Content uploaded by Jens Behnke
Author content
All content in this area was uploaded by Jens Behnke on Feb 28, 2017
Content may be subject to copyright.
The Current State
of Homeopathic
Research
• Healthcare research
• Randomised controlled clinical trials
• Meta-Analyses
• Basic research
Wissenschaftliche Gesellschaft für Homöopathie
Scientific Society for Homeopathy
With contributions from
Klaus von Ammon
Stephan Baumgartner
Jens Behnke
Martin Frei-Erb
Curt Kösters
Michael Teut
Loredana Torchetti
May 2016
Imprint
Publisher:
Scientic Society for Homeopathy (WissHom)
Wallstraße 48
06366 Köthen (Anhalt)
Germany
www.wisshom.de
Editing, design and layout:
Mag. phil. Christine Doppler
lektorat@wisshom.de
www.christine-doppler.net
© WissHom. Köthen (Anhalt). 2016
Acknowledgements
We wish to thank the Homeopathy Foundation of the German Central Association of Homeopathic Doctors (DZVhÄ) for
its nancial and administrative support. Without such kind support this report on the current status of homeopathy could not
have been written – at least not at this time.
www.homoeopathie-stiftung.de
Neither the Homeopathy Foundation of the German Central Association of Homeopathic Doctors (DZVhÄ) nor the
German Central Association of Homeopathic Doctors (DZVhÄ) itself has had any inuence on the content of this report.
Contents
2 Imprint
3 Contents
5 Summary
7 Michael Teut
Clinical outcome research in homeopathy
13 Klaus von Ammon, Loredana Torchetti, Martin Frei-Erb
Results of original RCTs with individual homeopathy and
high potencies versus placebo and standard treatments
25 Klaus von Ammon, Curt Kösters
Methodological problems of randomised double-blind trials
in homeopathic research
33 Jens Behnke
Meta-analyses in homeopathic clinical research
41 Stephan Baumgartner
Status of basic research in homeopathy
49 Authors
The current state of
homeopathic research
This report on the current state of homeopathic research provides a summary on the research areas
of healthcare research, randomised controlled clinical trials, meta-analyses and basic research.
It aims to contribute to the discussion within the field of homeopathy concerning the need for research,
the relevance of individual research fields and methods, and their role in future research strategies.
We are also publishing this report on the current status of scientific research, moreover, for the benefit
of medical science as a whole and the public.
Whilst the conventional development of medicinal products is based on research which must then
stand up to medical practice, homeopathy is primarily a successful medical practice that must stand up
to scientific research.
Outcome studies investigating homeopathic treatment under routine conditions have reported
clinically relevant improvements in symptoms and quality of life, often comparable with those under
conventional treatments, but with fewer adverse effects. In half of all health economic evaluations
homeopathic treatment showed less costs. Methodologically, a causal relationship between drug
therapy and therapeutic outcome cannot be deduced from those studies.
The randomised controlled clinical trials investigated here, studies with good methodology into
individualised homeopathy with high potencies only, have indicated, like earlier studies, that classical
homeopathy is superior to placebo, and that remedies in high potencies have a specific effect. A
definitive scientific statement cannot be made at present given the heterogeneous nature of the data
and the small number of studies of good quality.
A review of the meta-analyses of homeopathy reveals results which mostly are statistically significant
compared to placebo, suggesting specific efficacy from potentised remedies. Depending on the
selection criteria applied, different studies are thereby included in the analysis. The majority of the
studies reported in all the examined reviews (also Shang et al.), including those with good
methodology, suggest that homeopathic treatment is superior to placebo. These findings are in part
markedly qualified by the authors of the respective meta-analyses in their (comment/discussion and)
conclusion. The stated caveats, e.g. with respect to study quality, thereby do not always reflect the
usual scientific standards, or they actually refer explicitly to the postulated implausibility of the efficacy
of high-potency medicines.
There are a number of high-quality basic research studies that report specific effects also for high
potencies, inclusive also of independently replicated experimental models. There are initial empirical
results pointing to the physicochemical and pharmaceutical, as well as biological ways homeopathic
remedies work, but no theory is fully developed yet.
A summary analysis of the clinical research data offers sufficient evidence of the therapeutic
effectiveness of homeopathic treatment.
The results from numerous placebo-controlled trials and basic research experiments suggest,
moreover, that potentised medicines offer a specific efficacy.
6
Put in perspective, there are many important open research areas – notably:
• Basic research into the optimisation of laboratory models and the understanding of the mode
of action
• Independent replications of studies in clinical and basic research
• Exploration of the provision of homeopathic care in reality, also combined with conventional
medicine
• Health economic analyses to evaluate the costs and benefits (cost effectiveness)
Board of the Scientific Society for Homeopathy (WissHom)
Köthen (Anhalt), May 2016
Teut M. Clinical outcome research in homeopathy. 2016.
Clinical outcome research in homeopathy
Michael Teut
_________________________________________________________________________________
Background
Although randomised clinical trials (RCTs) are regarded as the “gold standard” in clinical research,
they do have drawbacks: As a rule, they are conducted at selected research centres by selected
physicians, with a selection of patients who in most cases have been enlisted.
Often, however, routine practice reveals many years later that the medicinal products are given to
other patients in another context and with different concomitant conditions and medications to those in
the registration trials, resulting in different outcomes and sometimes surprising adverse effects. The
results from RCTs thus can only be generalized to the clinical routine to a limited degree.
For this reason, additional clinical studies that examine the effectiveness and safety of medicines
under everyday routine care condtions and with everyday patients are important and increasingly
demanded by health authorities. Such clinical studies, which are performed in everyday settings, are
referred to as “clinical outcome research”.
In outcome studies homeopathy is usually studied as an integrated and complex therapeutic concept
(e.g. consisting of a specific consultation, examination and medication). Prospective observational or
cohort studies (without control group) describe medical care as it actually takes place in clinical reality.
However, uncontrolled studies do not permit to assess about the efficacy of a specific intervention as
the result may be influenced by other factors (e.g. the patients social status, income, lifestyle or other,
concurrent therapies).
Prospective observational or cohort studies including a control group (e.g. homeopathy versus
conventional treatment) permit to compare therapeutic alternatives in the routine care settings.
However, the outcomes may be influenced by other relevant features or characteristics than the
interventions in different patient groups (e.g. a more healthy lifestyle of the group undergoing
homeopathic treatment, a different social, economic or educational status).
It is possible to statistically adjust for differences in the patient groups to a certain extent. If optimal
comparability is desired, however, it is better to make use of randomisation technique: If patients are
randomly assigned to treatment groups that reflect the routine setting (randomised, pragmatic study),
efficacy can be judged most reliably.
Also health economic questions can be answered by outcome research. A cost analysis addresses
only the costs of treatment, but not the outcome. A cost-cost analysis compares the costs of two
alternative treatments – again, without taking into account benefits of the treatment. A cost-benefit
analysis examines the benefit in monetary units; a cost-effectiveness analysis examines costs related
to natural parameters (e.g. years of life gained). A cost-utility analysis examines costs in relation to
benefits, reported as utility value, mostly in “quality-adjusted life years” (QALYs).
8
Teut M. Clinical outcome research in homeopathy. 2016.
Results
To date homeopathic outcome research has focussed largely on therapy administered by physicians.
Most of the available clinical studies were performed in Europe and India.
The largest homeopathic cohort study to be conducted so far in German speaking countries included
3981 patients in the clinics of more than 100 homeopathic physicians in Germany and Switzerland
over a period of eight years (1, 2). The most common indications for homeopathic treatment were
long-term, chronic diseases: headaches and migraines in women; allergic rhinitis and hypertension in
men; atopic eczema and recurrent infections in children. The pre-/post-treatment-comparisons of
outcomes showed a reduction in the intensity of symptoms on average by almost 50%, and a marked
improvement in quality of life. The intensity of the symptoms was evaluated by the physicians and
patients on symptom scales (Numeric Rating Scale); quality of life was assessed using the SF-36
questionnaire. The decrease in symptoms was greatest within the first three months of treatment. The
noted improvements in the symptoms and quality of life were long-lasting. In follow-up questionnaires
the patients still reported lasting improvements eight years after starting treatment (2). Clinically
relevant improvements were seen in this study for groups with the following diagnoses: Migraine and
headache (3, 4), chronic rhinitis (5), back pain (6), geriatric patients (7), menstrual pain (8) and
psoriasis (9). As this study was designed without a control group, the question arises whether the
improvements reflect a therapeutic effect or the spontaneous course of the disease conditions. In this
regard the term “regression toward the mean” is used to describe the shift of extreme values (e.g.
when a disease is active) over time towards the mean value. A statistical analysis of the quality of life
data from this large cohort study did not suggest a sole “regression toward the mean” effect (10).
An earlier prospective comparative observational study in the 1990s (11) already compared outcomes
of patients on homeopathic treatment to those on conventional treatment in routine care. 465 patients
with diseases of the upper or lower respiratory tracts and ears (including allergies) were included, 281
received homeopathic and 175 conventional treatment. After 14 days, 82.6% of the patients on
homeopathic and 68% of those on conventional treatment were free of symptoms or greatly improved;
67.3% of the patients on homeopathic and 56.6% of the patients on conventional treatment improved
within three days. Adverse effects occurred in 22.3% of the patients on conventional treatment, but
only in 7.8% of the patients on homeopathic treatment. In 2007, the authors published the data from
an even larger multinational prospective obeservational study including 1577 patients with acute
respiratory and ear disorders: 857 patients received homeopathic treatment, and 720 patients
conventional therapy. The improvements were similar in both groups after seven days, but recovery
was more rapid in the homeopathic group (12).
In another comparative, prospective observational study (13), the outcomes of patients insured by a
German health insurance fund receiving homeopathic treatment were compared to those on
conventional medicine alone. A total of 459 patients took part. There was a more marked reduction of
symptom severity (recorded by patients and physicians) in the homeopathy group compared to the
conventional treatment group; in addition the costs were lower under homeopathic treatment.
In another prospective observational study including 1097 Norwegian patients given homeopathic
treatment seven out of ten patients reported a clear and clinically relevant symptoms reduction during
a six-month course of homeopathic treatment (14).
A prospective observational study conducted at a British hospital outpatient department in a
population of 6544 patients revealed a clinical improvement in 70.7% of those under homeopathic
treatment, and a good or very good therapeutic outcome in half of all patients (15).
9
Teut M. Clinical outcome research in homeopathy. 2016.
A prospective observational study at an Italian hospital also reported at least a moderate clinical
improvement in 74% of all included patients (16).
An observational study in 772 children conducted in six European countries and Brazil (17) reported a
high level of treatment satisfaction and an increase in quality of life after two months of homeopathic
treatment; adverse effects were observed in only 4.2% of the children.
Several observational studies are also available for diagnosis-related groups of patients:
• Comparative prospective observational studies into the routine treatment of children with atopic
eczema have been conducted in Germany (18, 19, 20). Similarly good outcomes from both
homeopathic and conventional treatment in terms of the skin and quality of life were reported,
although homeopathic treatment was more expensive.
• A comparative, prospective observational study in cancer patients with different diagnoses in
Germany compared conventional treatment plus homeopathy (259 patients) against
conventional therapy alone (380 patients) (21). Quality of life was found to be better in the group
receiving homeopathic treatment. Further studies in oncology revealed a high level of treatment
satisfaction in the case of complementary homeopathic treatment (22) and a decrease in
symptoms of oestrogen withdrawal in female patients with breast cancer under complementary
homeopathic treatment (23). In a recent randomised pragmatic trial in Austria cancer patients
were randomly assigned to receive either conventional therapy or conventional therapy plus
homeopathic treatment (24). For the combination group a significantly better quality of life and far
greater well-being was reported.
• A small comparative prospective observational study in patients with diabetic neuropathy
revealed an improvement in clinical symptoms of patients on homeopathic treatment (n=45)
compared to those on conventional therapy (n=32) (25). A prospective observational study in
India including 336 patients likewise demonstrated a clinical improvement of neuropathy under
homeopathic treatment within 12 months (26).
• The French EPI3-MSD cohort study reported that patients with musculoskeletal symptoms
receiving homeopathic treatment from their general practitioner had similar outcomes to patients
on conventional treatment, but that approx. 50% fewer were taking nonsteroidal antirheumatic
drugs (27).
• Clinically relevant improvements were reported by other homeopathic prospective observational
studies, amongst others, headaches (28, 29, 4), otitis media (30), male infertility (31), acne (32),
chronic sinusitis (33,5), Chickungunya disease (34), menopausal symptoms (35), asthma (36),
allergies (37) and injuries (38).
Health economic studies investigating homeopathy naturally reflect the costs within the national
healthcare system. A systematic review performed in 2014 (39) summarises the results from 14 health
economic analyses of homeopathy covering more than 3500 patients; ten studies included a control
group. In eight out of 14 studies improvements in health and cost savings were reported; in four
studies the outcomes corresponded to those of the conventional control, with equivalent costs.
Two studies found that the outcomes were comparable to conventional therapy, albeit with higher
costs. One of the first studies into the costs of medical homeopathy under the integrated care
contracts (Integrierte Versorgung) of the statutory health insurance funds in Germany revealed that
the costs of additional homeopathic treatment were higher than those of patients receiving only
conventional therapy, but yet without assessing the outcomes. Hence, the ratio of cost to outcome is
not yet clear (40).
10
Teut M. Clinical outcome research in homeopathy. 2016.
Discussion
On the whole, outcome studies document relatively uniform results: In patients undergoing
homeopathic treatment in routine care, relevant clinical improvements are observed, if compared to
conventional care with often similar outcomes, but with fewer adverse effects and, in half of all health
economic studies, lower costs.
It is important to understand that uncontrolled observational studies cannot examine whether high-
potency homeopathic remedies are more effective than placebos. Outcomes from clinical routine are
reported. Such outcomes have to be investigated with care and can be influenced by a number of
factors – drug effects, suggestive effects, expectations, therapeutic consultations, regression toward
the mean, and concomitant (non-homeopathic) therapies, amongst others.
The strengths of outcome studies, however, are that they allow to report on the outcomes of
homeopathy in everyday conditions, portraying a good picture of the experiences gained by
homeopathic patients and therapists in the daily routine.
Conclusion
Under routine conditions, clinically relevant improvements in symptoms and quality of life are reported
by the majority of homeopathic outcome studies. From a pragmatic perspective it can be concluded
that patients experience a clinical benefit. Methodologically, however, a causal relationship (efficacy)
between drug therapy and outcomes cannot be investigated in these studies.
References
1. Witt CM, Lüdtke R, Baur R, Willich S: Homeopathic medical practice: Long-Term Results of a cohort study
with 3981 patients. BMC Public Health 2005, 5:115 http://www.biomedcentral.com/1471-2458/5/115
2. Witt CM, Lüdtke R, Mengler N, Willich SN. How healthy are chronically ill patients after eight years of
homeopathic treatment? – Results from a long term observational study. BMC Public Health 2008; 8:413
http://www.biomedcentral.com/1471-2458/8/413
3. Witt CM, Lüdtke R, Willich SN. Homeopathic treatment of patients with migraine – a prospective
observational study with a two years follow-up. J Alt Complement Med 2010; 16(4): 347-355
http://www.ncbi.nlm.nih.gov/pubmed/20423206
4. Witt CM, Lüdtke R, Willich SN. Homeopathic treatment of patients with chronic headache – a prospective
observational study with a follow-up over two years. Forsch Komplementmed 2009; 16:227-235
http://www.ncbi.nlm.nih.gov/pubmed/19729933
5. Witt CM, Lüdtke R, Willich SN. Homeopathic treatment of patients with chronic sinusitis: a prospective
observational study with 8 years follow-up. BMC Ear Nose Throat Disord 2009; 9:7
http://www.ncbi.nlm.nih.gov/pubmed/19635154
6. Witt CM, Lüdtke R, Baur R, Willich SN. Homeopathic treatment of patients with chronic low back pain – a
prospective observational study with 2 years follow-up. Clin J Pain 2009; 25:334-339
http://www.ncbi.nlm.nih.gov/pubmed/19590483
7. Teut M, Lüdtke R, Schnabel K, Willich SN, Witt CM: Homeopathic treatment of elderly patients – a
prospective observational study with follow-up over a two year period. BMC Geriatrics 2010; 10:10
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2841176/
8. Witt CM, Lüdtke R, Willich SN. Homeopathic treatment of patients with dysmenorrhea – a prospective
observational study with 2 years follow-up. Arch Gynecol Obstet 2009; 280:603-611.
http://www.ncbi.nlm.nih.gov/pubmed/19229544
9. Witt CM, Lüdtke R, Willich SN. Homeopathic treatment of patients with psoriasis – a prospective
observational study with 2 years follow-up. J Eur Acad Dermatol Venereol 2009; 23:538-543
http://www.ncbi.nlm.nih.gov/pubmed/19192019
11
Teut M. Clinical outcome research in homeopathy. 2016.
10. Lüdtke R, Willich SN, Ostermann T. Are the effects of homeopathy attributable to a statistical artefact? A
reanalysis of an observational study. Evid Based Complement Alternat Med. 2013; 2013:612-890.
http://www.ncbi.nlm.nih.gov/pubmed/24396390
11. Riley D, Fischer M, Singh B, Haidvogl M, Heger M. Homeopathy and conventional medicine: an outcomes
study comparing effectiveness in a primary care setting. J Altern Complement Med. 2001 Apr; 7(2):149-59.
http://www.ncbi.nlm.nih.gov/pubmed/11327521
12. Haidvogl M, Riley DS, Heger M, Brien S, Jong M, Fischer M, Lewith GT, Jansen G, Thurneysen AE.
Homeopathic and conventional treatment for acute respiratory and ear complaints: a comparative study on
outcome in the primary care setting. BMC Complement Altern Med. 2007 Mar 2; 7:7.
http://www.biomedcentral.com/1472-6882/7/7
13. Witt C, Keil T, Selim D, Roll S, Vance W, Wegscheider K, Willich SN. Outcome and costs of homoeopathic
and conventional treatment strategies: a comparative cohort study in patients with chronic disorders.
Complement Ther Med. 2005 Jun; 13(2):79-86. http://www.ncbi.nlm.nih.gov/pubmed/16036164
14. Steinsbekk A, Lüdtke R. Patients' assessments of the effectiveness of homeopathic care in Norway: a
prospective observational multicentre outcome study. Homeopathy. 2005 Jan; 94(1):10-6.
http://www.ncbi.nlm.nih.gov/pubmed/15751329
15. Spence DS, Thompson EA, Barron SJ. Homeopathic treatment for chronic disease: a 6-year, university-
hospital outpatient observational study. J Altern Complement Med. 2005 Oct; 11(5):793-8.
http://www.ncbi.nlm.nih.gov/pubmed/16296912
16. Rossi E, Endrizzi C, Panozzo MA, Bianchi A, Da Frè M. Homeopathy in the public health system: a seven-
year observational study at Lucca Hospital (Italy). Homeopathy. 2009 Jul; 98(3):142-8.
http://www.ncbi.nlm.nih.gov/pubmed/19647207
17. Van Wassenhoven M, Goossens M, Anelli M, Sermeus G, Kupers P, Morgado C, Martin E, Bezerra M.
Pediatric homeopathy: a prospective observational survey based on parent proxy-reports of their children's
health-related Quality of Life in six European countries and Brazil. Homeopathy. 2014 Oct; 103(4):257-63.)
http://www.ncbi.nlm.nih.gov/pubmed/25439042
18. Keil T, Witt CM, Roll S, Vance W, Weber K, Wegscheider K, Willich SN. Homoeopathic versus conventional
treatment of children with eczema: a comparative cohort study. Complement Ther Med. 2008 Feb; 16(1):15-
21. http://www.ncbi.nlm.nih.gov/pubmed/18346624
19. Witt CM, Brinkhaus B, Pach D, Reinhold T, Wruck K, Roll S, Jäckel T, Staab D, Wegscheider K, Willich SN.
Homoeopathic versus conventional therapy for atopic eczema in children: medical and economic results.
Dermatology. 2009; 219(4):329-40. http://www.ncbi.nlm.nih.gov/pubmed/19828937
20. Roll S, Reinhold T, Pach D, Brinkhaus B, Icke K, Staab D, Jäckel T, Wegscheider K, Willich SN, Witt CM.
Comparative effectiveness of homoeopathic vs. conventional therapy in usual care of atopic eczema in
children: long-term medical and economic outcomes. PLoS One. 2013; 8(1):e54973.
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0054973
21. Rostock M, Naumann J, Guethlin C, Guenther L, Bartsch HH, Walach H. Classical homeopathy in the
treatment of cancer patients – a prospective observational study of two independent cohorts. BMC Cancer.
2011 Jan 17; 11:19. http://www.biomedcentral.com/1471-2407/11/19
22. Thompson EA, Reillly D. The homeopathic approach to symptom control in the cancer patient: a prospective
observational study. Palliat Med. 2002 May; 16(3):227-33. http://www.ncbi.nlm.nih.gov/pubmed/12046999
23. Thompson EA, Reilly D. The homeopathic approach to the treatment of symptoms of oestrogen withdrawal in
breast cancer patients. A prospective observational study. Homeopathy. 2003 Jul; 92(3):131-4.
http://www.ncbi.nlm.nih.gov/pubmed/12884894
24. Frass M, Friehs H, Thallinger C, Sohal NK, Marosi C, Muchitsch I, Gaertner K, Gleiss A, Schuster E,
Oberbaum M. Influence of adjunctive classical homeopathy on global health status and subjective wellbeing
in cancer patients – A pragmatic randomized controlled trial. Complement Ther Med. 2015 Jun; 23(3):309-17.
http://www.ncbi.nlm.nih.gov/pubmed/26051564
25. Pomposelli R, Piasere V, Andreoni C, Costini G, Tonini E, Spalluzzi A, Rossi D, Quarenghi C, Zanolin ME,
Bellavite P. Observational study of homeopathic and conventional therapies in patients with diabetic
polyneuropathy. Homeopathy. 2009 Jan; 98(1):17-25. http://www.ncbi.nlm.nih.gov/pubmed/19135955
26. Nayak C, Oberai P, Varanasi R, Baig H, Ch R, Reddy GR, Devi P, S B, Singh V, Singh VP, Singh H,
Shitanshu SS. A prospective multi-centric open clinical trial of homeopathy in diabetic distal symmetric
polyneuropathy. Homeopathy. 2013 Apr; 102(2):130-8. http://www.ncbi.nlm.nih.gov/pubmed/23622263
27. Rossignol M, Begaud B, Engel P, Avouac B, Lert F, Rouillon F, Bénichou J, Massol J, Duru G, Magnier AM,
Guillemot D, Grimaldi-Bensouda L, Abenhaim L; EPI3-LA-SER group. Impact of physician preferences for
homeopathic or conventional medicines on patients with musculoskeletal disorders: results from the EPI3-
MSD cohort. Pharmacoepidemiol Drug Saf. 2012 Oct; 21(10):1093-101.
http://www.ncbi.nlm.nih.gov/pubmed/22782803
12
Teut M. Clinical outcome research in homeopathy. 2016.
28. Walach H, Lowes T, Mussbach D, Schamell U, Springer W, Stritzl G, Haag G. The long-term effects of
homeopathic treatment of chronic headaches: one year follow-up and single case time series analysis. Br
Homeopath J. 2001Apr; 90(2):63-72. http://www.ncbi.nlm.nih.gov/pubmed/11341459
29. Danno K, Colas A, Masson JL, Bordet MF. Homeopathic treatment of migraine in children: results of a
prospective, multicenter, observational study. J Altern Complement Med. 2013 Feb; 19(2):119-23.
http://www.ncbi.nlm.nih.gov/pubmed/22978244
30. Otitis media - Friese KH, Kruse S, Lüdtke R, Moeller H. The homoeopathic treatment of otitis media in
children – comparisons with conventional therapy. Int J Clin Pharmacol Ther. 1997 Jul; 35(7):296-301.
http://www.ncbi.nlm.nih.gov/pubmed/9247843
31. Gerhard I, Wallis E. Individualized homeopathic therapy for male infertility. Homeopathy. 2002 Jul; 91(3):133-
44. http://www.ncbi.nlm.nih.gov/pubmed/12322866
32. Miglani A, Manchanda RK. Observational study of Arctium lappa in the treatment of acne vulgaris.
Homeopathy. 2014 Jul; 103(3):203-7.) http://www.ncbi.nlm.nih.gov/pubmed/24931753
33. Nayak C, Singh V, Singh VP, Oberai P, Roja V, Shitanshu SS, Sinha MN, Deewan D, Lakhera BC,
Ramteke S, Kaushik S, Sarkar S, Mandal NR, Mohanan PG, Singh JR, Biswas S, Mathew G. Homeopathy in
chronic sinusitis: a prospective multi-centric observational study. Homeopathy. 2012 Apr; 101(2):84-91.
http://www.ncbi.nlm.nih.gov/pubmed/22487367
34. Wadhwani GG. Homeopathic drug therapy. Homeopathy in Chikungunya Fever and Post-Chikungunya
Chronic Arthritis: an observational study. Homeopathy. 2013 Jul; 102(3):193-8.
http://www.ncbi.nlm.nih.gov/pubmed/23870379
35. Nayak C, Singh V, Singh K, Singh H, Gupta J, Lamba CD, Sharma A, Sharma B, Indira B, Bhuvaneshwari S,
Bindra SK, Luxmi KS. Management of distress during climacteric years by homeopathic therapy. J Altern
Complement Med. 2011 Nov; 17(11):1037-42. http://www.ncbi.nlm.nih.gov/pubmed/22087613
36. Shafei HF, AbdelDayem SM, Mohamed NH. Individualized homeopathy in a group of Egyptian asthmatic
children. Homeopathy. 2012 Oct; 101(4):224-30. http://www.ncbi.nlm.nih.gov/pubmed/23089218
37. Gründling C, Schimetta W, Frass M. Real-life effect of classical homeopathy in the treatment of allergies: A
multicenter prospective observational study. Wien Klin Wochenschr. 2012 Jan; 124(1-2):11-7.
http://www.ncbi.nlm.nih.gov/pubmed/22138796
38. Schneider C, Schneider B, Hanisch J, van Haselen R. The role of a homoeopathic preparation compared
with conventional therapy in the treatment of injuries: an observational cohort study. Complement Ther Med.
2008 Feb; 16(1):22-7. http://www.ncbi.nlm.nih.gov/pubmed/18346625
39. Viksveen P, Dymitr Z, Simoens S. Economic evaluations of homeopathy: a review. Eur J Health Econ. 2014
Mar; 15(2):157-74. http://www.ncbi.nlm.nih.gov/pubmed/23397477
40. Ostermann JK, Reinhold T, Witt CM. Can Additional Homeopathic Treatment Save Costs? A Retrospective
Cost-Analysis Based on 44500 Insured Persons. PLoS One. 2015 Jul 31; 10(7)
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0134657
von Ammon K, Torchetti L, Frei-Erb M. Results of original RCTs with individual homeopathy
and high potencies versus placebo and standard treatments. 2016.
Results of original RCTs with individual
homeopathy and high potencies versus
placebo and standard treatments
Klaus von Ammon, Loredana Torchetti, Martin Frei-Erb
_________________________________________________________________________________
Purpose
To compile an overview of the specific effects of individually prescribed ultramolecular medicines
(classical homeopathy with high potencies ≥C12) compared with placebo or standard therapy when
used in patients requiring treatment.
Method
A search for literature in published meta-analyses and the “Medline” database (search criteria:
“homeopath*” AND “RCT” AND publication date 01/01/2012-31/12/2014) resulted in high-quality
(minimum criterion: Jadad1 score ≥3), peer-reviewed original publications on randomised controlled
trials (RCTs) into the effect of individually prescribed ultramolecular medicines (potency ≥C12) which
were collated for the purpose of a descriptive assessment. Books or book contributions, dissertations
or conference articles were excluded from this analysis, as were pilot studies, studies that were not
double-blind, studies with a dropout rate of >20%, studies with laboratory or surrogate parameters,
and prevention studies. The original studies were used, if possible, or the data extracted from meta-
analyses.
Result
Based on the publications by Linde 19972, Linde and Melchart 19983, Clausius 19984 and Shang
20055, RCTs of good methodological quality fulfilling the minimum criterion of Jadad score ≥3 were
identified. In addition, all peer-reviewed RCTs into classical homeopathy could be selected from
Mathie 20136 and 20147, and an updated Medline search (period 2012-2014) was performed.
Altogether, 71 RCTs (without duplicate entries) were identified from the period 1982 to 2014. Of these,
a total of 62 trials were ruled out (for details, see Table 1, page 20): 26 due to the use of molecular
substances (potency < C12), 14 pilot studies, 11 single-blind studies, five due to laboratory/surrogate
parameters, and six for other reasons (two due to a lack of peer review, two due to the preventive use
of homeopathy, and two due to a high dropout rate).
A total of 24 to 126 patients took part in the remaining nine trials (Table 2, page 21), with a mean
number of 64 (SD=31, median 64). The clinical spectrum comprised three studies of diarrhoea, and
one study for each of the other diagnoses. In four studies, homeopathy was used as an add-on to
standard treatment (Frass 2005I3, Jacobs 1993I5, Jacobs1994I6, Jacobs 2000I7). In all nine studies,
homeopathy was compared against placebo. The clinical endpoint of the studies varied between five
days and 1.5 years. The intention-to-treat (ITT) analysis is a standard.
14
von Ammon K, Torchetti L, Frei-Erb M. Results of original RCTs with individual homeopathy
and high potencies versus placebo and standard treatments. 2016.
In meta-analyses of RCTs, the study quality (risk of bias) is assessed by evaluating the internal
validitiy. For this purpose, different instruments were used in the various meta-analyses: Linde 19972,
Linde and Melchart 19983, and Clausius 19984 applied the – now outdated – Jadad score1 (maximum
5 points, cut-off for low risk of bias ≥3) and their own scale to assess internal validity (maximum 6 or 7
points, cut-off ≥4 or ≥5 points; for details see Table 3, page 23). In the case of Shang 20055, the study
quality was judged on the basis of similar criteria to Jadad (higher quality = double-blinding and
adequate randomisation). Mathie 20147 used today’s commonly applied Cochrane method8,
supplemented by differentiated classification of the bias categories. For the purposes of comparison,
we evaluated the quality of the studies by Frei 2005I4 and Lökken 1995I8 using the Cochrane criteria
and according to Mathie’s classification. Some studies were thus evaluated with different tools. The
methodological quality of most of the studies was found to be consistent in the various meta-analyses,
with the exception of Jacobs 2000I7 (Shang low, Mathie high risk of bias) and Whitmarsh 1997I9 (high
risk of bias in Mathie, and Shang, good quality in Linde and Melchart). Only the study by Jacobs
1994I6 was judged by Mathie to offer reliable evidence. In our opinion, the study by Frei 2005I4 can
likewise be deemed reliable.
On the whole, whilst the study quality is not entirely optimal, it is satisfactory. Of the nine studies, only
Frei 2005I4 revealed a low risk of bias according to the Cochrane criteria; in five studies the risk of bias
was deemed uncertain, and in three the potential for bias was rated as high. The results of these
studies must therefore be interpreted with certain reservations.
To compare the studies and subject them to meta-analysis, the results of the original studies were
converted into effect size measures, such as into odds ratios (ORs). If the accompanying 95%
confidence interval (CI) did not include 1, it was assumed that one group was superior to the other.
Of the eight RCTs with a calculated OR, six indicated that classical homeopathy was superior to
placebo, whereas the situation was reversed in two studies (Table 4, page 23). Given the small
sample sizes in all the individual studies, however, the confidence intervals were too wide to be able to
suggest that the effect of homeopathy or placebo was clearly superior. The only exception was the
study by Frass 2005I3, the result of which clearly favoured homeopathy.
Whereas no OR was given in the case of Frei 2005I4, the significance test to ascertain group
differences was applied. It indicated superiority for classical homeopathy over placebo, even if such
significance tests do not constitute measures of effect size and thus are not particularly suitable for
making a summarised assessment.
By restricting the review to the studies with the best quality (reliable evidence) according to Mathie’s
criteria omly, the OR in Jacobs 1994I6 then favoured classical homeopathy whilst the CI started
precisely at 1. Homeopathy also appeared to be superior to placebo in Frei 2005I4, even if no OR with
CI was available for making a direct comparison.
Discussion
As suggested by the majority of systematic reviews and meta-analyses performed to date, there is
some evidence that classical homeopathy is superior to placebo. Also consistent with previous meta-
analyses is the fact that the quality of the studies is not satisfactory right through. By applying very
strict parameters (low or uncertain risk of bias, as per Mathie 20147), the number of usable studies is
reduced to six, with the remaining three revealing a high risk of bias in accordance with Mathie 20147.
15
von Ammon K, Torchetti L, Frei-Erb M. Results of original RCTs with individual homeopathy
and high potencies versus placebo and standard treatments. 2016.
Given that these results were obtained from relatively few studies and small sample sizes, the
following conclusions must be interpreted with caution (cf. Mathie 20147).
The criterion of this review is the individual application of high potencies, which were used for different
diagnoses with varying outcome measures and endpoints. In this respect the available data are
inhomogeneous, allowing only for descriptive analysis.
It is surprising that all the studies were placebo-controlled, there were no comparative studies with
standard therapy only, i.e. standard therapy without placebo, and that in a large number (4 out of 9) of
studies homeopathy was used as an add-on. This reflects the experimental approach of most of the
studies and invokes the question of transfer and applicability into daily practice. The broadness of the
diagnostic spectrum, covering acute and chronic, mild (Lökken 1995I8) to severe (Frass 2005I3)
conditions, is just as surprising. Apart from the three studies by JacobsI5-I7 into childhood diarrhoea,
none of these studies has as yet been replicated.
To achieve more definitive results, future clinical studies should be conducted prospectively with larger
sample sizes and methods that closely reflect the clinical routine.
This review did not consider any non-peer-reviewed studies into classical homeopathy, nor any
studies that employed non-individualized methods of homeopathy – clinical, complex homeopathy,
complex (integrative) treatments and isopathy – which are covered by the pool of experience in
applied homeopathy.
Summary
In the last 25 years, a number of good-quality studies into individual homeopathy and ultramolecular
medicines have been published subsequent to peer review in high-level, Medline-indexed journals.
The majority (7 out of 9) indicate that these medicines have a specific effect, even if the sample size of
the individual studies was generally too small to demonstrate, on its own, that they were clearly
superior.
A similar overview should be produced for non-peer-reviewed studies (n=15 in Mathie 20136) and non-
individually prescribed homeopathic treatment (n=279 in Mathie 20136).
References
1) Jadad AR, Moore RA, Carroll D, et al. Assessing the quality of reports of randomized clinical trials: is blinding
necessary? Control Clin Trials. 1996; 17:1–12.
2) Linde K, Clausius N, Ramirez G, Melchart D, Eitel F, Hedges LV, Jonas WB. Are the clinical effects of
homeopathy placebo effects? A meta-analysis of placebo-controlled trials. Lancet. 1997 Sep;
350(9081):834–43.
3) Linde K, Melchart D. Randomized controlled trials of individualized homeopathy: a state-of-the-art review. J
Altern Complement Med. 1998 Winter; 4(4):371–88.
4) Clausius N. Kontrollierte klinische Studien zur Homöopathie. Eine systematische Übersichtsarbeit mit
Metaanalyse. Essen: Karl und Veronica Carstens-Stiftung [Controlled clinical studies into homeopathy. A
systematic review with meta-analysis. Essen: Karl and Veronica Carstens Foundation; in German]; 1998.
5) Shang A, Huwiler-Müntener K, Nartey L, Jüni P, Dörig S, Sterne JA, Pewsner D, Egger M. Are the clinical
effects of homeopathy placebo effects? Comparative study of placebo-controlled trials of homeopathy and
allopathy. Lancet. 2005 Aug 27–Sep 2; 366(9487):726–32.
16
von Ammon K, Torchetti L, Frei-Erb M. Results of original RCTs with individual homeopathy
and high potencies versus placebo and standard treatments. 2016.
6) Mathie RT, Hacke D, Clausen J, Nicolai T, Riley DS, Fisher P. Randomised controlled trials of homeopathy in
humans: characterising the research journal literature for systematic review. Homeopathy. 2013 Jan;
102(1):3–24. doi: 10.1016/j.homp.2012.10.002.
7) Mathie RT, Lloyd SM, Legg LA, Clausen J, Moss S, Davidson JR, Ford I. Randomised placebo-controlled
trials of individualised homeopathic treatment: systematic review and meta-analysis. Syst Rev. 2014 Dec 6;
3:142. doi: 10.1186/2046-4053-3-142.
8) Higgins JP, Altman DG, Gøtzsche PC, Jüni P, Moher D, Oxman AD, Savovic J, Schulz KF, Weeks L, Sterne
JA; Cochrane Bias Methods Group; Cochrane Statistical Methods Group. The Cochrane Collaboration's tool
for assessing risk of bias in randomised trials. BMJ. 2011 Oct 18; 343:d5928. doi: 10.1136/bmj.d5928.
References – original studies
a) Included original studies (I)
I1) Bonne O, Shemer Y, Gorali Y, Katz M, Shalev AY. A randomized, double-blind, placebo-controlled study of
classical homeopathy in generalized anxiety disorder. J Clin Psychiatr. 2003; 64:282–287.
I2) Chapman EH, Weintraub RJ, Milburn MA, Pirozzi TO, Woo E. Homeopathic treatment of mild traumatic brain
injury: a randomized, double-blind, placebo-controlled clinical trial. J Head Trauma Rehabil. 1999; 14:521–
542.
I3) Frass M, Linkesch M, Banyai S, Resch G, Dielacher C, Löbl T, Endler C, Haidvogl M, Muchitsch I, Schuster
E. et al. Adjunctive homeopathic treatment in patients with severe sepsis: a randomized, double-blind,
placebo-controlled trial in an intensive care unit. Homeopathy. 2005; Apr; 94:75–80.
I4) Frei H, Everts R, von Ammon K, et al. Homeopathic treatment of children with attention deficit hyperactivity
disorder: a randomized, double blind, placebo controlled crossover trial. Eur J Pediatr. 2005; 164:758–767.
I5) Jacobs J, Jimenez LM, Gloyds SS, Casares FE, Gaitan MP, Crothers D. Homoeopathic treatment of acute
childhood diarrhoea. A randomized clinical trial in Nicaragua. Br Homoeopath J. 1993; 82:83–86.
I6) Jacobs J, Jimenez LM, Gloyds SS, Gale JL, Crothers D. Treatment of acute childhood diarrhea with
homeopathic medicine; a randomized clinical trial in Nicaragua. Pediatrics. 1994; 93:719–725.
I7) Jacobs J, Jimenez LM, Malthouse S, et al. Homeopathic treatment of acute childhood diarrhoea: results from
a clinical trial in Nepal. J Altern Complement Med. 2000; 6:131–139.
I8) Lökken P, Straumsheim PA, Tveiten D, Skjelbred P, Borchgrevink CF. Effect of homoeopathy on pain and
other events after acute trauma: placebo controlled trial with bilateral oral surgery. BMJ. 1995; 310:1439–42.
I9) Whitmarsh TE, Coleston-Shields DM, Steiner TJ. Double-blind randomized placebo-controlled study of
homoeopathic prophylaxis of migraine. Cephalalgia. 1997; 17:600–604.
b) Excluded original studies (E)
E1) Adler UC, Paiva NM, Cesar AT, Adler MS, Molina A, Padula AE, Calil HM. Homeopathic Individualized Q-
Potencies versus Fluoxetine for Moderate to Severe Depression: Double-Blind, Randomized Non-Inferiority
Trial. Evid Based Complement Alternat Med. 2011; 2011:520182. doi: 10.1093/ecam/nep114. Epub 2011
Jun 8.
E2) Adler UC, Krüger S, Teut M, Lüdtke R, Schützler L, Martins F, Willich SN, Linde K, Witt CM. Homeopathy
for depression: A randomized, partially double-blind, placebo-controlled, four-armed study (DEP-HOM).
PLoS One. 2013 Sep 23; 8(9):e74537. doi: 10.1371/journal.pone.0074537. eCollection 2013.
E3) Andrade L, Ferraz MB, Atra E, Castro A, Silva MSM. A randomized controlled trial to evaluate the
effectiveness of homoeopathy in rheumatoid arthritis. Scand J Rheumatol. 1991; 20:204–208.
E4) Awdry R. Homeopathy and chronic fatigue – The search for proof. Int J Altern Complement Med.
1996;14:19–22.
E5) Bell IR, Lewis DA 2nd, Brooks AJ, Schwartz GE, Lewis SE, Caspi O, Cunningham V, Baldwin CM.
Individual differences in response to randomly assigned active individualized homeopathic and placebo
treatment in fibromyalgia: implications of a double-blinded optional crossover design. J Altern Complement
Med. 2004; 10:269–283.
17
von Ammon K, Torchetti L, Frei-Erb M. Results of original RCTs with individual homeopathy
and high potencies versus placebo and standard treatments. 2016.
E6) Bell IR, Lewis DA 2nd, Schwartz GE, Lewis SE, Caspi O, Scott A, Brooks AJ, Baldwin CM.
Electroencephalographic cordance patterns distinguish exceptional clinical responders with fibromyalgia to
individualized homeopathic medicines. J Altern Complement Med. 2004; 10:285–299.
E7) Bell IR, Lewis DA 2nd, Lewis SE, Schwartz GE, Brooks AJ, Scott A, Baldwin CM. EEG alpha sensitization
in individualized homeopathic treatment of fibromyalgia. Int J Neurosci. 2004; 114:1195–1220.
E8) Bell I, Lewis D, Brooks A, Schwartz G, Lewis S, Walsh B, Baldwin C. Improved clinical status in
fibromyalgia patients treated with individualized homeopathic remedies versus placebo. Rheumatology.
2004; 43:577–582.
E9) Brien S, Lachance L, Prescott P, McDermott C, Lewith G. Homeopathy has clinical benefits in rheumatoid
arthritis patients that are attributable to the consultation process but not the homeopathic remedy: a
randomized controlled clinical trial. Rheumatology (Oxford). 2011; 50:1070–1082.
E10) Brigo B. Homeopathic treatment of migraine: A sixty case, double-blind, controlled study (homeopathic
remedy vs. placebo). Proceedings of the 42nd LMHI Congress. Arlington, VA, March 29–April 2, 1987,
pp318–329. (see also Brigo B, Serpellini G. Homeopathic treatment of migraines: A randomized doubleblind
study of sixty cases (homoeopathic remedy versus placebo). Berl J Res Hom. 1991;1:98–106.) →
JADAD=2 bei Clausius.
E11) Carlini EA, Braz S, Troncone LRP, Tufik S, Romanach AK, Pustiglione M, Sposati MC, Cudizio Filho O,
Prado MIA. Efeito hipnótico de medicação homeopática e do placebo. Avaliação pela técnica de duplocego
e cruzamento [Hypnotic effect of homeopathic medication and placebo. Evaluation by double-blind and
crossover techniques]. Rev Assoc Méd Bras. 1987; 33:83–88.
E12) Cavalcanti AM, Rocha LM, Carillo R Jr, Lima LU, Lugon JR. Effects of homeopathic treatment on pruritus of
haemodialysis patients: a randomized placebo-controlled double-blind trial. Homeopathy. 2003; 92:177–
181.
E13) Chakraborty PS, Lamba CD, Nayak D, John MD, Sarkar DB, Poddar A, Arya JS, Raju K, Vivekanand K,
Singh HBK, Baig H, Prusty AK, Singh V, Nayak C. Effect of individualized homoeopathic treatment in
influenza like illness: A multicenter, single blind, randomized, placebo controlled study. Indian J Res
Homoeopathy. 2013; 7:22–30.
E14) Chakraborty PS, Varanasi R, Majumdar AK, Banoth K, Prasad S, Ghosh MS, Sinha MN, Reddy GRC,
Singh V, Nayak, C. Effect of homoeopathic LM potencies in acute attacks of haemorrhoidal disease: A
multicentric randomized single-blind placebo-controlled trial. Indian J Res Homoeopathy. 2013; 7:72–80.
E15) Chand KS, Manchanda RK, Mittal R, Batra S, Banavaliker JN, De I. Homeopathic treatment in addition to
standard care in multi drug resistant pulmonary tuberculosis: a randomized, double blind, placebo controlled
clinical trial. Homeopathy. 2014 Apr; 103(2):97–107. doi: 10.1016/j.homp.2013.12.003.
E16) Chapman EH, Angelica J, Spitalny G, Strauss M. Results of a study of the homoeopathic treatment of PMS.
J Am Inst Hom. 1994; 87:14–21.
E17) Chauhan VK, Manchanda RK, Narang A, Marwaha RK, Arora S, Nagpal L, Verma SK, Sreenivas V.
Efficacy of homeopathic intervention in subclinical hypothyroidism with or without autoimmune thyroiditis in
children: an exploratory randomized control study. Homeopathy. 2014 Oct; 103(4):224–31. doi:
10.1016/j.homp.2014.08.004. Epub 2014 Sep 27.
E18) de Lange de Klerk ESM, Blommers J, Kuik DJ, Bezemer PD, Feenstra L. Effects of homoeopathic
medicines on daily burden of symptoms in children with recurrent upper respiratory tract infections. Br Med
J. 1994; 309:1329–1332.
E19) Fisher P. An experimental double-blind trial method in homeopathy. Use of a limited range of remedies to
treat fibrositis. Br Homeopath J. 1986; 75:142–147.
E20) Fisher P, Greenwood A, Huskisson EC, Turner P, Belon P. Effect of homeopathic treatment on fibrositis
(primary fibromylgia). Br Med J. 1989; 299:365–366.
E21) Fisher P, Scott DL. A randomized controlled trial of homeopathy in rheumatoid arthritis. Rheumatology.
2001; 40:1052–1055.
E22) Fisher P, McCarney R, Hasford C, Vickers A. Evaluation of specific and non-specific effects in homeopathy:
Feasibility study for a randomised trial. Homeopathy. 2006; 95:215–222.
E23) Gaucher C, Jeulin D, Peycru P, Amengual C. A double blind randomized placebo controlled study of
cholera treatment with highly diluted and succussed solutions. Br Homoeopath J. 1994; 83:132–134.
E24) Gerhard I, Monga B, Roebruck P, Runnebaum B. Homoeopathy versus conventional therapy in female
infertility: interim analysis of a randomized study. Forsch Komplementarmed. 1997; 5:262–269.
E25) Gmünder R, Kissling R. The efficacy of homeopathy in the treatment of chronic low back pain compared to
standardized physiotherapy. Z Orthop Grenzgeb. 2002; 140:503–508.
18
von Ammon K, Torchetti L, Frei-Erb M. Results of original RCTs with individual homeopathy
and high potencies versus placebo and standard treatments. 2016.
E26) Haila S, Koskinen A, Tenovuo J. Effects of homeopathic treatment on salivary flow rate and subjective
symptoms in patients with oral dryness: a randomized trial. Homeopathy. 2005; 94:175–181.
E27) Harrison H, Fixsen A, Vickers A. A randomized comparison of homoeopathic and standard care for the
treatment of glue ear in children. Complement Ther Med. 1999; 7:132–135.
E28) Hitzenberger G, Korn A, Dorcsi M, Bauer P, Wohlzogen FX. Kontrollierte randomisierte doppelblinde Studie
zum Vergleich einer Behandlung von Patienten mit essentieller Hypertonie mit homöopathischen und
pharmakologisch wirksamen Medikamenten [Controlled, randomised, double-blind study to compare
homeopathic versus pharmacological treatment in patients with essential hypertension]. Wien Klin
Wochenschr. 1982; 94:665–670.
E29) Jacobs J, Springer DA, Crothers D. Homeopathic treatment of acute otitis media in children: a preliminary
randomized placebo–controlled trial. Pediatr Infect Dis J. 2001; 20:177–183.
E30) Jacobs J, Herman P, Heron K, Olsen S, Vaughters L. Homeopathy for menopausal symptoms in breast
cancer survivors: a preliminary randomized controlled trial. J Altern Complement Med. 2005b; 11:21–27.
E31) Jacobs J, Williams A–L, Girard C, Njike VY, Katz D. Homeopathy for attention-deficit/hyperactivity disorder:
a pilot randomized-controlled trial. J Altern Complement Med. 2005a; 11:799–806.
E32) Jansen GRHJ, van der Veer ALJ, Hagenaars J, van der Juy A. Lessons learnt from an unsuccessful clinical
trial of homoeopathy. Results of a small-scale, double-blind trial in proctocolitis. Br Homoeopath J. 1992;
81:132–138.
E33) Kainz JT, Kozel G, Haidvogl M, Smolle J. Homoeopathic versus placebo therapy of children with warts on
the hands: a randomized, double-blind clinical trial. Dermatology. 1996; 193:318–320.
E34) Katz T, Fisher P, Katz A, Davidson J, Feder G. The feasibility of a randomised, placebo-controlled clinical
trial of homeopathic treatment of depression in general practice. Homeopathy. 2005; 94:145–52.
E35) Kuzeff RM. Homeopathy, sensation of well-being and CD4-levels – A placebo-controlled, randomized trial.
Complement Ther Med. 1998; 6:4–9.
E36) Mourão LC, Moutinho H, Canabarro A. Additional benefits of homeopathy in the treatment of chronic
periodontitis: a randomized clinical trial. Complement Ther Clin Pract. 2013 Nov; 19(4):246–50. doi:
10.1016/j.ctcp.2013.05.002. Epub 2013 Jul 27.
E37) Mourão LC, Cataldo DM, Moutinho H, Fischer RG, Canabarro A. Additional effects of homeopathy on
chronic periodontitis: a 1-year follow-up randomized clinical trial. Complement Ther Clin Pract. 2014 Aug;
20(3):141–6. doi: 10.1016/j.ctcp.2014.03.003. Epub 2014 Apr 8.
E38) Mousavi F, Mojaver YN, Asadzadeh M, Mirzazadeh M. Homeopathic treatment of minor aphthous ulcer: a
randomized, placebo-controlled clinical trial. Homeopathy. 2009; 98:137–141.
E39) Naudé DF, Couchman IMS, Maharaj A. Chronic primary insomnia: efficacy of homeopathic simillimum.
Homeopathy. 2010; 99:63–68. [Published erratum: Homeopathy. 2010; 99:151]
E40) Oberai P, Gopinadhan S, Varanasi R, Mishra A, Singh V, Nayak C. Homoeopathic management of attention
deficit hyperactivity disorder: A randomised placebo-controlled pilot trial. Indian J Res Homoeopathy. 2013;
7:158–167.
E41) Paterson C, Ewings P, Brazier JE, Britten N. Treating dyspepsia with acupuncture and homeopathy:
reflections on a pilot study by researchers, practitioners and participants. Complement Ther Med. 2003;
11:78–84.
E42) Peckham EJ, Relton C, Raw J, Walters C, Thomas K, Smith C, Kapur K, Said E. Interim results of a
randomised controlled trial of homeopathic treatment for irritable bowel syndrome. Homeopathy. 2014 Jul;
103(3):172–7. doi: 10.1016/j.homp.2014.05.001. Epub 2014 May 29.
E43) Rastogi DP, Singh VP, Singh V, Dey SK, Rao K. Homeopathy in HIV infection: a trial report of double-blind
placebo controlled study. Br Homoeopath J. 1999; 88:49–57.
E44) Relton C, Smith C, Raw J, et al. Healthcare provided by a homeopath as an adjunct to usual care for
fibromyalgia (FMS): results of a pilot randomised controlled trial. Homeopathy. 2009; 98:77–82.
E45) Relton C, O'Cathain A, Nicholl J. A pilot 'cohort multiple randomised controlled trial' of treatment by a
homeopath for women with menopausal hot flushes. Contemp Clin Trials. 2012 Sep; 33(5):853–9. doi:
10.1016/j.cct.2012.04.003. Epub 2012 Apr 20.
E46) Saha S, Koley M, Hossain SI, Mundle M, Ghosh S, Nag G, Datta AK, Rath P. Individualized homoeopathy
versus placebo in essential hypertension: A double-blind randomized controlled trial. Indian J Res
Homoeopathy. 2013; 7:62–71.
E47) Sajedi F, Alizad V, Alaeddini F, Fatemi R, Mazaherinezhad A. The effect of adding homeopathic treatment
to rehabilitation on muscle tone of children with spastic cerebral palsy. Complement Ther Clin Pract. 2008;
14:33–37.
19
von Ammon K, Torchetti L, Frei-Erb M. Results of original RCTs with individual homeopathy
and high potencies versus placebo and standard treatments. 2016.
E48) Schreier T, Hartmann M, Petzoldt D, et al. Homöopathie versus konventionelle Therapie bei männlicher
Unfruchtbarkeit: Zwischenbericht einer randomisierten Studie [Homeopathy versus conventional therapy for
male infertility: Interim report from a randomised study]. Forsch Komplementarmed. 1997; 4:325–331.
E49) Schwab G. Lässt sich eine Wirkung homöopathischer Hochpotenzen nachweisen? [Can an effect be
demonstrated with high-potency homeopathic remedies?] Karlsruhe: Deutsche Homöopathische Union;
1990.
E50) Siebenwirth J, Lüdtke R, Remy W, Rakoski J, Borelli S, Ring J. Wirksamkeit einer klassisch-
homöopathischen Therapie bei atopischem Ekzem. Eine randomisierte, placebokontrollierte
Doppelblindstudie [Effectiveness of classical homeopathic treatment in atopic eczema. A randomised
placebo-controlled double-blind clinical trial]. Forsch Komplementärmed. 2009; 16:315–323.
E51) Sinha MN, Siddiqui VA, Nayak C, Singh V, Dixit R, Dewan D, Mishra A. Randomized controlled pilot study
to compare Homeopathy and Conventional therapy in Acute Otitis Media. Homeopathy. 2012 Jan;
101(1):5–12. doi: 10.1016/j.homp.2011.08.003.
E52) Steinsbekk A, Bentzen N, Fønnebø V, Lewith G. Self treatment with one of three self selected,
ultramolecular homeopathic medicines for the prevention of upper respiratory tract infections in children. A
double-blind randomized placebo controlled trial. Br J Clin Pharmacol. 2005; 59:447–455.
E53) Steinsbekk A, Fønnebø V, Lewith G, Bentzen N. Homeopathic care for the prevention of upper respiratory
tract infections in children: a pragmatic, randomized, controlled trial comparing randomized homeopathic
care and waiting-list controls. Complement Ther Med. 2005; 13:231–238.
E54) Straumsheim P, Borchgrevink C, Mowinckel P, Kierulf H, Hafslund O. Homeopathic treatment of migraine: a
double blind, placebo controlled trial of 68 patients. Br Homeopath J. 2000; 89:4–7.
E55) Thompson EA, Montgomery A, Douglas D, Reilly D. A pilot, randomized, double-blinded, placebo-controlled
trial of individualized homeopathy for symptoms of estrogen withdrawal in breast-cancer survivors. J Altern
Complement Med. 2005; 11:13–20.
E56) Thompson EA, Shaw A, Nichol J, et al. The feasibility of a pragmatic randomised controlled trial to compare
usual care with usual care plus individualised homeopathy, in children requiring secondary care for asthma.
Homeopathy. 2011; 100:122–130.
E57) van Erp VMA, Brands M. Homoeopathic treatment of malaria in Ghana: open study and clinical trial. Br
Homoeopath J. 1996; 85:66–70.
E58) Walach H, Häusler W, Lowes T, Mussbach D, Schamell U, Springer W, Stritzl G, Haag G. Classical
homeopathic treatment of chronic headaches. Cephalalgia. 1997; 17:119–126.
E59) Weatherley-Jones E, Nicholl JP, Thomas KJ, Parry GJ, McKendrick MW, Green ST, Stanley PJ, Lynch SP.
A randomized, controlled, triple-blind trial of the efficacy of homeopathic treatment for chronic fatigue
syndrome. J Psychosom Res. 2004; 56:189–197.
E60) White A, Slade P, Hunt C, Hart A, Ernst E. Individualised homeopathy as an adjunct in the treatment of
childhood asthma: a randomised placebo controlled trial. Thorax. 2003; 58:317–321.
E61) Witt A, Kaufmann U, Bitschnau M, et al. Monthly itraconazole versus classic homeopathy for the treatment
of recurrent vulvovaginal candidiasis: a randomised trial. BJOG: Int J Obstet Gynaecol. 2009; 116:1499–
1505.
E62) Yakir M, Kreitler S, Brzezinski A, Vithoulkas G, Oberbaum M, Bentwich Z. Effects of homeopathic treatment
in women with premenstrual syndrome: a pilot study. Br Homeopath J. 2001; 90:148–153.
20
von Ammon K, Torchetti L, Frei-Erb M. Results of original RCTs with individual homeopathy
and high potencies versus placebo and standard treatments. 2016.
Table 1 Sources and primary exclusion criteria of RCTs not included in the overview
No.
Authors
Year
Sources
Primary reason for exclusion
Linde
1997
Clausius
1998
Linde &
Melchart
1998
Shang
2005
Mathie
2013
Mathie
2014
Medline
search
(2012-2014)
Potency
<C12
Pilot
study
Single-
blind
Lab/
surrogate
parameters Other
E1 Adler 2011 x
x
E2 Adler 2013 x x
x
E3 Andrade 1991 x x x x
x
E4 Awdry 1996 x
x1
E5 Bell 2004
x
x
E6 Bell 2004 x
x
E7 Bell 2004 x
x
E8 Bell 2004 x
x
E9 Brien 2011 x
x
E10 Brigo 1991 x
x1
E11 Carlini 1987 x x
x
E12 Cavalcanti 2003 x
x
E13 Chakraborty 2013 x
x
E14 Chakraborty 2013 x
x
E15 Chand 2014 x
x
E16 Chapman 1994 x x x x
x
E17 Chauhan 2014 x
x
E18 de Lange 1994 x x x x
x
E19 Fisher 1986 x
x
E20 Fisher 1989 x x
x
E21 Fisher 2001 x
x
E22 Fisher 2006 x
x
E23 Gaucher 1994 x
x
E24 Gerhard 1997 x
x
E25 Gmünder 2002 x
x
E26 Haila 2005 x
x
E27 Harrison 1999 x
x
E28 Hitzenberger 1982 x x
x
E29 Jacobs 2001 x x
x
E30 Jacobs 2005 x
x
E31 Jacobs 2005 x
x
E32 Jansen 1992 x
x
E33 Kainz 1996 x x
x
E34 Katz 2005 x
x
E35 Kuzeff 1998 x x
x
E36 Mouräo 2013
x
x
E37 Mouräo 2014 x
x
E38 Mousavi 2009 x
x
E39 Naude 2010 x
x
E40 Oberai 2013 x
x
E41 Paterson 2003 x
x
E42 Peckham 2014 x
x
21
von Ammon K, Torchetti L, Frei-Erb M. Results of original RCTs with individual homeopathy
and high potencies versus placebo and standard treatments. 2016.
E43 Rastogi 1999 x
x
E44 Relton 2009 x
x
E45 Relton 2012 x
x
E46 Saha 2013 x
x
E47 Sajedi 2008 x
x
E48 Schreier 1997 x
x
E49 Schwab 1990 x x x
x
E50 Siebenwirth 2009
x
x
E51 Sinha 2012 x
x
E52 Steinsbekk 2005 x
x2
E53 Steinsbekk 2005 x
x2
E54 Straumsheim 2000 x x
x
E55 Thompson 2005 x
x
E56 Thompson 2011 x
x
E57 van Erp 1996 x
x3
E58 Walach 1997 x x x
x
E59
Weatherley
-
Jones 2004 x
x
E60 White 2003 x
x
E61 Witt 2009 x
x3
E62 Yakir 2001 x
x
Remarks
1 No peer review
2 Prevention
3 High dropout rate: E57 van Erp 1996: 26%, E61 Witt 2009: 53%
Table 2 Description of included trials
No.
Authorship Year Diagnosis n ITT n PP Duration Study group Control group
I1 Bonne 2003 Anxiety disorder 44 39 10 weeks HOM PLA
I2 Chapman 1999 Traumatic brain injury 61 50 4 months HOM PLA
I3 Frass 2005 Sepsis 70 67 180 days Standard + HOM Standard + PLA
I4 Frei 2005 ADHD 62 57 180 days HOM PLA
I5 Jacobs 1993 Acute diarrhoea in children 34 33 ≤ 6 days Standard + HOM Standard + PLA
I6 Jacobs 1994 Acute diarrhoea in children 92 81 ≤ 5 days Standard + HOM Standard + PLA
I7 Jacobs 2000 Acute diarrhoea in children 126 116 5 days Standard + HOM Standard + PLA
I8 Lökken 1995 Removal of wisdom teeth 24 24 5 days HOM PLA
I9 Whitmarsh 1997 Migraine 63 60 4 months HOM PLA
Remarks
ITT = intention-to-treat analysis; PP = per protocol analysis; ADHD = attention deficit hyperactivity disorder;
HOM = homeopathy; PLA = placebo; standard = standard treatment
22
von Ammon K, Torchetti L, Frei-Erb M. Results of original RCTs with individual homeopathy
and high potencies versus placebo and standard treatments. 2016.
Table 3 Risk of bias of the included trials
No. Authorship Year
RoB Linde
1997:
Jadad
(5; ≥3)1
RoB
Linde
1997:
IV
(7; ≥5)2
RoB
Clausius
1998:
MV
(10; ≥5)3
RoB
Linde &
Melchart
1998: Jadad
(5; ≥3)1
RoB
Linde &
Melchart
1998:
IV
(6; ≥4)4
Study
quality
Linde &
Melchart
19985
RoB
Shang
20056
RoB
Mathie
20147
RoB after
Mathie8
I1 Bonne 2003 Uncertain (B3)
I2 Chapman 1999 Low Uncertain (B1)
I3 Frass 2005 Uncertain (B2)
I4 Frei 2005 Low (A)
I5 Jacobs 1993 3 3.5 7 3 3 B High High(C1.4)
I6 Jacobs 1994 5 6 9 5 5 A Low Uncertain (B1)9
I7 Jacobs 2000 Low High (C2.0)
I8 Lökken 1995 5 6 5 5 5.5 A Uncertain (B1)10
I9 Whitmarsh 1997 4 4 A High High (C1.4)
Remarks
RoB = risk of bias; IV = internal validity; MV = model validity; MA = meta analysis.
1 Jadad 1996: maximum 5 points (cut-off ≥3) for: 1) Randomisation; 2) Adequate randomisation; 3) Double-blind;
4) Adequate blinding; 5) Reporting of dropouts and discontinuations.
2 Internal validity Linde 1997: maximum 7 points (cut-off ≥5) for: 1) Randomisation; 2) Adequate randomisation;
3) Comparability of the groups before treatment; 4) Blinding of the patients; 5) Blinding of examining
physicians; 6) Selection error after onset of treatment 7) Adequate statistical analysis.
3 Model validity Clausius 1998: maximum 10 points (cut-off ≥5): 1) Compliance with similarity principle (0-5
points); 2) Validated drug prescribing (0-1); 3) Level of trust in prescribing (0-1); 4) Medicine known to
homeopathy (0-1); 5) Unum remedium (0-1); 6) Specialist homeopathic expertise of the physician (0-1).
4 Internal validity Linde & Melchart 1998: maximum 6 points (cut-off ≥4), precise criteria not determinable:
5 Study quality Linde & Melchart 1998: A) Good methodological quality likely; B) Major deficiencies unlikely; C)
Significant minor or moderate problems; D) Not evaluable/major deficiencies.
6 Shang 2005: low risk of bias = double-blind, adequate randomisation (i.e. adequate generation of
randomisation sequence and adequate masking of randomisation).
7 Risk of bias Mathie 2014: 7 Cochrane ranges 1) Adequate generation of randomisation sequence; 2) Adequate
masking of randomisation; 3a) Blinding of participants and study personnel; 3b) Blinding of examiners; 4)
Completeness of outcome data; 5) Selective reporting; 6) Other biases.
A = low risk of bias in all seven ranges; Bx = uncertain risk of bias in x ranges, low risk of bias in all other
ranges; Cy.x = high risk of bias in y ranges, uncertain risk of bias in x ranges, low risk of bias in all other
ranges.
8 The studies whose risk of bias had not already been evaluated in the meta-analysis by Mathie 2014 were
assessed according to the Cochrane criteria and evaluated using the Mathies classification.
9 Reliable study as per Mathie 2014, i.e. at least bias category B1, and free of bias in the Cochrane ranges 1, 2,
3a and 3b.
10 Risk of bias in Lökken 1995 (as per Cochrane ranges): unclear in range 1 (i.e. unclear whether the
randomisation sequence was adequately generated), and low risk of bias in all other ranges.
23
von Ammon K, Torchetti L, Frei-Erb M. Results of original RCTs with individual homeopathy
and high potencies versus placebo and standard treatments. 2016.
Table 4 Results of included studies
No. Authorship Year
OR
Linde 1997 &
Clausius 1998
(>1 favours HOM)1
CI
Linde
19971
OR
Shang 2005 (<1
favours HOM)2
CI
Shang
20052
OR Mathie2014
(>1 favours
HOM)3
CI
Mathie
2014
Outcome of
testing for group
differences4:
I1 Bonne 2003
0.87 0.28-2.73
I2 Chapman 1999
approx. 15 incl. 1
1.986 0.72-5.49
I3 Frass 2005
3.13 1.10-8.86
I4 Frei 2005
HOM>placebo7
I5 Jacobs 1993 1.97 incl. 1
<1 incl. 1
n.i. n.i.
I6 Jacobs 1994 2.24 incl. 1?
<1 incl. 1?
2.22 1.00-4.94
I7 Jacobs 2000
<1 incl. 1?
n.i. n.i.
I8 Lökken 1995 0.72 incl. 1
I9 Whitmarsh 1997
approx. 1 incl. 1
1.72 0.69-4.34
Remarks
OR = odds ratio; CI = 95% confidence interval; HOM = homeopathy; MA = meta-analysis
1 OR and CI for Linde 1997 are depicted on a graph only; these ORs (without CIs) are reported for Clausius
1998, on the other hand,
2 Shang 2005: OR and CI are presented on a graph only.
3 n.i.= not included in the meta-analysis for Mathie 2014, as data could not be extracted.
4 In the case of the study by Frei 2005 not yet included in a meta-analysis, a report is provided on the outcome
of statistical testing for group differences.
5 Outcome: impaired activity.
6 Outcome: symptoms.
7 Frei 2005: Crossover study to compare homeopathy against placebo; the linear mixed model delivers a
significant within-person difference in favour of homeopathy (-1.67 points on the symptom-outcome scale, with
a 95% CI of -3.316 to -0.016; p=0.0479).
von Ammon K, Kösters C. Methodological problems of randomised double-blind trials
in homeopathic research. 2016.
Methodological problems of randomised
double-blind trials in homeopathic research
Klaus von Ammon, Curt Kösters
_________________________________________________________________________________
RCTs (randomised controlled trials) are aimed at examining the efficacy of medical treatments. They
entail experiments with (at least) two groups of patients or volunteers randomised to either the study
group (under treatment to be studied) or the control group. The control treatment thereby comprises
placebo or the best available therapy to date, or the study treatment is administered as an add-on to
the best proven treatment. Multi-group comparisons can therefore be performed.
The aim is to control any external confounding factors (such as the study location, methods of
administration etc.) as far as possible, or ensure that they do not differ between patients, so as to
create the same conditions for the study groups. To control the “internal” – known and unknown –
confounding factors amongst the subjects (e.g. expectations, predispositions, responsiveness),
randomisation is used: it can thus be assumed that such factors are balanced out between the groups,
or will differ only through coincidence, respectively. In such a way, the specific net effect of the
treatment to be studied can be determined.
Such a study is deemed double-blind if neither the study subjects nor the investigators know to which
study group a subject belongs. Triple-blind refers to a study design in which the rater also is unaware
of the randomisation and the nature of the different treatments.
These types of study are generally chosen when investigating new medicinal products and are
regarded as the “gold standard” – especially if the design is at least double-blind.
The RCT is surpassed at the top of the evidence pyramid (evidence-based medicine or EBM) only by
a methodological review of such RCTs (meta-analysis).
With respect to the understanding of methodological problems, it must be noted that epistemologically
RCTs are based on the paradigm of the classical natural science experiment. In such an experiment,
all known variables are kept constant and only one input variable is modified. The resulting
reproducible change in the output variables leads not only to the conclusion of a causal relationship
between the input and output variables, but at the same time also permits this relationship to be
quantified precisely.
If such an experiment cannot be fully reproduced in each individual case, it must be assumed that
further, as yet unidentified variables exist. Before the experiment is sensibly repeated, these variables
must be identified and eliminated.
The adaptation of this experimental approach to therapeutic research (and here in particular to
pharmacotherapy) has brought with it methodological limitations, however:
26
von Ammon K, Kösters C. Methodological problems of randomised double-blind trials
in homeopathic research. 2016.
• Biological organisms exhibit individual reactions – full reproducibility in an individual case
generally cannot be achieved. Such a requirement has been abandoned in favour of a statistical
statement concerning the differences in the respective group as a whole.1
• Repetition in an individual case is replaced through replication of the study as a whole.
• To keep the variables constant and to control them, randomisation is added – it should
compensate for the potential internal “confounding factors”.
• On account of the rising number, the inconsistent results and incomplete reproducibility of
studies and trials, they are summarised in meta-analyses.
What remains is the requirement that the greatest possible homogeneity be achieved within the study
groups. Resulting problems in particular, but also further methodological problems, are still a topic of
discussion, as is the value of this concept to medical practice.
In terms of medical practice, RCTs come under criticism for the following reasons, amongst others:
1. As experiments, RCTs represent an essentially different situation to that encountered in
everyday routine medical consultation.
2. The behaviour of the principal investigator could – from the design of the study alone – be
altered by his/her expectations (Rosenthal effect).
3. The individuals involved are aware of this difference and possibly behave differently
(Hawthorne effect for study subjects).
4. Different reactions from one individual to the next cannot be reflected by the statistical
statement.2
5. Any resulting statistically significant differences are no guarantee that the obtained findings are
clinically relevant.
6. The applicability and relevance to routine practice with mostly multimorbid, old, (very) young,
female, or pregnant patients, for instance, prove problematic, as the study subjects are often
young men.
7. Adverse effects or interactions with other medicinal products cannot be monitored during the
usually short period of study.
8. Long-term effects thus can only be recorded insufficiently and are systematically undervalued.
Long-term effects can look very different, and under certain circumstances also contrast with
the short-term effects.
9. Non-pharmacological treatment methods can be investigated only to a limited extent in double-
blind studies.3
1 A probability of error must thereby be taken into consideration. A probability of error of 5% (or 1% and 0.1%,
respectively) is now common.
2 Georg Ivanovas raised the eloquent question in relation to a double-blind study into the influence of the firmness
of a mattress on chronic back pain (Kovacs FM et al: Effect of firmness of mattress on chronic non-specific low-
back pain: randomised, double-blind, controlled, multicentre trial. Lancet 2003, 362: 1599–604), namely whether
all patients should sleep on a moderately firm mattress, including the 10% who experienced increased pain from
such a mattress and were more comfortable on another. Similar issues resulting from individual reactions have
also been seen with medicinal products, however.
3 This proves problematic if, when writing a guideline on depression, the evidence pertaining to pharmacological
measures must be weighed against that of psychotherapeutic interventions.
27
von Ammon K, Kösters C. Methodological problems of randomised double-blind trials
in homeopathic research. 2016.
10. Complex therapeutic strategies with individualised elements (as is common practice today in
pain therapy, for instance) cannot be investigated appropriately by means of the RCT.4
11. The applicability of this experimental approach to entire therapeutic systems has not yet been
examined systematically, let alone validated.
The classic double-blind study is designed with a specific, almost experimental study scenario in mind:
one medicinal product is tested in one indication. It is very well suited to the examination of such a
focused question (with the mentioned limitations).
If this more or less successfully implemented strategy for conventional pharmaceutical research is
adapted to homeopathic research, a number of other problems emerge which are inherent to the
research subject and possibly, in turn, will necessitate adaptation of the research tool – and limit its
validity or applicability, respectively. Some of these issues are already recognised5, but possibly have
not yet been methodologically examined to an adequate degree.
Problems arise with the application of the methodological tool of double-blind RCTs in homeopathic
research due to individualisation, due to the iterative approach to finding an appropriate agent and the
assessment of the course, and due to the different way in which therapeutic success is evaluated in
homeopathy.
Changes in symptoms can occur as a result of the natural course of the disease but also, in particular,
as a reaction to a previously administered medicine. Successive approximation solutions are thus
involved here, in the sense of an iterative approach.
The course during homeopathic treatment is evaluated largely on the basis of the change in well-
being, as well as the relevant clinical symptoms. A marked improvement in well-being, albeit without
any major improvement in the clinically relevant symptoms to begin with, is a justified reason for
waiting – because experience has shown that an improvement in clinically relevant symptoms will
follow.
Problems due to individualisation
RCTs with adequate individual remedy finding and prescribing in line with classical homeopathic
principles have seldom been conducted (see contribution by Klaus von Ammon, Loredana Torchetti,
Martin Frei-Erb: Results from original RCTs with individual homeopathy and high potencies versus
placebo and standard treatments, with excluded studies therein).
In a study with individual remedy selection (as is standard practice in classical homeopathy), the
overall outcome is inevitably poorer than the result from the individual agent with the highest degree of
probability if the different remedies available for selection do not offer the same – or at least roughly
the same – degree of probability.
The degree of probability – if initially ignoring the involved therapist and his/her level of expertise –
depends largely on the information available on the individual agents and their reliability.
4 In complex therapeutic strategies, the interplay of different treatments (e.g. pharmaceuticals, physiotherapy,
relaxation techniques, psychotherapy, acupuncture) individually adapted to suit the patient in question are
expected to deliver the therapeutic effect, rather than a single intervention.
5 Gaus W, Högel J. Studies on the efficacy of unconventional therapies. Problems and designs.
Arzneimittelforschung. 1995 Jan; 45(1):88–92.
28
von Ammon K, Kösters C. Methodological problems of randomised double-blind trials
in homeopathic research. 2016.
The information available on the individual agents differs considerably, ranging from remedies with
more than 1,000 known symptoms (originating mainly from drug proving and confirmed, moreover, by
clinical experience – if taken together a criterion for high validity in homeopathy) to those with fewer
than 10 known symptoms, which in addition result from clinical experience only and hence are of low
validity in homeopathy.
In the clinical routine, this difference in the level of knowledge about medicines and the resulting,
differing degree of probability are a much less serious problem (see above). The effect size of a study,
which precisely evaluates an intervention, is therefore lower than the effect of an individual
homeopathic treatment in clinical routine. Under the pressure of a study to offer the “optimal”
prescription spontaneously (instead of the practical approach to approximation), a negative Hawthorne
effect could in fact emerge.
Besides, therapists selecting medicines on an individualised basis cause an additional, fundamental
methodological problem. The double-blind studies of conventional medicine themselves offer only
limited reproducibility compared to classical physical experiments, because aside from the medicinal
products as the actual study variables, the patients and their individual reactions are also involved.
The therapists, as the third independent variable, additionally hamper replication – including
replication by independent examiners, in particular.
Proposed solutions:
• Limitation of the choice of agents and introduction of a standardised treatment protocol is
potentially suitable as a means of minimising the “therapist” variable. In an open pilot phase,
the problem of differing probabilities can potentially be minimised. A restriction to the
effectiveness of treatment vis-à-vis the results from everyday practice must probably be
accepted. In case of simple, acute conditions this approach would appear to be justifiable
based on experience to date; in chronic conditions the same would certainly then apply if at
the same time an exclusion criterion is introduced for patients with symptoms that do not
correspond to those of one of the approved agents – an exclusion criterion that at the same
time clearly narrows the external validity (validity of the study for average patients).
• One innovative solution is the idea of an open treatment phase followed by the actual double-
blind study with the inclusion criterion of “responder” (Bell: fibromyalgia6; Frei: ADHD 7).
However, a large proportion of the therapeutic effect already exhausted in the open treatment
phase will no longer be captured by the double-blind study. Whether this concept can be
successfully applied to other indications and other prescribing strategies also has not yet been
fully explained or examined. Yet again, external validity is limited in this case by the inclusion
criterion. In the design addressed specifically here, moreover, undesirable crossover effects
(e.g. take-over instead of wash-out effects) cannot be completely ruled out. Provisionally, this
concept must still be regarded as experimental by nature.
• An interesting, but certainly also still experimental – and largely untested – approach is the
“N=1 study” with the option of consolidated meta-analyses.8 In this case individualisation is not
a methodological problem. Somewhat unclear in this study design is the documentation of
changing prescriptions (see Iteration below); also unclear is the external validity.
6 Bell I, Lewis D, Brooks A, Schwartz G, Lewis S, Walsh B, Baldwin C. Improved clinical status in fibromyalgia
patients treated with individualized homeopathic remedies versus placebo. Rheumatology. 2004; 43:577–582.
7 Frei H, Everts R, von Ammon K, et al. Homeopathic treatment of children with attention deficit hyperactivity
disorder: a randomized, double blind, placebo controlled crossover trial. Eur J Pediatr. 2005; 164:758–767.
8 Teut M, Linde K. Scientific case research in complementary and alternative medicine-a review. Complement
Ther Med. 2013 Aug; 21(4):388–95
29
von Ammon K, Kösters C. Methodological problems of randomised double-blind trials
in homeopathic research. 2016.
• The solution that appears to come closest – whilst also the oldest: To permit consecutive
prescriptions with consistent blinding (Walach, headaches9) is probably, methodologically,
more problematic (see Problems due to iteration below).
Problems due to iteration
To date no sufficient methodological evidence has been forthcoming to demonstrate that a double-
blind study is a suitable experimental tool for evaluating several different successive and (within the
meaning of iteration) coordinated prescriptions. Successive prescriptions permitted in a double-blind
study could intensify the blurring of individual prescriptions.
It is recommended, therefore, to apply the double-blind study initially to individual interventions only
(including repetitions of the same agent). This tends to restrict the applicability of the double-blind
study in homeopathic research to simple, acute diseases, however.
Problems due to the outcome
Unlike conventional pharmaceuticals, homeopathic remedies are not suitable for forcing the organism
in question into a specific state – the postulated effect is stimulated self-regulation. The reactions thus
induced are far more individual in their nature and course over time compared to the response to
conventional pharmaceuticals. Assessment of the course in clinical practice is therefore much more
strongly related, in the case of homeopathy, to the symptoms as a whole and also, in particular, to
changes in well-being.
In the planning of a study this alternative approach to assessing the course must at least be
considered – this requirement again tends to conflict, however, with the methodologically
standardised, pre-defined examination times for the respective primary endpoints of a study.10
Also to be considered, however, are other clinical courses and how they can be given adequate
consideration in the design of a study.11
These issues are also mostly relevant to chronic indications and largely negligible in the case of acute
diseases.
9 Walach H, Häusler W, Lowes T, Mussbach D, Schamell U, Springer W, Stritzl G, Haag G. Classical
homeopathic treatment of chronic headaches. Cephalalgia. 1997; 17:119–126.
10 If headache is the primary endpoint, but the mood and energy levels of a patient initially clearly improve, then
in practice (and from the perspective of the patient) this is regarded as a marked improvement – in a study it is
not.
11 The outcome of a study into the treatment of neurodermatitis, compared directly against ongoing cortisone
treatment, is predictable: the effect of cortisone is faster and far better than that of a homeopathic treatment.
Patients who seek homeopathic treatment for neurodermatitis are also aware of this fact. This is not the
question that needs to be answered by a study, therefore. As far as the comparison between homeopathic
treatment and cortisone therapy is concerned, it would be much more interesting if it were conducted in each
case three weeks after discontinuing the last agent (in both study groups). In turn, however, the question
arises whether this scenario also offers an appropriate and fair design from the perspective of conventional
medicine.
30
von Ammon K, Kösters C. Methodological problems of randomised double-blind trials
in homeopathic research. 2016.
Other problems
Further phenomena characteristic of homeopathy, such as primary aggravation or a symptom shift
(centripetal shift in symptoms vs. Hering’s law), also have not yet been considered by the design or
the analysis of studies, respectively. The individual development of such phenomena also conflicts at
this point with the rigid scheduling of follow-up appointments required for the purpose of statistical
analysis. But this, also, tends to be more relevant to chronic conditions.
Entanglement effects were suggested by Walach; hence, they should be accounted for planning a
study. To date, however, neither the existence nor the power of such effects have been adequately
examined or proven.
Appropriate research relating to these problems would be useful in the field of conventional
pharmacology. The question as to possible substance-, remedy- or drug-specific effects in the placebo
group (including proper storage of the products) has not yet been investigated. Clinical experience
tends to suggest that this phenomenon is not particularly relevant.
Meta-analyses of diverse diagnoses in the strict sense of the definition (homogeneous total
population) tend to be questionable from the basic perspective of methodology. Such vast meta-
analyses could be practical as an extension to systematic review.
Implications for homeopathic clinical research
Better and more extensive data are required with respect to routine homeopathic practice.
Conclusions can be drawn from such data with a view to study planning: useful indications, effect
sizes, appropriate study sizes and durations, along with data on the probability of selecting individual
agents. Comprehensive and unselected case documentation will most likely deliver such data.
So far, study planning sometimes appears to be founded upon estimations and individual clinical
advisers and their personal experience. There is room for improvement, accordingly, when it comes to
communication between practice and research. And at least until valid data from everyday practice are
available, extensive pilot phases are recommendable as part of study planning (see above).
The highly deficient development of theories in homeopathy should be stimulated. However, there is
still a need for clarification with respect to individual research tools (appropriately modified RCT
designs, e.g. by excluding non-responders and N=1; see above).
It is not possible to demonstrate at present, however, that a double-blind study, especially with primary
blinding, is a suitable research tool for examining homeopathic treatment in any given indication –
especially in chronic indications.
The justified, understandable and as yet inconclusively resolved question, namely “what contribution
do context effects make to the clinical efficacy of homeopathy?”, in turn can be answered only, and
exclusively, by (adequately elaborated) double-blind studies.
31
von Ammon K, Kösters C. Methodological problems of randomised double-blind trials
in homeopathic research. 2016.
With this in mind, but also in the light of the long-standing and not only scientific debate surrounding
homeopathy12 – a debate that has long since become a vehement social dialogue on a global scale –
the question arises whether research should not concentrate on the central point of this debate.
The clinical effectiveness of the method is not denied, even by sceptics and scientific critics. This
debate does not (or at least not primarily) revolve around whether a homeopathic treatment is effective
in one or another indication, but whether this clinical effectiveness can be attributed to and thus
explained by a specific effect (efficacy) caused by the medicines applied (plausibility paradox, see
footnote 13 below). High potencies, above all, naturally attract considerable attention.
Quite clearly, the relevant number of clinical studies performed in recent decades did not deliver a
definitive conclusion13.
Perhaps the two essential questions raised here – “In which indications is homeopathic treatment
effective?” and “What is the basis for the effect of homeopathic treatment?” – should be addressed
separately. And the second question should be answered by developing independent treatment
models. Distinguishing between questions and developing methodologically appropriate models are
also suitable approaches when considering theoretical scientific aspects. RCTs in the field of
conventional pharmacology address the question whether a specific medicine is effective in a specific
indication. RCTs in homeopathic research address the (clinical) effectiveness of the method, the
efficacy (specific effect) of the drug, and the mechanism of action.
These therapeutic models should then be replicated multiple times – by different examiners – ideally
with a pre-defined number of replications.
Given the methodological problems arising from individualisation and iteration, the therapeutic models
most appropriate here are those based on a limited number of agents which are prescribed according
to a rigid regimen. Clinical situations in which, as a rule, the same agent is prescribed
homeopathically, are ideal. For this reason alone, acute diseases/conditions must be clearly
prioritised.
Very simple models (one condition/one agent) would also offer the advantage that they can potentially
also be analysed and replicated by examiners who have no prior knowledge of homeopathy. A pre-
defined number of replications would also permit a falsification criterion to be established. If the
negative hypothesis applies, the results from replications should equate to the probability of chance.
Testing of the models in appropriate pilot studies is advisable, at least until a clinical database in the
form of unselected case documentation is available.
12 In 2005, the esteemed scientific journal The Lancet officially announced “The end of homeopathy” in its
editorial.
13 “All relevant published reviews [...] reveal that the majority of the available studies produced positive results
(this also applies to that by Shang et al.!). At the same time, it cannot be denied that positive results are not as
common from studies with good methodology as from those that are not so good. Naturally this means that
positive results can certainly also be found in the good studies. The debate amongst clinical researchers
therefore is not whether positive evidence is derived from placebo-controlled studies, but whether such
evidence is sufficient to prove the efficacy of homeopathy in view of the low plausibility from the perspective of
natural science.” (Prof. Klaus Linde, Munich Technical University, personal communication 2007).
32
von Ammon K, Kösters C. Methodological problems of randomised double-blind trials
in homeopathic research. 2016.
Simple models that have already been tested are, for example:
• Treatment of childhood diarrhoea (Jacobs)14
• Treatment of respiratory patients with high-potency potassium dichromate (Frass)15
The treatment of hay fever with pollen (Reilly)16 is also, in principle, a very suitable isopathy model,
given its simplicity, for investigating the effect of potentised substances.
Beyond these therapeutic models, and without unduly influencing the therapeutic setting, the question
of which indication is successfully treated by homeopathy can be answered scientifically to current
methodological standards, depending on the indication, by epidemiological studies, cohort studies,
randomised unblinded studies, and also by unselected case series. This applies in particular to
chronic illnesses.
For reasons of logic, future meta-analyses will then be based predominantly on model studies that are
easy to replicate, and their respective replication(s).
14 Jacobs J, Jonas WB, Jimenez-Perez M, Crothers D: Homeopathy for childhood diarrhea: combined results and
metaanalysis from three randomized, controlled clinical trials. Pediatr Infect Dis J 2003; 22(3):229–234.
15 Frass M, Dielacher C, Linkesch M, Endler C, Muchitsch I, Schuster E, Kaye A. Influence of potassium
dichromate on tracheal secretions in critically ill patients. Chest. 2005; 127:936–41
16 Reilly D, Taylor M, Beattie N, et al., Is Evidence for Homoeopathy Reproducible? Lancet, December 10, 1994,
344:1601–6.
Behnke J. Meta-analyses in homeopathic clinical research. 2016.
Meta-analyses in homeopathic clinical
research
Jens Behnke
_________________________________________________________________________________
1. Introduction
The idea behind evidence-based medicine (EBM) is that one or more meta-analyses of high-quality,
randomised clinical trials will furnish reliable data on the efficacy of a therapeutic method. A treatment
for which sufficient positive data are available within this context is proven by the highest level of
evidence, namely Ia.1 In response to the question whether and to what extent a particular method
should be used in the clinical routine, it is awarded recommendation grade A or a “strong”
recommendation, respectively.2
As homeopathy is a topic of constant debate, both proponents and opponents frequently refer to the
results from meta-analyses of the available clinical studies to support their respective stance in line
with the highest scientific standards of EBM. With this in mind, an attempt will be made below to
provide an outline of the most important global meta-analyses performed on clinical research into
homeopathy. All relevant publications that are not limited to a single clinical indication are presented
and discussed individually so as to formulate an overall opinion on this basis on the status of clinical
research into homeopathy, insofar as this has been addressed by the given publications. The question
that is relevant here is whether the clinical effects of homeopathy are placebo effects, or whether
potentised medicines can produce specific effects.
2.1. Kleijnen, Knipschild and ter Riet (1991)3
This meta-analysis of homeopathy was published in 1991, namely during the era when EBM was in its
infancy. The authors start by referring to what is often claimed: that homeopathy on the one hand is
implausible, and on the other hand has not been investigated using modern methods (controlled
studies). Kleijnen, Knipschild and ter Riet identified 105 evaluable studies that could be included in
their analysis. Of these, 14 examined classical homeopathy involving the individual remedy selection,
58 the prescription of a single homeopathic agent following clinical diagnosis (proven indication), 26 a
combination of remedies, and 16 isopathy. The studies were evaluated for their quality using an
independent method and awarded scores, with the results being included in the final analysis with
different weightings based on their score. A total of 81 studies indicated efficacy for homeopathy in
excess of placebo effects. Of these, the majority were rated as methodologically good in terms of
randomisation, blinding, patient number and similar methodological criteria (15 out of 22). Overall,
Kleijnen, Knipschild and ter Riet (1991) criticise the fact that many studies tend to be of low quality,
whilst at the same time noting the trend in favour of homeopathy, both in those with a sophisticated
design and those which, in their methodology, are rather weak.
1 Cf. Cochrane (2014)
2 Cf. Harbour & Miller (2001)
3 Kleijnen, Knipschild and ter Riet (1991)
34
Behnke J. Meta-analyses in homeopathic clinical research. 2016.
The authors conclude: “At the moment the evidence of clinical trials is positive but not sufficient to
draw definitive conclusions because most trials are of low methodological quality and because of the
unknown role of publication bias. This indicates that there is a legitimate case for further evaluation of
homoeopathy, but only by means of well performed trials.”4
2.2. Linde et al. (1997)5
In the light of the question whether the clinical effectiveness of homeopathy can be explained solely by
placebo effects, Linde et al. (1997) conducted a meta-analysis of all placebo-controlled, randomised
and/or double-blind clinical trials of homeopathy published in the renowned journal The Lancet. Of the
119 studies identified, 89 contained sufficient data to be included in the meta-analysis. Linde et al.
evaluated this subset for its methodological quality using their own method of assessment, performed
by two independent experts whose results were compared against one another. A high level of
agreement was achieved in the assessment (interrater reliability: к=0.76). The average quality of the
examined studies (n=89) with regard to randomisation, double-blinding, handling of study dropouts,
etc., was 52% of the maximum value on a generally accepted scale for evaluating the quality of clinical
trials (Jadad Scale).
The analysis included 13 publications on individualised single-agent homeopathy (classical
homeopathy), 49 that examined prescribing based on clinical diagnosis, 20 that used complex
homeopathic remedies, and seven based on isopathy. Of the studies included, 22% examined
moderate potencies (according to the definition by Linde et al. D9 – D23 or C5 – C11) and 37% high
potencies (above D23 or C11, likewise as per Linde et al.), whereby both substance groups according
to the authors theoretically contained too few molecules of the starting substance to exert any
pharmacological activity (estimated total concentration per patient of less than 10-13 mol/L).
The 89 studies included in the meta-analysis revealed that homeopathy was significantly superior to
placebo (mean odds ratio 2.45; 95% confidence interval 2.05-2.93). In the case of the 26 studies rated
as good methodologically, Linde et al. calculated lower, yet still significant efficacy versus placebo
(odds ratio 1.66; 95% confidence interval 1.33-2.08). This result proved to be robust, moreover, in
various sensitivity analyses (e.g. only the studies of the best quality with pre-defined endpoints which
are listed in the MEDLINE database, n=5: odds ratio 1.97; 95% confidence interval 1.04-3.75).
Correction of the results due to potential errors that could have arisen from selective reporting
(publication bias) also did not eliminate the positive effects of homeopathy. Considering the initial
question of their meta-analysis, the researchers ultimately concluded:
“The results of our meta-analysis are not compatible with the hypothesis that the clinical effects of
homoeopathy are completely due to placebo. However, we found insufficient evidence from these
studies that homoeopathy is clearly efficacious for any single clinical condition. Further research on
homoeopathy is war-ranted provided it is rigorous and systematic.”
The criticism arising from this meta-analysis was aimed primarily at the quality of the studies included,
which Linde et al. (1997) evaluated using their own system and included, accordingly, in their final
analysis. In response to this, the authors screened the material again and created subsets depending
on the Jadad score attained by the studies, at which point they introduced an external assessment
scale.6 In this reanalysis it transpired that the superiority of homeopathy over placebo was lessened to
4 Kleijnen, Knipschild and ter Riet (1991)
5 Linde et al. (1997)
6 Linde et al. (1999)
35
Behnke J. Meta-analyses in homeopathic clinical research. 2016.
some extent if studies of higher quality were considered, without disappearing completely and without
this correlation being linear: The ten studies with the highest Jadad score of 5 demonstrated that the
effect of homeopathic treatment was greater than in the 19 studies with a Jadad score of 3 and the 11
studies that achieved a score of 4. On the whole, it was found that homeopathy was significantly
superior to placebo treatment in each of the six subsets created on the basis of the Jadad score. The
argument that the higher the quality of the study design, the lower the effect of homeopathic treatment,
was thus refuted.
2.3. Cucherat et al. (2000)7
This meta-analysis was undertaken as part of a report for the European Parliament. The authors
conducted a systematic literature review and contacted pharmaceutical companies to identify
randomised, placebo-controlled trials into homeopathy. Only those studies were considered in which
potencies above C3 were used, or products designated by the manufacturer as “homeopathic”. A lack
of blinding was not deemed to be an exclusion criterion. Only studies with a clearly defined primary
endpoint were taken into consideration. Two experts were appointed to evaluate the quality of the
studies. If their opinions differed, a third expert was consulted. Determination of the mean significance
level (p-value) was used as the statistical method for meta-analysis, because the included trials were
very heterogeneous as regards the treated diseases, prescribing methods and endpoints. A combined
p-value of less than 0.05 signified in this case that the null hypothesis was incorrect and the
homeopathic treatment thus differed significantly from placebo.
Of the 118 studies identified, 16 were included in the meta-analysis. The others, in the opinion of the
authors, either had no clearly defined endpoint (92.9%) or were methodologically deficient. One of
these 16 studies contained a total of three groups (standard treatment, homeopathy and placebo).
Hence, 17 comparisons of “homeopathy versus control” were analysed in total. Of these 17 study
outcomes, 11 (65%) delivered a result in favour of homeopathy; three suggested superiority for
placebo that was below the significance level. As the overall conclusion from analysing all the studies
that fulfilled their inclusion and exclusion criteria, Cucherat et al. thus reported a highly significant
mean p-value of 0.000036, which demonstrates clear efficacy for homeopathy versus placebo. The
authors concluded that this outcome was not significantly influenced by publication bias, as by their
estimation 155 notional trials with a negative or non-significant outcome would have been required to
increase the composite p-value of their meta-analysis to above 0.05. A subset analysis revealed,
however, that the significance level in an analysis of only those studies with a dropout rate of less than
5% (n=5) was not achieved (p=0.082). Based on this conclusion, Cucherat et al. postulated:
”There is some evidence that homeopathic treatments are more effective than placebo; however, the
strength of this evidence is low because of the low methodological quality of the trials. Studies of high
methodological quality were more likely to be negative than the lower quality studies.“8
This cautious overall judgement is based on the definition of “methodological quality”, which in the
given case is largely based on the rate of dropouts. Use of the same criterion as the main parameter
for judging the quality of the study is questionable, however. What is essential to the reliability of the
study outcome is not primarily the number of patients ending treatment prematurely, but rather the
total number of cases, the correct randomisation of the subjects to the study groups, the blinding of
patients and physicians, etc. In terms of the dropout rate the statistical method used to compensate for
7 Cucherat et al. (2000)
8 Cucherat et al. (2000)
36
Behnke J. Meta-analyses in homeopathic clinical research. 2016.
the resulting loss of data is mainly relevant to study quality. 9 To this end there are different
approaches, some of which are suitable for rendering dropout rates of far more than 5%
manageable.10 The corresponding methods are standard practice in epidemiology, and the criterion
chosen by Cucherat et al. for evaluating the study quality in their meta-analysis, along with the
limitation of the conclusion of their meta-analysis that resulted, is therefore very unusual from the
viewpoint of methodology.
2.4. Shang et al. (2005)11
This meta-analysis, published in the renowned journal The Lancet, triggered the loudest media echo
of all scientific papers on the subject of homeopathy. The article was accompanied in the Lancet by an
editorial announcing “The end of homeopathy”12.
Shang et al. (2005) referred more or less to the same pool of data as that at the time of publication by
Linde et al. (1997); the final analysis included only eight out of the 110 studies initially reviewed,
however. These eight studies were selected as the largest from a pool of 21 rated at first as
demonstrating good methodology. The combined odds ratio of these homeopathic trials was 0.88
(95% confidence interval 0.65-1.19). For comparison, the authors initially selected 110 studies from
the field of conventional medicine from a Cochrane database, of which six were rated as high quality
and included in the final analysis with a combined odds ratio of 0.58 (95% confidence interval 0.39-
0.85). Shang et al. concluded:
“Biases are present in placebo-controlled trials of both homoeopathy and conventional medicine.
When account was taken for these biases in the analysis, there was weak evidence for a specific
effect of homoeopathic remedies, but strong evidence for specific effects of conventional interventions.
This finding is compatible with the notion that the clinical effects of homoeopathy are placebo
effects.“13
This meta-analysis attracted criticism from various authors who pointed out a number of weaknesses:
Fisher 14 complained about the lack of transparency in study selection insofar as the original
publication by Shang et al. (2005) contained no details of which eight studies were ultimately
analysed. Such a situation is a glaring deviation from the scientific standards used in meta-analyses,
as specified in the QUORUM guideline.15 This guideline demands, amongst others, that all studies
considered in a review – both those ultimately analysed and those excluded from final analysis – be
presented in detail alongside the criteria for such a selection process so that the results of the meta-
analysis and the path taken to obtain them can be understood. Although such a procedure is self-
evident with any scientific publication, Fisher (2006) believes that close observance of such standards
must be demanded, especially of reviews that contain far-reaching, definitive conclusions, such as the
one by Shang et al. (2005).
9 Cf. Sakpal (2010)
10 Cf. EMA (2010)
11 Shang et al. (2005)
12 The Lancet (2005)
13 Shang et al. (2005)
14 Fisher (2006)
15 Moher (1999)
37
Behnke J. Meta-analyses in homeopathic clinical research. 2016.
Fisher also doubts the accuracy of the fit of the 110 studies in conventional medicine against those in
homeopathy. The latter were, on average, of much higher quality, as a result of which the probability
of a positive result diminishes. Shang et al. failed, moreover, to conduct sensitivity analyses which
would have revealed the extent to which the negative outcome depends on study selection.16 Rutten
and Stolper (2008) were of a similar opinion:
“Re-analysis of Shang’s post-publication data did not support the conclusion that homeopathy is a
placebo effect. The conclusion that homeopathy is and that conventional is not a placebo effect was
not based on comparative analysis and not justified because of heterogeneity and lack of sensitivity
analysis.“17
Such a detailed sensitivity analysis was therefore produced by Lüdtke and Rutten (2008) once it
became known which eight studies had led to the presented result. It was found that the analysis of
the 21 high-quality studies from the pool collated by Shang et al. (2005) demonstrated significant
superiority for homeopathy over placebo. The negative conclusion ultimately presented by the authors
is influenced essentially by one single, large study that examines the efficacy of one homeopathic
remedy in the prevention of muscle soreness.18
Another weakness of the meta-analysis performed by Shang et al. is the heterogeneity of the very few
selected studies in relation to the global statement derived from them as a result. The authors
analysed a pool of studies that examined the efficacy of potentised medicines in the prevention of
colds, treatment of warts, prevention of aching muscles, treatment of migraines, childhood diarrhoea,
and severe brain injuries, jointly in only one study in each case. This is a method whereby
consideration should be given to the possibility that homeopathy is an effective treatment for certain
indications, but not for others. If one were to compare, as part of a meta-analysis, three clearly
negative studies of conventional medicine in the treatment of cancer, soft-tissue injuries and rheumatic
disease with specific drugs against a slightly positive study in the treatment of headaches with aspirin,
its negative outcome would not likely be suitable for concluding that all conventional pharmacological
interventions are generally ineffective.
As the author of the first major meta-analysis of homeopathy, likewise published in the Lancet, Linde
also agreed with many of the presented criticisms, as did other experts in the area of researching
complementary medicine.19 On the whole, it must be concluded that there were clear methodological
weaknesses in the review by Shang et al. and that their conclusion is based on a relatively small
number of studies that were selected according to questionable criteria.
2.5. Mathie et al. (2014)
The most recent meta-analysis of homeopathy was performed by Mathie and colleagues in 2014. It
included only those studies with a method of treatment that could be classified as “individualised
homeopathy”. This approach is based on the authors’ assumption that prescribing potentised
medicines cannot necessarily be interpreted as an adequate criterion for it to be associated with a
single coherent treatment procedure. Clinical homeopathy, complex homeopathy, isopathy, etc., differ
at least from the individualised approach of classical homeopathy in that the latter regularly involves
16 Cf. Fisher (2006)
17 Rutten & Stolper (2008)
18 Vickers et al. (1998)
19 Cf. Linde and Jonas (2005), and Bell et al. (2005)
38
Behnke J. Meta-analyses in homeopathic clinical research. 2016.
much longer diagnostic periods. This situation could explain a serious difference compared to the
other prescribing practices.
Mathie et al. thus take an initial step away from the concept of global meta-analysis, which attempts to
answer the question whether any method that can be subsumed under the generic term of
homeopathy is effective. This approach permits a more differentiated view of some of the available
data than some previous reviews, even if Linde et al. (1997) already performed subset analyses which
refer to different methods of prescribing potentised medicines.
Yet the limitation to 32 studies that examine individualised homeopathy is not the only characteristic
feature of the meta-analysis by Mathie et al. The review of the quality of the individual studies, by
assessing the risk of bias using a tool of the Cochrane Collaboration 20 , is also new. Their
methodological guidelines represent the state of the art of EBM, especially with regard to systematic
reviews and meta-analysis. In their methodology, therefore, Mathie et al. are at the height of the game
with their approach.
Of the 32 included studies, 22 delivered sufficient data to warrant inclusion in the final analysis. These
studies delivered a significantly positive result for homeopathy (OR 1.53; 95% confidence interval
1.22-1.91). The assessment of the retrieved study material performed by Mathie et al. (2014) with
respect to the seven aspects (domains) of methological quality (blinding, randomisation, selective
reporting, etc.) according to the Cochrane criteria, however, meant that only three studies were graded
as reliable. The combined odds ratio of these publications, at 1.98 (95% confidence interval 1.16-
3.38), was higher than the average, but given the exclusion of the majority of the retrieved material the
database was so small that the authors stated: „Though our conclusions can be made most securely
from three trials with reliable evidence, this sub-set of studies is too small to enable a decisive answer
to our tested hypothesis.“21
The main results of the meta-analysis by Mathie et al. confer, in their opinion, with those of previous,
comparable reviews: Specific (minor) effects are identifiable with homeopathic treatment. These are
robust, insofar as sensitivity analyses for different subsets of the analysed studies allow similar effect
sizes to be identified. The quality of the evidence found was judged overall to be low or unclear.
Hence, no definitive conclusions could be drawn. Consequently, more high-quality RCTs of
individualised homeopathy are required in order to formulate reliable statements.22
3. Conclusion
In the light of the relevant global meta-analyses of homeopathy published to date it is clear, on the
whole, that in four out of five cases potentised medicines tend to reveal specific efficacy in excess of
placebo. The overall outcome is only negative (homeopathy = placebo) if a large amount (90%-95%)23
of the available data is excluded from the analysis and/or dubious statistical methods are employed. In
each of these cases measures are used that deviate from the usual scientific standards insofar as the
desired increase in reliable data, by excluding studies with certain features, is not reasonably in
proportion with the accepted narrowing of the database (for example, a dropout rate of <10% = 9
studies versus <5% = 5 studies; see above).24
20 Higgins & Altman (2011)
21 Mathie et al. (2014)
22 Ibid.
23 Cf. Hahn (2013)
24 Cf. Cucherat et al. (2000)
39
Behnke J. Meta-analyses in homeopathic clinical research. 2016.
An exception here is the latest study by Mathie et al. (2014), which likewise judges a large proportion
of the data found as (relatively) unreliable. In doing so, however, the authors refer to accepted
standard procedures, and their study produces a positive result for homeopathy. The only study to
ultimately conclude that the clinical effects of homeopathy can be explained entirely by placebo
effects25 reveals considerable shortcomings in terms of methodology.
In homoepathic research, the evaluation of data on the basis of (in)compatibility with certain
theoretical preconceptions appears to play a significant role. As far as scientific theories are
concerned, this phenomenon is referred to as plausibility bias.26 Hahn (2013), for instance, analysed
the inclusion and exclusion criteria of studies in meta-analyses of homeopathy with a negative or
indifferent tendency, revealing, on the basis of statistical considerations amongst others, that they
were probably formulated retrospectively. He suspects that this methodological approach was
ideologically motivated, which would diametrically oppose the scientific demands of EBM.27
The authors of the reviewed publications, however, more or less agree that the available evidence
does not permit any definitive conclusions to be drawn concerning the efficacy of potentised medicines
in individual diseases – the reason being that independent replications of high-quality randomised,
placebo-controlled, double-blind trials referring to one and the same indication are lacking.
Along with other considerations, this observation points the way for the future of homeopathic
research: In this respect an evaluation of the available data from the perspective of model validity
would be useful, along with the design of new, harmonised studies that then results. This could be
achieved with meta-analyses which, with greater differentiation than hitherto, consider individual
prescribing methods and/or, in particular, indications for homeopathic treatment. New data could also
lead to reviews that cover material not previously considered, including non-interventional studies for
example, and on this basis offer a more comprehensive picture of the evidence available on
homeopathy.
References
Bell I.R., Belon P., Bolognani, F. et al. (2005): Letter to the Editor: Are the Clinical Effects of Homeopathy Placebo
Effects? In: The Lancet, 366, 2082.
Das Deutsche Cochrane Zentrum (2014): Von der Evidenz zur Empfehlung (Klassifikationssysteme) [The
German Cochrane Centre (2014): From evidence to recommendation (classification systems)].
http://www.cochrane.de/de/evidenz-empfehlung
Cucherat, M., Haugh, M.C., Gooch, M. & Boissel, J. P. (2000): Evidence of Clinical Efficacy of Homeopathy. A
Meta-Analysis of Clinical Trials. HMRAG. Homeopathic Medicines Research Advisory Group. In: European
Journal of Clinical Pharmacology, 56 (1), 27–33.
Egger, M., Smith, G.D., Schneider, M. & Minder, C. (1997). Bias in Meta-Analysis Detected by a Simple,
Graphical Test. In: British Medical Journal, 15(7109), 624–634.
European Medicines Agency (2010): Guideline on Missing Data in Confirmatory Clinical Trials.
EMA/CPMP/EWP/1776/99 Rev. 1.
[http://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2010/09/WC500096793.pdf;
downloaded on 17.12.2015]
Ernst, E. & Pittler, M.H. (2000): Re-Analysis of Previous Meta-Analysis of Clinical Trials of Homeopathy. In:
Journal of Clinical Epidemiology, 53 (11), 1188.
25 Shang et al. (2005)
26 Rutten et al. (2013)
27 Cf. Hahn (2013)
40
Behnke J. Meta-analyses in homeopathic clinical research. 2016.
Evidence Based Medicine Working Group (1992): Evidence-Based Medicine: A New Approach to Teaching the
Practice of Medicine. In: Journal of the American Medical Association, 286, 2420-2425.
Fisher, P. (2006): Homeopathy and The Lancet. In: Evidence Based Complementary and Alternative Medicine,
3(1), 145–147.
Hahn, R. G. (2013): Homeopathy: Meta-Analyses of Pooled Clinical Data. In: Forschende Komplementärmedizin,
20 (5), 376–381.
Higgins, J.P.T., Altman, D.G. (2011): Chapter 8: Assessing risk of bias in included studies. In: Cochrane
Handbook for Systematic Reviews of Interventions, Version 5.1.0. Edited by: Higgins, J.P.T. & Green, S. The
Cochrane Collaboration, 2011.
Kleijnen, J., Knipschild, P., ter Riet, G. (1991): Clinical Trials of Homoeopathy. In: British Medical Journal,
302(6772), 316–323.
Linde, K. & Jonas, W. (2006): Are the Clinical Effects of Homoeopathy Placebo Effects? In: The Lancet,
366(9503), 2081–2082.
Linde, K. & Melchart, D. (1998): Randomized Controlled Trials in Individualized Homeopathy: A State-of-the-Art
Review. In: Journal of Alternative and Complementary Medicine, 4, 371–388.
Linde, K., Clausius, N., Ramirez, G., Melchart, D., Eitel, F., Hedges, L.V. & Jonas, W.B. (1997): Are the Clinical
Effects of Homoeopathy Placebo Effects? A Meta-Analysis of Placebo-Controlled Trials. In: The Lancet,
350(9081), 834–843.
Linde, K., Scholz, M., Ramirez, G., Clausius, N., Melchart, D. & Jonas, W.B. (1999): Impact of Study Quality on
Outcome in Placebo-Controlled Trials of Homeopathy. In: Journal of Clinical Epidemiology, 52, 631–636.
Lüdtke, R. & Rutten A.L. (2008): The Conclusions on the Effectiveness of Homeopathy Highly Depend on the Set
of Analyzed Trials. In: Journal of Clinical Epidemiology, 61(12), 1197–204.
Mathie, R.T., Lloyd, S.M., Legg, L.A., Clausen, J., Moss, S., Davidson, J.R.T., Ford, I. (2014): Randomised
placebo-controlled trials of individualised homeopathic treatment: Systematic review and meta-analysis.
Systematic Reviews 2014, 3:142 doi:10.1186/2046-4053-3-142
Moher D., Cook, D.J., Eastwood, S., Olkin, I., Rennie, D. & Stroup, D.F. (QUORUM Group) (1999): Improving the
Quality of Reports of Meta-Analyses of Randomised Controlled Trials: The QUORUM Statement. In: The Lancet,
354, 1896–1900.
Peters, D. (2005): Shang et al. Carelessness, Collusion, or Conspiracy? In: Journal of Alternative Complementary
Medicine, 11 (5), 779–780.
Rutten, L., Mathie, R.T., Fisher, P., Goossens, M. & Wassenhoven, M. van (2013): Plausibility and Evidence: The
Case of Homeopathy. In: Medicine, Health Care and Philosophy, 16 (3), 525–532.
Rutten, A.L.B. & Stolper, C.F (2008): The 2005 Meta-Analysis of Homeopathy: The Importance of Postpublication
Data. In: Homeopathy, 97 (4), 169–177.
Sakpal, T.V. (2010): Sample Size Estimation in Clinical Trial. Perspectives on Clinical Research, 1(2): 67–69.
Shang, A. et al. (2005): Are the Clinical Effects of Homoeopathy Placebo Effects? Comparative Study of Placebo-
controlled Trials of Homoeopathy and Allopathy. In: The Lancet, 366 (9487), 726–732.
The Lancet (2005): The End of Homoeopathy. In: The Lancet 366 (9487), 690.
Vickers, A.J., Fisher, P., Smith, C., Wyllie, S.E. & Rees, R. (1998): Homeopathic Arnica 30x Is Ineffective for
Muscle Soreness After Long-distance Running: A Randomized, Double-blind, Placebo-controlled Trial. In: The
Clinical Journal of Pain, 14 (3), 227–231.
Baumgartner S. Status of basic research in homeopathy. 2016.
Status of basic research in homeopathy
Stephan Baumgartner
_________________________________________________________________________________
Homeopathy is based on three principles: testing the remedy in healthy individuals, the similarity
principle, and potentisation. Basic research in homeopathy focuses on scientific evidence for the
similarity principle, and potentisation as a pharmaceutical process.
Criticism of homeopathy is aimed primarily at the process of potentisation, as from the perspective of
natural sciences and pharmacology there are various arguments that question the meaningfulness of
using potentisation as a pharmaceutical process. This includes, amongst others:
• The content of active substances decreases exponentially, to a large extent, on potentisation.
With simple anorganic compounds (such as calcium, sodium, silicium), lower dilutions, which
in practice correspond roughly to 6x-9x, cannot be produced beyond the ubiquitarity limit
(ppm-ppb depending on the material used). With plant and animal extracts it can be assumed
that as of a potency level equivalent to the inverse of the Avogadro number in terms of its
degree of dilution (approx. 24x), the probability of tracing just a single molecule of the starting
substance in the potentised drug rapidly reaches zero. For this reason, questions are raised
concerning specific effects from drugs diluted to higher potencies.
• At lower potencies (to approx. 6x/9x), the substance content can deviate considerably from
the nominally expected concentration – for instance due to minimal solubility in water (e.g.
with metals), or due to adsorption at the vessel walls or dissolution from the same. From the
viewpoint of natural science and pharmacology, this results in undefined concentrations in
potentised drugs.
Goals of basic homeopathic research
In response to the specificity argument, preclinical research into potentised drugs in recent decades
has primarily addressed the question of the existence of possible specific effects from higher-potency
drugs in biological models. Given the difficulties reproducing results achieved (1), the latter question
has developed in the last 20 years into one concerning the development of apt scientific methods:
methods that as far as possible should satisfy the object of the study in its specific properties and
effects – also with a view to obtaining reproducible results (2). Another question concerned
establishment of a correlation between potency level and effect, i.e. determining the dose-effect
relationship. Only few studies addressed the question of pharmaceutical optimisation (e.g. number of
shakes on potentisation) or resistance of the potentised drugs to environmental influences. With a
view to identifying the pharmaceutical mechanism of action, studies were also conducted to determine
any specific physicochemical properties of potentised drugs.
Models and methods of basic homeopathic research
Scientific literature related to preclinical research into potentised drugs is comprehensively
documented in the HomBRex database (3). In 2013, this database contained entries on 1,868
experimental studies discussed in 1,383 publications (3). Four large preclinical research areas with
42
Baumgartner S. Status of basic research in homeopathy. 2016.
potentised drugs can be distinguished: (A) physicochemical measurements, (B) in vitro assays (cell
cultures, microorganisms), (C) bioassays with plants and (D) animal experiments, the latter accounting
for more than 50% of all studies (4).
Research related to physicochemical studies into potentised drugs was last summarised in a review
conducted in 2003 (5). The results from 36 studies were presented in 44 publications. The methods of
analysis were divided into six categories: electrical impedance, electrochemistry, spectroscopy,
nuclear magnetic resonance, Raman spectrosopcy and methods with unknown principles (black box
methods). Half of the studies used nuclear magnetic resonance, where reproducible results were also
achieved across various studies. Based on the currently available data, there are no empirical
indications of stable water clusters in potentised drugs (6, 7), which have been hypothetically
demanded as carriers of potential drug information (8). In several independent studies that measured
NMR relaxation times (T1/T2), however, there were clear and consistent differences between
potentised silicea products and relevant shaken controls which were interpreted as modified molecular
dynamics in water (9). UV spectroscopy also revealed consistent differences between potentised
substances and corresponding controls (10). A theoretical model that interprets such modifications as
carriers of specific drug effects has not yet been devised, however (9). Since 2000, there has been a
marked upswing in physicochemical research into potentised drugs. Hence, by the end of 2015 more
than 150 publications had been identified as part of a systematic review, the content of which is
currently being evaluated by an international study team. The research field is characterised by many
different methods which in some cases are employed only in individual studies and often have not
been replicated. Examples are the investigation for nanoparticles (11) or the reactivity of dyes (12).
The study of potentised products by means of in vitro assays was last addressed in a review
performed in 2007 (13). The 67 experimental studies that were evaluated can be divided into cell-free
systems (e.g. enzymatic models), cell cultures, and models with cells from donated blood. Amongst
the latter, those involving human basophils account for a considerable percentage (42%) of all the
studies. The model used most frequently is inhibition of basophil degranulation by potentised
histamine. Several replication studies and one multicentre study are available on this subject (14). It is
noticeable from the studies with human basophils that effects could also be seen in various
independent studies as a result of high potencies (beyond the inverse of the Avogadro number). The
pattern of active and inactive potency levels within a given sequence of potencies (the “potency
curve”) differed in each study (15), however – both between the different laboratories and on repeating
the process at the same laboratory (16). As the basophils originated in each case from different blood
donors, it is conceivable that the results depend on the individual donor. There is no empirical
evidence of such to date, however.
Plant bioassays are the third research field with respect to the effect and efficacy of potentised
substances. The status of the research was last examined systematically in 2009/2011 (17-19). In the
three main fields, i.e. bioassays with healthy plants, intoxication models and phytophathological
studies, a total of 167 experimental studies were counted in 157 publications. Of these, 48 fulfilled
higher standards of quality and were analysed more closely (20). The test organism used most
frequently was wheat, followed by duckweed and peas. The most commonly used stressor was
arsenic. Silver nitrate was most often used as the potentised substance, followed by arsenic and
gibberellic acid. In various studies, specific effects – also of potencies beyond the inverse of the
Avogadro number – were observed. In the models that examined continuous series of potencies, a
discontinuous relationship between effect and potency level was observed in all cases, i.e. effective
and ineffective potencies alternated according to a defined, albeit different sequence depending on the
study.
43
Baumgartner S. Status of basic research in homeopathy. 2016.
With respect to the fourth large research field, namely animal experiments, only reviews that offer a
cursory description or are limited in terms of time or subject matter have been found recently –
probably on account of the large number of scientific studies overall (4). The studies were in most
cases performed on rats (35%) or mice (29%) (4). The employed research concepts were categorised
by way of example using the rat (21). In most of the experiments a specific disease was artificially
induced in the animals. In the majority of cases this entailed intoxication (e.g. with arsenic, lead, CCl4),
induction of behavioural disorders (e.g. with ethanol, caffeine), inflammatory models (e.g. with
carrageenan) or induced hormonal disorders. The animals were then treated either prophylactically or
curatively with potentised medicines (before or after induction of the disease, respectively). A
systematic review of 105 publications is available on the studies in experimental toxicology (22). A
total of 26 studies were included in a quantitative meta-analysis. A mean protective index of 19.7%
resulted for the examined indications (survival rate, excretion of toxins) after administering potentised
substances, which were selected according to the isopathic principle. Aside from artificially induced
disease models, there were two other model clusters in which the effect of potentised substances was
investigated: development models and animal behaviour models. In case of the development models,
studies repeatedly involved an amphibian model in which the effect of thyroxine 30x on the
metamorphosis of Rana temporaria was examined. The inhibitory effect of thyroxine 30x on such
metamorphosis was judged to be significant in a meta-analysis of 26 studies. The effect appears to be
quite robust, moreover, as both internal and external reproducibility was rated as positive (23). A more
recent review of animal behaviour models identified 18 publications that spanned a wide range of
models and methods (24). There was an emphasis on the potentised drugs Ignatia, Gelsemium and
Chamomilla. Potentised Gelsemium products, used in a number of studies, exhibited highly significant
effects in behavioural experiments on mice which were of the same magnitude as conventional
psychopharmaceuticals at substantial dosage (24).
In summary it can be stated that preclinical research into potentised drugs is characterised by a wide
variety of methods and tested drug potencies. The large number of scientific publications also includes
many studies of good quality that deliver empirical evidence of specific effects for potentised drugs,
also for highly diluted potentised products.
Research into the dose-effect relationship
The outcome of preclinical research into potentised products that has been most adequately
confirmed empirically is the phenomenon of the non-linear relationship between the effect and potency
level. To my knowledge, in every preclinical study that investigated several potency levels of the same
substance, effective and ineffective potency levels were reported to appear consecutively according to
a particular pattern (referred to, after Kolisko (25), as the “potency curve”) (17-19, 24). These patterns
generally prove to be stable within one series of tests (26-28), but over time can also change (28) or
differ between independent laboratories (14, 16). To date, no potency curve has been identified that
remained stable either for the tested potentised substance, the experimental method, or a combination
of substance and method. It is not possible to say at the moment whether this is due to as yet
unidentified determinants or to a particular phenomenon that is inherent to the effects of potentised
drugs.
Research into the similarity principle
The principle of similarity has been investigated far less intensively than that of potentisation: the
current status of research was last reviewed in 2011 (29). Research into hormesis clearly
demonstrated that low doses of a stressor administered before or after an acutely toxic, higher dose of
44
Baumgartner S. Status of basic research in homeopathy. 2016.
the same stressor can ameliorate these harmful effects – both clinically and preclinically (29). Within
the context of homeopathy, this may be regarded as empirical evidence of the isopathic principle. The
best empirical evidence for applicability of the similarity principle (treating “like with like”) in the field of
preclinical research has been obtained by a Dutch study team with studies into cell cultures. The cells
in these studies were firstly exposed to a heat shock. Subsequent treatment with low doses of a series
of different stressors revealed that the survival rate was better, the more similar the respective stressor
was to the heat shock in terms of the spectrum of proteins that developed (30).
Research into the biological mechanism of action
Difficulties reproducing the results from research into potentised substances could be caused by the
fact that the experimental systems and/or parameters chosen to document the nature of the effect of
potentised substances are inadequate. For example, it has been postulated that the activity of
potentised substances is not aimed primarily at exerting a particular effect in a particular direction
(irrespective of the system and its condition), but rather at having compensatory effects or promoting
homeostasis. Translated into an experimental language, this means that primarily effects are not to be
anticipated as regards the mean value of a population, which is reproducibly shifted in a certain
direction, but that the scattering of the system is reduced. This is what was observed in a meta-
analysis across various experimental systems in which the activity of potentised arsenic was
investigated (31). Whether this phenomenon should be attributed specifically to the effect of
potentised arsenic, or whether it is characteristic in general to the effect of potentised drugs, cannot be
determined at present due to the lack of experimental data.
There are some indications that the status of an experimental system decisively influences the
response to a treatment with potentised substances. The response of Pisum sativum to treatment with
potentised gibberellic acid depends on the batch of seed, for instance (32). The metamorphosis of
Rana temporaria is only influenced by potentised thyroxine if the amphibians originate from highland
habitats (33, 34). Lemna gibba responds to potentised gibberellic acid only if the organisms are in a
gibbous state (28). The therapeutic effect exerted by potentised Lyopodium in the phytopathological
model system Malus domestica / Dysaphis plantaginea is especially pronounced if the stress from D.
plantaginea is not too high (35). These examples clearly indicate that potentised substances are seen
to exert an effect in particular if the test organisms are in a mildly, yet not excessively, stressed state.
The degree of divergence from homeostasis evidently may only be so great that the organisms can
regain a state of equilibrium on their own. This statement is to be taken as a working hypothesis at
present and should be carefully examined in the process of developing further methods. After
verification, it could be rated as evidence for a part of the biological mechanism of action of potentised
drugs.
Another aspect of the mechanism of action is formulated in the hypothesis that potentised drugs exert
their effect at the level of the ability of the whole organisms to regulate itself. Empirical evidence is
available in this regard from a number of experimental oncology studies in animals. By systematically
administering potentised drugs in the animals inoculated with cancer cells, the number and size of the
developing tumours were decreased; treatment of the cancer cell lines with the same potentised drugs
in vitro revealed no activity whatsoever (36-39). These results speak clearly in favour of an effect at
the higher level of the organism as a whole. This aspect of the biological mechanism of action of
potentised drugs must also be determined more precisely in future studies. The relevance to the
design of preclinical models is evident: If the hypothesis fits, any efficacy in potentised drugs cannot
be expected from studies in cancer cell lines (40).
45
Baumgartner S. Status of basic research in homeopathy. 2016.
Reproducibility of basic homeopathic research
In 2010, a systematic review to address the question of the reproducibility of effects in basic research
of potentised products identified 24 experimental models that had been used in a total of 107 studies
and were repeated internally at the laboratory or externally (41). Repeat studies in 24 of these models
revealed similar results in 22, differing results in 6, and no effects in 15 models. Independent
reproductions delivered significant results with 7 models. It can therefore be stated that preclinical
research into potentised drugs in the last 10 years has made clear progress: in 1999, no single
experimental model was known in which the results could be reproduced independently (1).
Outlook
Given the advances in methodology, the prospect of addressing diverse questions of scientific and
pharmaceutical relevance is becoming more realistic. Consideration must first be given to determining
the mechanism of action of potentised drugs, at both the pharmaceutical and biomedical level. From
the pharmaceutical perspective, two concepts are conceivable: (1) consolidation and methodological
expansion of the physicochemical studies for determining the specific molecular structure and
dynamics of potentised drugs, and (2) physical interventions into potentised drugs, the effects of which
can be investigated with appropriate bioassays and which then permit conclusions to be drawn on the
mechanism of action. Furthermore, established test methods could be used to examine
pharmaceutically relevant issues, especially with respect to stability under external influences (heat,
pressure, sterile filtration, electromagnetic radiation, etc.) and also with regard to stability in general.
Working groups and organisations
In Europe there are five working groups at present that focus closely on questions of basic research
into potentised products. These working groups are based at the University of Witten/Herdecke
(Germany, Dr Stephan Baumgartner), the University of Bern (Switzerland, Prof. Ursula Wolf), the
Interuniversity College of Graz (Austria, Prof. Christian Endler), the University of Verona (Italy, Prof.
Paolo Bellavite) and the University of Bologna (Italy, Prof. Lucietta Betti). Elsewhere, there are several
university-based working groups in Brazil and India.
Notable organisations are the GIRI (Groupe International de Recherche sur l'Infinitésimal,
www.giriweb.com), a scientific association dedicated to basic research into potentised products, the
Karl and Veronica Carstens Foundation as the supporters of the HomBRex database (www.carstens-
stiftung.de), and the HRI (Homeopathy Research Institute), as the organiser of conferences on the
subject of homeopathic research (www.homeoinst.org).
Summary
The following can be concluded on the status of preclinical research into potentised substances:
Amongst the available specialist scientific publications, totalling more than 1,000, there is a significant
collection of studies of high quality that has delivered empirical evidence of specific efficacy, also with
highly diluted, potentised drugs. Likewise, there are several experimental models with which
significant specific effects were observed for potentised products when replicated independently. The
empirically ascertained modification of molecular dynamics in potentised drugs could give indications
of the physicochemical and pharmaceutical mechanisms of action, which as a whole must still be
determined, however. There is also initial empirical evidence of the biological mechanism of action,
described by a regulatory reaction of the entire organism to deviations from homeostasis.
46
Baumgartner S. Status of basic research in homeopathy. 2016.
References
1. Vickers AJ. Independent Replication of Pre-Clinical Research in Homoeopathy: A Systematic Review.
Forschende Komplementärmedizin. 1999; 6(6):311-20.
2. Baumgartner S. The State of Basic Research in Homeopathy. In: Witt C, Albrecht H, editors. New Directions
in Homeopathy Research – Advice From an Interdisciplinary Conference. Essen: KVC-Verlag; 2009. p. 107–
30.
3. Clausen J, van Wijk R, Albrecht H. Geographical and temporal distribution of basic research experiments in
homeopathy. Homeopathy. 2014; 103(3):193-7.
4. van Wijk R, Albrecht H. Classification of systems and methods used in biological basic research on
homeopathy. Homeopathy. 2007; 96(4):247-51.
5. Becker-Witt C, Weisshuhn TER, Lüdtke R, Willich SN. Quality Assessment of Physical Research in
Homeopathy. The Journal of Alternative and Complementary Medicine. 2003; 9(1):113-32.
6. Aabel S, Fossheim S, Rise F. Nuclear magnetic resonance (NMR) studies of homeopathic solutions. British
Homeopathic Journal. 2001; 90:14-20.
7. Anick DJ. High sensitivity 1H-NMR spectroscopy of homeopathic remedies made in water. BMC
Complementary and Alternative Medicine. 2004; 4:15.
8. Anagnostatos GS, Vithoulkas G, Garzonis P, Tavouxoglou C. A Working Hypothesis for Homeopathic
Microdiluted Remedies. The Berlin Journal on Research in Homoeoepathy. 1991; 1(3):141-7.
9. Demangeat JL. Nanosized solvent superstructures in ultramolecular aqueous dilutions: twenty years'
research using water proton NMR relaxation. Homeopathy. 2013; 102(2):87-105.
10. Klein SD, Sandig A, Baumgartner S, Wolf U. Differences in Median Ultraviolet Light Transmissions of Serial
Homeopathic Dilutions of Copper Sulfate, Hypericum perforatum, and Sulfur. Evid Based Complement
Alternat Med. 2013; 2013:370-609.
11. Chikramane P, Suresh A, Bellare J, Kane S. Extreme homeopathic dilutions retain starting materials: A
nanoparticulate perspective. Homeopathy. 2010; 99:231-42.
12. Cartwright SJ. Solvatochromic dyes detect the presence of homeopathic potencies. Homeopathy. 2015; in
press http://dx.doi.org/10.1016/j.homp.2015.08.002.
13. Witt CM, Bluth M, Albrecht H, Weisshuhn TE, Baumgartner S, Willich SN. The in vitro evidence for an effect
of high homeopathic potencies – A systematic review of the literature. Complement Ther Med. 2007;
15(2):128-38.
14. Belon P, Cumps J, Ennis M, Mannaioni PF, Roberfroid M, Sainte-Laudy J, et al. Histamine dilutions modulate
basophil activation. Inflamm Res. 2004; 53:181–8.
15. Guggisberg AG, Baumgartner S, Tschopp CM, Heusser P. Replication study concerning the effects of
homeopathic dilutions of histamine on human basophil degranulation in vitro. Complementary Therapies in
Medicine. 2005; 13(2):91-100.
16. Baumgartner S, Guggisberg A. Journal Club: Basophilendegranulation, dritter Akt: Homöopathie und
Basophilenreaktion – weniger klar, als manche das gern hätten. Kommentar [Journal Club: Degranulation of
basophils, third act: Homeopathy and basophilic reaction – less transparent than many would like to believe.
Commentary]. Forsch Komplentarmed Klass Naturheilkd. 2005; 12:52–4.
17. Majewsky V, Arlt S, Shah D, Scherr C, Jager T, Betti L, et al. Use of homeopathic preparations in
experimental studies with healthy plants. Homeopathy. 2009; 98(4):228-43.
18. Betti L, Trebbi G, Majewsky V, Scherr C, Shah-Rossi D, Jager T, et al. Use of homeopathic preparations in
phytopathological models and in field trials: a critical review. Homeopathy. 2009; 98(4):244-66.
19. Jäger T, Scherr C, Shah D, Majewsky V, Betti L, Trebbi G, et al. Use of homeopathic preparations in
experimental studies with abiotically stressed plants. Homeopathy. 2011; 100(4):275-87.
20. Jäger T, Scherr C, Shah D, Majewsky V, Wolf U, Betti L, et al. The use of plant-based bioassays in
homeopathic basic research. Homeopathy. 2015; in press http://dx.doi.org/10.1016/j.homp.2015.06.009.
21. van Wijk R, Clausen J, Albrecht H. The rat in basic therapeutic research in homeopathy. Homeopathy. 2009;
98(4):280-6.
22. Linde K, Jonas WB, Melchart D, Worku F, Wagner H, Eitel F. Critical Review and Meta-Analysis of Serial
Agitated Dilutions in Experimental Toxicology. Hum Exp Toxicol. 1994; 13:481-92.
23. Harrer B. Replication of an experiment on extremely diluted thyroxine and highland amphibians.
Homeopathy. 2013; 102(1):25-30.
47
Baumgartner S. Status of basic research in homeopathy. 2016.
24. Bellavite P, Magnani P, Marzotto M, Conforti A. Assays of homeopathic remedies in rodent behavioural and
psychopathological models. Homeopathy. 2009; 98(4):208-27.
25. Kolisko L. Physiologischer und physikalischer Nachweis der Wirksamkeit kleinster Entitäten [Physiological
and physical evidence of the efficacy of the smallest entities]. Stuttgart: Der Kommende Tag AG; 1923.
26. Scherr C, Simon M, Spranger J, Baumgartner S. Effects of potentised substances on growth rate of the water
plant Lemna gibba L. Complement Ther Med. 2009; 17:63–70.
27. Jäger T, Scherr C, Simon M, Heusser P, Baumgartner S. Effects of homeopathic arsenicum album, nosode,
and gibberellic acid preparations on the growth rate of arsenic-impaired duckweed (Lemna gibba L.).
ScientificWorldJournal. 2010; 10:2112-29.
28. Majewsky V, Scherr C, Arlt SP, Kiener J, Frrokaj K, Schindler T, et al. Reproducibility of effects of
homeopathically potentised gibberellic acid on the growth of Lemna gibba L. in a randomised and blinded
bioassay. Homeopathy. 2014; 103(2):113-26.
29. van Wijk R, Wiegant FAC. Postconditioning hormesis and the similia principle. Frontiers in Bioscience. 2011;
E3:1128-38.
30. Wiegant F, Van Wijk R. The similia principle: Results obtained in a cellular model system. Homeopathy.
2010; 99(1):3-14.
31. Nani D, Brizzi M, Lazzarato L, Betti L. The role of variability in evaluating ultra high dilution effects:
considerations based on plant model experiments. Forschende Komplementärmedizin. 2007; 14(5):301-5.
32. Baumgartner S, Shah D, Schaller J, Kämpfer U, Thurneysen A, Heusser P. Reproducibility of dwarf pea
shoot growth stimulation by homeopathic potencies of gibberellic acid. Complementary Therapies in
Medicine. 2008; 16:183–91.
33. Endler PC, Lüdtke R, Heckmann C, Zausner C, Lassnig H, Scherer-Pongratz W, et al. Pretreatment with
thyroxin (10-8 parts by weight) enhances a "curative" effect of homeopathically prepared thyroxin (10-13) on
lowland frogs. Forsch Komplentarmed Klass Naturheilkd. 2003; 10:137-42.
34. Zausner C, Lassnig H, Endler PC, Scherer-Pongratz W, Haidvogl M, Frass M, et al. Die Wirkung von
"homöopathisch" zubereitetem Thyroxin auf die Metamorphose von Hochlandamphibien – Ergebnisse einer
multizentrischen Kontrollstudie [The action of “homeopathically” prepared thyroxine on the metamorphosis of
highland amphibians – results of a controlled multicentre study]. Perfusion. 2002; 15(7):268-76.
35. Wyss E, Tamm L, Siebenwirth J, Baumgartner S. Homeopathic preparations to control the rosy apple aphid
(Dysaphis plantaginea Pass.). ScientificWorldJournal. 2010; 10:38-48.
36. Jonas WB, Gaddipati JP, Rajeshkumar NV, Sharma A, Thangapazham RL, Warren J, et al. Can
homeopathic treatment slow prostate cancer growth? Integr Cancer Ther. 2006; 5(4):343-9.
37. Thangapazham RL, Gaddipati JP, Rajeshkumar NV, Sharma A, Singh AK, Ives JA, et al. Homeopathic
medicines do not alter growth and gene expression in prostate and breast cancer cells in vitro. Integr Cancer
Ther. 2006; 5(4):356-61.
38. Thangapazham RL, Rajeshkumar NV, Sharma A, Warren J, Singh AK, Ives JA, et al. Effect of homeopathic
treatment on gene expression in Copenhagen rat tumor tissues. Integr Cancer Ther. 2006; 5(4):350-5.
39. Saha S, Hossain DM, Mukherjee S, Mohanty S, Mazumdar M, Mukherjee S, et al. Calcarea carbonica
induces apoptosis in cancer cells in p53-dependent manner via an immuno-modulatory circuit. BMC
Complement Altern Med. 2013; 13:230.
40. Wälchli C, Baumgartner S, Bastide M. Effect of low doses and high homeopathic potencies in normal and
cancerous human lymphocytes: an in vitro isopathic study. Journal of Alternative and Complementary
Medicine. 2006; 12(5):421-7.
41. Endler P, Thieves K, Reich C, Matthiessen P, Bonamin L, Scherr C, et al. Repetitions of fundamental
research models for homeopathically prepared dilutions beyond 10(-23): a bibliometric study. Homeopathy.
2010; 99(1):25-36.
Authors
Dr. med. Klaus von Ammon
Neurosurgery FMH, Specialist in homeopathy FMH/SVHA
Born 1955 in Munich
1974-1980 Study of human medicine in Hamburg, Marburg/Lahn and Munich
1980-1988 Advanced training to become a neurosurgeon, Neurosurgical Department, Surgical Hospital
and Outpatients Clinic of the Technical University of Munich “rechts der Isar”
1989-1996 Consultant and lecturer at the Neurosurgical Department of Zurich University Hospital
1997-1999 Advanced training in homeopathy (SVHA/ZAKH, Zurich), with continuing professional
development ever since
2003 Certificate of qualification in homeopathy
Consultant in homeopathic research, University of Bern, Institute for Complementary Medicine (IKOM),
publishing, teaching and consulting on scientific matters since 2000
Medical practice in classical homeopathy in Stäfa, Zürichsee
Since 2005 provision of continuing professional development lectures nationally and internationally
2010-2016 WissHom speaker for research department
Contact
Rebbergstraße 7
CH-8712 Stäfa
Phone: 0041-(0)44-926 89 59
E-mail: klaus@vonammon.ch
Institut für Komplementärmedizin (IKOM), Universität Bern
Inselspital, Freiburgstraße 46, CH-3010 Bern
Phone: 0041-(0)31-632 42 66
E-mail: klaus.vonammon@ikom.unibe.ch
PD Dr. sc. nat. Stephan Baumgartner
Born 1965 in Munich. 1984-1990 Study of physics, mathematics and astronomy at the Basel University.
1995 Doctorate in environmental sciences at the Federal Institute of Technology (ETH) in Zurich. Since
1993 scientific associate in the department of basic research at the Hiscia Institute, Society for Cancer
Research, Arlesheim, and also since 1996 at the Institute of Complementary Medicine at the University of
Bern. Since 2011 scientific associate at the Institute for Integrative Medicine of the University of
Witten/Herdecke. 2013 Postdoctorate from the University of Witten/Herdecke.
Focus of research: Development of experimental methods to investigate pharmaceutical processes in
complementary medicine (focusing on potentisation methods in homeopathy and anthroposophic
medicine); systematic reviews on basic research into potentised products.
50 publications in peer-reviewed scientific journals, 144 presentations at scientific conferences. Teaching:
Talks and seminars for students of medicine and pharmacy; continuing professional development courses for physicians,
pharmacists and health practitioners; mentoring for master theses and dissertations; conference organisation.
Memberships: German Physical Society (DPG) and Scientific Society for Homeopathy (WissHom). Founding member of the
International Society for Complementary Medicine Research (ISCMR) and the Association for Anthroposophically Extended
Pharmacy in Switzerland (VAEPS). Currently president of the International Research Group on Very Low Dose and High
Dilution Effects (GIRI).
Contact
Institut für Integrative Medizin (IfIM), Universität Witten/Herdecke
Gemeinschaftskrankenhaus, Gerhard-Kienle-Weg 4, D-58313 Herdecke
E-mail: stephan.baumgartner@uni-wh.de
Dr. phil. Jens Behnke
Jens Behnke heads the department of “Homeopathy Research and Teaching” of the Carstens Foundation:
Nature and Medicine. His responsibilities include devising research strategies, managing and continuously
developing the CORE-Hom database (clinical research into homeopathy), and managing university-based
homeopathy working groups along with the corresponding optional courses for medical degrees.
Contact
Karl und Veronica Carstens Stiftung, Am Deimelsberg 36, D-45130 Essen
Phone: 0049-(0)201-5 63 05-13
Websites: www.carstens-stiftung.de / www.naturundmedizin.de
Mail: j.behnke@carstens-stiftung.de
Dr. med. Martin Frei-Erb, MD
Medical studies at University of Bern (Switzerland), federal diploma 1985. Specialization in general internal
medicine 1993 with postgraduate training in internal medicine, surgery, rheumatology and psychiatry.
Doctoral thesis in 1990, Faculty of Medicine, University of Bern (Switzerland). Training in homeopathy from
1990 to 1992 in Bern, further training at different international courses. Proficiency certificate in
Homeopathy (FMH/SVHA) in 2000. Working as general practitioner since 1993 in his own medical practice
at Thun together with two colleagues. Co-director/managing director, Institute of Complementary Medicine
IKOM, University of Bern (Switzerland), since 4/2008. Lecturer for family medicine since 2009, Institute of
Primary Care BIHAM, University of Bern (Switzerland).
Contact
Institut für Komplementärmedizin (IKOM), Universität Bern
PH 4, Inselspital, CH-3010 Bern
Mail: martin.frei@ikom.unibe.ch
50
Curt Kösters
Studies at Free University of Berlin, clinical training at a department for internal medicine in Hamburg
(1987-1989), training in homeopathy as of 1989 in Berlin and Hamburg; after a study visit to India in 1992,
private practice opened in Hamburg.
2004-2011: Various positions on the board of the German Central Society of Homeopathic Doctors
(DZVhÄ), chairman 2007-2010; speaker in the department for quality improvement at the Scientific Society
for Homeopathy (WissHom) since November 2010.
Since 1996: Lecturer in advanced homeopathic training and lecture series at the universities of Kiel,
Lübeck, Hamburg; chairman of the Society of Homeopathic Physicians in Schleswig-Holstein and the
Hanseatic Cities.
1996-2004: Contributor to a homeopathic development aid project in Nepal (Bhaktapur International Homeopathic Clinic).
As of 1998 licence to provide training in homeopathy and consultant at Hamburg Medical Association.
2001: Organisation and scientific management of the Annual Congress of the German Central Association of Homeopathic
Doctors (DZVhÄ) in Hamburg.
2002/2003: Management of homeopathic provings as part of advanced training courses.
As of 2002: Involvement in InHom (European Institute for Homeopathy) – focusing on databases, case documentation, chronic
illnesses and their treatment, interventions in accordance with the similarity principle in social contexts.
2003: Co-editor of “International Debate – Principles of Homeopathy”.
As of 2004: Collaboration in an interdisciplinary working group on the application of homeopathic principles to town planning (at
the International Building Exhibition 2010).
Contact
Eggerstedtstraße 56, D-22765 Hamburg
E-mail: Curt.Koesters@hamburg.de
Dr. med. Michael Teut
Specialist in general medicine, additional qualifications in homeopathy, nutritional medicine, hypnosis/hypnotherapy. Clinical trial
investigator.
Study of medicine in Göttingen and Leiden, parallel training in classical homeopathy and phytotherapy.
Since 2007, clinical work and research at the Institute for Social Medicine, Epidemiology and Health
Economics of Charité – Universitätsmedizin Berlin.
Teaching and lecturing on complementary medicine, nutritional medicine, social medicine, prevention,
research methodology. Member of Commission D of the BfArM, the Pluralism in Medicine dialogue forum
at the Federal Medical Association of the German Society for General Medicine, the German Central
Association of Homeopathic Doctors, the German Society of Hypnosis and Hypnotherapy, the International
Society for Hypnosis, and at Nature and Medicine e.V., the association supporting the Karl and Veronica
Carstens Foundation.
Contact
Institut für Sozialmedizin, Epidemiologie und Gesundheitsökonomie, Charité – Universitätsmedizin Berlin
Luisenstraße 5, D-10117 Berlin
Phone: 0049-(0)30-450 529 234
Websites: www.michael-teut.de/www.hochschulambulanz-naturheilkunde.de/unser_team/dr_michael_teut/
E-mail: michael.teut@charite.de
Dr. phil. Loredana Torchetti
1998-2000 / 2002-2006 Study of psychology (clinical and social psychology) and psychopathology
at the University of Bern
2007-2013 Dissertation on personality psychology at the University of Bern
2009-2011 Diploma in applied statistics, Federal Institute of Technology (ETH) in Zurich
Since 2013 scientific associate at the Institute for Complementary Medicine (IKOM) at the University of
Bern
Contact
Institut für Komplementärmedizin (IKOM), Universität Bern
PH 4, Inselspital, CH-3010 Bern
E-mail: loredana.torchetti@ikom.unibe.ch
www.wisshom.de
