosomatic processes. Social Science Medicine 26(3):327-32. (abstract)
27. Frei, H., Everts, R., von Ammon, K., Kaufmann, F., Walther, D., Hsu Schmitz, SF., Collenberg, M., Steinlin, M., Lim, C., Thurneysen, A. (2007) Randomised controlled trials of homeopathy in hyperactive children: treatment procedure leads to an unconventional study design; Experience with open-label homeopathic treatment preceding the Swiss ADHD placebo controlled, randomised, double-blind, cross-over trial Homeopathy 96 pp. 35-41
APPENDIX I Outcome measurements scales references
Consultation and Relational Empathy Measure: The Consultation and Relational Empathy Measure has been developed as a tool for assessing the patients’ perceptions of relational empathy in the consultation. Relevance and practical use of the Consultation and Relational Empathy (CARE) Measure in general practice. Mercer SW, McConnachie A, Maxwell M, Heaney D, Watt GC. Fam Pract. 2005 Jun;22(3):328-34. Epub 2005 Mar 16.
Childrens Dermatological Life Quality Index: developed in 1995 by research dermatologists The Children’s Dermatology Life Quality Index aims to measure the impact of skin disease on children’s quality of life. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=7662573&dopt=Abstract
The Fatigue index scale was initially developed in 1994 to improve understanding of the effects of fatigue on quality of life. FIS items reflect perceived impact on cognitive, physical, and psychosocial functioning. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=8148458&dopt=Citation
Glasgow Homeopathic Hospital Outcome Scale; specifically developed for homeopathy treatment at GHH http://www.adhom.com
Validity and reliability demonstrated in a report published in 1997 http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&list_uids=9218081&dopt=Abstract
BMJ 1996;312:1016-1020 (20 April) MYMOP Measure Your Medical Outcome Profile has been developed as a patient generated generic instrument to evaluate one’s health status, and is sensitive to improvement or deterioration, in both specific symptoms and global quality of life assessment