Homeopathy (2017) -, 1e9

Ó 2017 The Faculty of Homeopathy. Published by Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.homp.2017.07.003, available online at http://www.sciencedirect.com

ORIGINAL PAPER

Model validity of randomised

placebo-controlled trials of
non-individualised homeopathic
treatment
Robert T Mathie1,*, Michel Van Wassenhoven2, ALB (Lex) Rutten3, Christien T Klein-Laansma4, Jos
e Eizayaga5,
Anna Pla i Castellsagu
e6, Miek C Jong4,7,8, Raj K Manchanda9, Fl
avio Dantas10, Menachem Oberbaum11,
Joyce Frye12, Helmut Roniger13, Stephan Baumgartner14, Robbert van Haselen15, Ton Nicolai3 and Peter Fisher13
1
Homeopathy Research Institute, 124 Cromwell Road, London SW7 4ET, UK
2
Belgian Homeopathic Medicines Registration Committee, FAMHP (Federal Agency for Medicines and Health Products),
Belgium
3
Independent Researcher
4
Department Nutrition & Health, Louis Bolk Institute, Driebergen, The Netherlands
5
Department of Homeopathy, Maimonides University, Buenos Aires, Argentina
6
Research Sub-committee, European Committee for Homeopathy, Belgium
7
Department of Health Sciences, Mid-Sweden University, Sundsvall, Sweden
8
National Information and Knowledge Centre for Integrative Medicine, The Netherlands
9
Central Council for Research in Homoeopathy, India
10
Faculty of Medicine, Federal University of Uberl^ andia, Uberl^andia, Brazil
11
Shaare Zedek Medical Center, Jerusalem, Israel
12
Center for Integrative Medicine, University of Maryland School of Medicine, Baltimore, MD, USA
13
Royal London Hospital for Integrated Medicine, London, UK
14
Institute of Integrative Medicine, University of Witten-Herdecke, Germany
15
International Institute for Integrated Medicine, Kingston, UK

Background: The comprehensive systematic review of randomised placebo-controlled

trials (RCTs) in homeopathy requires examination of a study’s model validity of homeo-
pathic treatment (MVHT) as well as its risk of bias (extent of reliable evidence).
Objective: To appraise MVHT in those RCTs of non-individualised homeopathy that an
associated investigation had judged as ‘not at high risk of bias’.
Design: Systematic review.
Methods: An assessment of MVHT was ascribed to each of 26 eligible RCTs. Another 49
RCTs were ineligible due to their high risk of bias.
Main outcome measures: MVHT and the prior risk of bias rating per trial were merged
to obtain a single overall quality designation (‘high’, ‘moderate’, ‘low’), based on the
GRADE principle of downgrading.
Results: The trials were rated as ‘acceptable MVHT’ (N = 9), ‘uncertain MVHT’ (N = 10)
and ‘inadequate MVHT’ (N = 7); and, previously, as ‘reliable evidence’ (N = 3) and ‘non-
reliable evidence’ (N = 23). The 26 trials were designated overall as: ‘high quality’
(N = 1); ‘moderate quality’ (N = 18); ‘low quality’ (N = 7).

*Correspondence: Robert T Mathie, Homeopathy Research Institute, 124 Cromwell Road, London SW7 4ET, UK.
E-mail: robertmathie@hri-research.org, Michelvw@homeopathy.be, praktijk@dokterrutten.nl, ctlaansma@planet.nl, jose.eizayaga@
gmail.com, pla.c.anna@gmail.com, m.jong@louisbolk.nl, rkmanchanda@gmail.com, dantas@ufu.br, oberbaum@netvision.net.il,
joyce.frye@gmail.com, helmut.roniger@uclh.nhs.uk, stephan.baumgartner@uni-wh.de, vanhaselen@intmedi.com, tonnicolai@me.
com, peter.fisher@uclh.nhs.uk
Received 1 February 2017; revised 23 June 2017; accepted 31 July 2017

Please cite this article in press as: Mathie RT, et al., Model validity of randomised placebo-controlled trials of non-individualised homeopathic treatment,
Homeopathy (2017), http://dx.doi.org/10.1016/j.homp.2017.07.003
 Model validity of randomised placebo-controlled trials
RT Mathie et al
2
Conclusion: Of the 26 RCTs of non-individualised homeopathy that were judged ‘not at
high risk of bias’, nine have been rated ‘acceptable MVHT’. One of those nine studies was
designated ‘high quality’ overall (‘acceptable MVHT’ and ‘reliable evidence’), and is thus
currently the only reported RCT that represents best therapeutic practice as well as un-
biased evidence in non-individualised homeopathy. As well as minimising risk of bias,
new RCTs in this area must aim to maximise MVHT and clarity of reporting. Homeop-
athy (2017) -, 1e9.

Keywords: Model validity; Non-individualised homeopathy; Randomised placebo-

controlled trials; Risk of bias; Systematic review

Background Objective: To appraise MVHT in the RCTs of non-

Our programme of systematic reviews and meta-analyses
of randomised placebo-controlled trials (RCTs) in homeop-
athy includes examination of each eligible trial’s risk of bias
as well as of its model validity (MV). The latter attribute re-
flects the concordance between the trial design and ideal
practice for the intervention under investigation, and is a
key facet of study quality in RCTs of complementary/alter-
native medicine (CAM) therapies such as homeopathy.1e3
We have previously reported our evaluations of RCTs of
individualised homeopathic treatment, presenting the
findings in connected papers.4e6 We concluded that an
individually prescribed homeopathic medicine may have
a small treatment effect beyond that of placebo; however,
decisive interpretation was undermined by the paucity of
high-quality evidence.4,6
The present paper focuses on placebo-controlled RCTs
of non-individualised homeopathic treatment. Trials of
this nature comprise the majority of the RCT literature
in homeopathy.7 Akin to a conventional drug trial, these
are studies in which the same homeopathic medicine (or
its corresponding placebo) has been given to each of
the trial participants. The in-depth, individualised, ho-
meopathic prescribing approach is therefore not involved.
Instead, non-individualised homeopathy trials have some
basis in clinical homeopathy, in which a single medicine,
or combination of medicines, is prescribed to an individ-
ual patient on the basis of a specified somatic symptom or
set of symptoms.8 The medicine may be a single tradi-
tional homeopathic remedy (e.g. Arnica), or a formulation
of several remedies that is either a proprietary medicine
(‘complex medicinal product’) or a unique formulation
prepared for the purposes of the research. Alternatively,
the single medicine may be a homeopathic nosode based
on the principle of isopathy (‘the use of medicines
derived from a causative agent of the disease itself, or
from a product of the disease process, to treat the condi-
tion’8).
The extent to which these several modes of homeopathy
are successfully reflected in RCTs of non-individualised
treatment is the subject of the present paper. As previ-
ously,5 we assess model validity of homeopathic treatment
(MVHT), defined as the extent to which a homeopathic
intervention and the main measure of its outcome, as im-
plemented in an RCT, reflect best clinical practice in home-
opathy.
individualised homeopathy which, in an associated investi-
gation,9 were judged to be not at ‘high risk of bias’.
Through this approach, we aimed to identify the overall
high-, moderate- and low-quality evidence in RCTs of
non-individualised homeopathic treatment.
Methods
Inclusion criteria for RCTs
We applied our MVHT assessment method (which was
designed to examine RCTs of either individualised or
non-individualised homeopathy) to peer-reviewed papers
reporting randomised, placebo-controlled trials of non-
individualised homeopathic treatment, published up to
and including 2014. Through formal literature search
methods, 110 records were identified as being potentially
eligible for systematic review in this RCT category: after
application of pre-defined exclusion criteria, 72 records
(reporting a total of 75 RCTs) remained eligible for sys-
tematic review.9 Of those 75 RCTs, 49 were rated as
‘high’ risk of bias (relevant details in Appendix 1)9; the re-
maining 26 RCTs (‘uncertain’ or ‘low’ risk of bias) were
therefore the material for the present study on MVHT (de-
tails, including reference citations,10e35 in Table 1).
Assessment of model validity
The development of our criteria-based method for
MVHT has been described in detail elsewhere.3,5 The
assessment domains are as follows:
Domain I (Rationale): Would a significant body of ac-
credited homeopaths support the rationale for the interven-
tion used in the study?
Domain II (Principles): Is the specific intervention used
consistent with homeopathic principles?
Domain III (Practitioner): Does the study have suitably
qualified and experienced homeopathic practitioner input?
Domain IV (Outcome measure): Does the main
outcome measure reflect the main effect expected of the
intervention used?
Domain V (Outcome sensitivity): Is the main outcome
measure capable of detecting change?
Domain VI (Follow-up): Is the length of follow-up for
the main outcome measure appropriate to detect the in-
tended effect of the intervention?

Homeopathy
Please cite this article in press as: Mathie RT, et al., Model validity of randomised placebo-controlled trials of non-individualised homeopathic treatment,
Homeopathy (2017), http://dx.doi.org/10.1016/j.homp.2017.07.003
Homeopathy (2017), http://dx.doi.org/10.1016/j.homp.2017.07.003
Please cite this article in press as: Mathie RT, et al., Model validity of randomised placebo-controlled trials of non-individualised homeopathic treatment,

Table 1 The 26 trials that were assessed for model validity of homeopathic treatment (MVHT)

No. Author, date [Ref] Category Condition Main outcome measure Endpoint Hom. type
10
A47 Baker, 2003 Mental disorder Anxiety Revised test anxiety (RTA) scale 4d Single
A48 Balzarini, 200011 Dermatology Radiodermatitis Index of total severity during recovery (RTSI) (re: skin 7e8 wk Combination
colour, temp, oedema, pigmentation)
A272 Colau, 201212 Obstetrics & gynaecology Menopausal syndrome Hot flash score 12 wk OTC complex
A61 Cornu, 201013 Surgery & anaesthesiology Post-operative bleeding Cumulated blood loss at drain removal Up to 7 d Combination
A62 Diefenbach, 199714 Respiratory infection Bronchitis Treatment success (‘very good’ + ‘good’ results) e Up to 3 wk OTC complex
physician assessed
A64 Ferley, 198915 Respiratory infection Influenza Proportion of patients recovered (from 5 cardinal By 48 h Isopathy*
symptoms and from temp>37.5)
A67 Frass, 200516 Respiratory infection Tracheal secretions Total volume of tracheal secretions per day 2d Single
A75 GRECHO, 198917 Surgery & anaesthesiology Post-operative ileus Number of hours from operation until first stool Up to 100 h approx. Combination
A79 Hofmeyr, 199018 Obstetrics & gynaecology Postpartum pain Daily questionnaire responses: perineal pain 4d Single
A83 Kaziro, 198419 Surgery & anaesthesiology Post-operative pain/swelling Pain score (VAS) 8d Single
A92 Leaman, 198920 Dermatology Minor burns Pain (0e10 VAS) e area under curve 6h Single
A93 Lewith, 200221 Allergy & asthma Allergic asthma Asthma VAS 16 wk Isopathy
A292 Malapane, 201422 Ear, nose & throat Tonsillitis Tonsillitis pain score (Wong-Baker FACES) 6d OTC complex
A290 Naidoo, 201323 Allergy & asthma Allergic skin reaction Wheal diameter 4 wk Combination
A100 Oberbaum, 200124 Oral/dental Stomatitis Area-under-curve (AUC) score for stomatitis 14 d minimum OTC complex
symptoms (severity and duration)

RT Mathie et al
Model validity of randomised placebo-controlled trials
A103 Padilha, 201125 Toxicology Lead poisoning Proportion of workers with Pb decrease of at least 30 d Single
25%
A104 Papp, 199826 Respiratory infection Influenza Proportion of patients with physician-assessed By 48 h Isopathy*
recovery in health (i.e. ‘no symptoms’)
A105 Paris, 200827 Surgery & anaesthesiology Post-operative analgesic intake Cumulated consumption of morphine < 10 mg/day 24 h post-op Combination
A112 Reilly, 199428 Allergy & asthma Allergic asthma Proportion with improvement in daily overall VAS 4 wk Isopathy
score
A120 Singer, 201029 Surgery & anaesthesiology Post-operative pain Maximum daily pain at rest, using 11-point numerical 13 d OTC complex
rating scale (NRS-11)
30
A123 Taylor, 2000 Allergy & asthma Perennial allergic rhinitis Proportion with improvement in daily overall VAS 3e4 wk Isopathy
score
A125 Tveiten, 199131 Musculoskeletal Muscle soreness Muscle soreness (VAS) 3d Single
A126 Tveiten, 199832 Musculoskeletal Muscle soreness Muscle soreness (VAS) 3d Single
A128 Vickers, 199833 Musculoskeletal Muscle soreness Muscle soreness (VAS) 2d Single
A135 Wiesenauer, 199534 Allergy & asthma Seasonal allergic rhinitis Free of nasal symptoms 4 wk Single
A137 Zabolotnyi, 200735 Ear, nose & throat Sinusitis Sinusitis severity score (SSS) cf. Day 0 21 d OTC complex

VAS: visual analogue scale.

* Equivocal categorisation.
Homeopathy

3
 Model validity of randomised placebo-controlled trials
RT Mathie et al
4
For each paper, three members of our group were
randomly allocated to judge MVHT; no individual was
allocated to assess any of his/her own published papers,
thus removing any potential conflict of interest. The pa-
pers’ authorship was not anonymised, and assessors were
asked to read the Results and Discussion sections only if
essential for clarifying relevant methodological details.
On a dedicated Excel spreadsheet, each assessor recorded
a judgment of ‘Yes’, ‘Unclear’ or ‘No’ to each domain’s
question. For the domain III (Practitioner) criterion, we
added clarification that, for these non-individualised
RCTs, relevant details of the practitioners was neces-
sarily stated in the paper only if focusing its MV assess-
ment solely on homeopathic experience (i.e. if the
assessment was not being based, alternatively, on citation
of a suitable literature source that informed the choice of
the medicine(s) used) e see Appendix 2. The assessors
sent their independent reports, by e-mail, to the study
coordinator (RTM), who collected the findings and ana-
lysed the data.
In an RCT for which no domain was judged ‘No’ by any
independent assessor, and if no more than two domains
presented minor discrepancies per domain (e.g. two ‘Yes’
and one ‘Unclear’ judgment), the recorded overall
MVHT was accepted by default, using the majority vote
for the domains indicated. For an RCT in which there
were widely discrepant assessments in at least one domain
(e.g. two ‘Yes’ and one ‘No’ judgment), or in which there
were minor discrepancies in at least three domains, the
final assessment per domain was reached through
consensus discussion (via e-mail), arbitrated if necessary
by the study coordinator.
Overall MVHT ratings and classifications
We rated MVHT for each trial across all six domains and
using the following nomenclature7:
A = ‘Yes’, as above, in all six domains.
Bx = ‘Unclear’, as above, in x domains; ‘Yes’ in all other
domains.
Cy.x = ‘No’, as above, in y domains; ‘Unclear’ in x do-
mains; ‘Yes’ in all other domains.
As in the Cochrane approach to risk of bias,36 an overall
classification approach was used to ensure that the most
important aspects of study quality prevailed in assessment.
An ‘A’-rated trial was automatically designated ‘accept-
able MVHT’. A ‘B1’-rated trial was regarded as having
acceptable MVHT if it was judged ‘Yes’ for each of do-
mains I, II, IV and V and ‘Unclear’ for domain III or
domain VI: in tabulations and text below, this rating is
shown as ‘B1*’. Other ‘B’-rated trials were designated ‘un-
certain MVHT’. A ‘C’-rated trial was designated ‘inade-
quate MVHT’.
Table 2 Frequency of unanimous assessments per domain in first (independent) round of assessment of 26 eligible trials
Domain I II
Rationale Principles
Frequency 10/26 7/26
(38.5%) (26.9%)
Homeopathy
Please cite this article in press as: Mathie RT, et al., Model validity of randomised placebo-controlled trials of non-individualised homeopathic treatment,
Homeopathy (2017), http://dx.doi.org/10.1016/j.homp.2017.07.003
Merging MVHT and risk of bias into overall quality
designation
Risk of bias, including the description reliable evidence
as applicable, was attributed per trial as previously re-
ported.4 In brief, a trial was designated ‘low’, ‘uncertain’
or ‘high’ risk of bias, based on the standard seven Cochrane
assessment domains (Appendix 3): reliable evidence was
attributed to a low risk-of-bias trial or to one whose uncer-
tainty in risk of bias was restricted to a single pre-specified
domain; all other trials were rated non-reliable evidence.9
For each of the 26 trials, our separate ratings for risk of
bias36 and for MVHT were merged to obtain a single over-
all quality designation (‘high’, ‘moderate’ or ‘low’ qual-
ity), based on the GRADE principle of ‘downgrading’37
e Appendix 4. Thus, for example, a trial attained an overall
designation of ‘high quality’ if it had no important defi-
ciency in either MVHT or risk of bias rating.
Results
Assessment ratings per domain
Over all 26 trials (156 domains of assessment in total),
there were 78 domains (50% of the total number) that
achieved unanimous votes (Table 2) during independent
assessment; unanimity was least evident for domain II.
Consensus discussion took place for each of the remaining
78 domains.
In some cases, application of the MVHT method was
distinctly challenging: this occurred notably for domain I
in trials that involved the use of Arnica for running-
associated muscle soreness or for some complex medicinal
products such as Traumeel. Follow-up discussion on these
matters concluded that both categories typically comprised
acceptable MVHT, though consensus was not achieved for
domain I in two of the trials that involved a complex medic-
inal product (A120 Singer29; A292 Malapane22): the ques-
tion was whether one or more of the homeopathic
ingredients could reasonably be used for the treated condi-
tion or specified symptoms. It was agreed that ‘isopathy’
was an equivocal categorisation for the two Oscillococci-
num trials (A64 Ferley15; A104 Papp26).
MVHT rating and classification per trial
Table 3 details the final MVHT assessments, by domain,
per trial. For three of the trials, unanimity was achieved for
all six domains without the need for consensus discussion
(see Table 3). Overall, deficiencies of MVHT were most
frequently recorded for domains I (Rationale), II (Princi-
ples), III (Practitioner) and V (Outcome sensitivity).
Seven trials were rated ‘A’, and were thus classified
‘acceptable MVHT’; two ‘B1*’ trials (i.e. ‘Unclear’
for domain III or domain VI only) were also classified
‘acceptable MVHT’, giving a total of nine RCTs in this
III IV V VI Total
Practitioner Outcome Sensitivity Follow-up
11/26 18/26 13/26 19/26 78/156
(42.3%) (69.2%) (50.0%) (73.1%) (50.0%)
 Model validity of randomised placebo-controlled trials
RT Mathie et al
5
Table 3 Consensus assessments per domain, with overall MVHT rating and classification per trial

No. First author Domain of assessment No. of ‘Y’ No. of ‘U’ No. of ‘N’ MVHT rating Classification

I II III IV V VI

A272 Colau Y Y Y Y Y Y 6 0 0 A Acceptable MVHT

A67 Frass* Y Y Y Y Y Y 6 0 0 A
A93 Lewith Y Y Y Y Y Y 6 0 0 A
A112 Reilly Y Y Y Y Y Y 6 0 0 A
A123 Taylor Y Y Y Y Y Y 6 0 0 A
A126 Tveiten Y Y Y Y Y Y 6 0 0 A
A292 Malapane Yy Y Y Y Y Y 6 0 0 A
A48 Balzarini* Y Y U Y Y Y 5 1 0 B1*
A135 Wiesenauer Y Y U Y Y Y 5 1 0 B1*
A62 Diefenbach Y Y Y Y U Y 5 1 0 B1 Uncertain MVHT
A83 Kaziro* U Y Y Y Y Y 5 1 0 B1
A125 Tveiten Y Y Y Y U Y 5 1 0 B1
A128 Vickers Y Y Y Y U Y 5 1 0 B1
A290 Naidoo U Y Y Y Y Y 5 1 0 B1
A75 GRECHO U Y U Y Y Y 4 2 0 B2
A104 Papp U U Y Y Y Y 4 2 0 B2
A137 Zabolotnyi U U Y Y Y Y 4 2 0 B2
A64 Ferley U U U Y Y Y 3 3 0 B3
A79 Hofmeyr Y Y U Y U U 3 3 0 B3
A100 Oberbaum Y N Y Y Y Y 5 0 1 C1.0 Inadequate MVHT
A120 Singer Yy N U Y Y Y 4 1 1 C1.1
A61 Cornu U N Y Y U Y 3 2 1 C1.2
A105 Paris Y Y Y U N U 3 2 1 C1.2
A103 Padilha U U U Y U N 1 4 1 C1.4
A92 Leaman U N U N Y Y 2 2 2 C2.2
A47 Baker U U N U U N 0 4 2 C2.4
No. of ‘Y’ per domain 16 17 17 23 18 22
No. of ‘U’ per domain 10 5 8 2 7 2
No. of ‘N’ per domain 0 4 1 1 1 2

* Unanimity achieved for each domain without need for consensus discussion.
y
Majority vote: consensus not achieved among assessors.

classification. The remaining five ‘B1’ trials (with ‘Un-
clear’ for domain I or domain V), together with five that
were rated ‘B2’ or lower, made a total of ten classified as
‘uncertain’ MVHT. Seven trials were rated ‘C’, and thus
classified ‘inadequate’ MVHT. There were no obvious dif-
ferences in MVHT e either by domain or by overall
rating e depending on the type of non-individualised
homeopathy (single, combination or complex medicine,
or isopathy).
Overall quality designation
Only one of the three trials previously rated as reliable
evidence had acceptable MVHT (A272 Colau12): this sin-
gle trial was therefore designated ‘high quality’ overall;
the other two trials regarded as reliable evidence (A103 Pa-
dilha25; A120 Singer29) had inadequate MVHT, and so
were down-graded to ‘low quality’ overall (Table 4). Of
the remaining 23 trials (i.e. those with non-reliable evi-
dence), eight had acceptable MVHT, 10 had uncertain
MVHT, and five had unacceptable MVHT. The 26 trials
were thus designated overall as: ‘high quality’ (N = 1);
‘moderate quality’ (N = 18); ‘low quality’ (N = 7).
Table 5 lists the overall quality designation of each trial
by the clinical condition it investigated (as previously char-
acterised7). For clinical conditions in which there have
been at least two RCTs without high risk of bias, the overall
quality designation was ‘moderate’ for allergic asthma,
muscle soreness and influenza, and ‘moderate’/‘low’ in
post-operative pain. The clinical category Obstetrics & Gy-
naecology had the highest overall quality designation of
‘moderate’/‘high’, reflecting the sole high-quality trial’s
topic of menopausal syndrome. The overall quality was
‘moderate’ within four categories (Allergy & Asthma,
Ear Nose & Throat, Musculoskeletal, Respiratory Infec-
tion), and in five categories it was ‘low’ or ‘moderate’/
‘low’.
Discussion
This study represents the second practical application of
our MVHT assessment method. We have reconfirmed that,
in general, there is no ‘undue difficulty in the ability of
three specialists to achieve a reasonable consensus agree-
ment in judging the relevant issues’5: for 50% of the 156
domains in total, independent assessment attained unanim-
ity without any need for consensus discussion. However,
these non-individualised trials did prove more challenging
to assess for MVHT than their counterparts in individual-
ised homeopathy.5 Extra effort to reach consensus for
domain I (rationale), for example, was sometimes needed
for those trials that used a complex homeopathic prepara-
tion.
In the current study, some assessors needed clarification
about the correct interpretation of the criteria for domain I,
and specifically about the meaning of ‘Whether a substan-
tial number of experienced homeopaths would support the
choice of this intervention for this type of patient’: ‘this

Homeopathy
Please cite this article in press as: Mathie RT, et al., Model validity of randomised placebo-controlled trials of non-individualised homeopathic treatment,
Homeopathy (2017), http://dx.doi.org/10.1016/j.homp.2017.07.003
 Model validity of randomised placebo-controlled trials
RT Mathie et al
6
Table 4 Ordering of 26 trials by overall quality designation

Ref. First author MVHT Risk of bias9 Downgrading Overall designation

A272 Colau Acceptable Low risk* 0 High quality

A67 Frass Acceptable Uncertain risk 1 Moderate quality
A93 Lewith Acceptable Uncertain risk 1 Moderate quality
A112 Reilly Acceptable Uncertain risk 1 Moderate quality
A123 Taylor Acceptable Uncertain risk 1 Moderate quality
A126 Tveiten Acceptable Uncertain risk 1 Moderate quality
A292 Malapane Acceptable Uncertain risk 1 Moderate quality
A48 Balzarini Acceptable Uncertain risk 1 Moderate quality
A135 Wiesenauer Acceptable Uncertain risk 1 Moderate quality
A62 Diefenbach Uncertain Uncertain risk 1 Moderate quality
A83 Kaziro Uncertain Uncertain risk 1 Moderate quality
A125 Tveiten Uncertain Uncertain risk 1 Moderate quality
A128 Vickers Uncertain Uncertain risk 1 Moderate quality
A290 Naidoo Uncertain Uncertain risk 1 Moderate quality
A75 GRECHO Uncertain Uncertain risk 1 Moderate quality
A104 Papp Uncertain Uncertain risk 1 Moderate quality
A137 Zabolotnyi Uncertain Uncertain risk 1 Moderate quality
A64 Ferley Uncertain Uncertain risk 1 Moderate quality
A79 Hofmeyr Uncertain Uncertain risk 1 Moderate quality
A103 Padilha Inadequate Low risk* 2 Low quality
A120 Singer Inadequate Uncertain risk* 2 Low quality
A100 Oberbaum Inadequate Uncertain risk 2 Low quality
A61 Cornu Inadequate Uncertain risk 2 Low quality
A105 Paris Inadequate Uncertain risk 2 Low quality
A92 Leaman Inadequate Uncertain risk 2 Low quality
A47 Baker Inadequate Uncertain risk 2 Low quality

* Reliable evidence (regarding risk of bias).

intervention’ was agreed by all to mean ‘this specific kind
of homeopathy’, as was its original intention. Likewise, for
domain III, it was initially unclear to some assessors that a
paper’s citation of a suitable reference e e.g. materia
medica or repertory e was in fact sufficient to accept there
had been ‘experienced homeopathic input’ to the design of
a non-individualised trial. The descriptive details for
domain III (Practitioner) were updated accordingly
(Appendix 2).
Concern was expressed within our group about the po-
tential redundancy of domain III in the context of these
non-individualised trials: domain I (rationale) and domain
II (principles) might be seen as sufficient, since practi-
tioners are not directly involved in prescribing the homeo-
pathic treatment. However, assessment of each MV domain
is an independent task. Thus evidence of satisfactory ho-
meopathic rationale and principles (domains I and II)
does not necessarily guarantee that a trial actually had

Table 5 Overall quality designation of trials by category of clinical condition

Category of clinical condition Clinical condition Ref. First author Year Overall quality designation

Allergy & asthma Allergic asthma A93 Lewith 2002 Moderate quality
Allergic asthma A112 Reilly 1994 Moderate quality
Allergic skin reaction A290 Naidoo 2013 Moderate quality
Perennial allergic rhinitis A123 Taylor 2000 Moderate quality
Seasonal allergic rhinitis A135 Wiesenauer 1995 Moderate quality
Dermatology Minor burns A92 Leaman 1989 Low quality
Radiodermatitis A48 Balzarini 2000 Moderate quality
Ear, nose & throat Sinusitis A137 Zabolotnyi 2007 Moderate quality
Tonsillitis A292 Malapane 2014 Moderate quality
Mental disorder Anxiety A47 Baker 2003 Low quality
Musculoskeletal Muscle soreness A125 Tveiten 1991 Moderate quality
Muscle soreness A126 Tveiten 1998 Moderate quality
Muscle soreness A128 Vickers 1998 Moderate quality
Obstetrics & gynaecology Menopausal syndrome A272 Colau 2012 High quality
Postpartum pain A79 Hofmeyr 1990 Moderate quality
Oral/dental Stomatitis A100 Oberbaum 2001 Low quality
Respiratory infection Bronchitis A62 Diefenbach 1997 Moderate quality
Influenza A64 Ferley 1989 Moderate quality
Influenza A104 Papp 1998 Moderate quality
Tracheal secretions A67 Frass 2005 Moderate quality
Surgery & anaesthesiology Post-operative analgesic intake A105 Paris 2008 Low quality
Post-operative bleeding A61 Cornu 2010 Low quality
Post-operative ileus A75 GRECHO 1989 Moderate quality
Post-operative pain A120 Singer 2010 Low quality
Post-operative pain/swelling A83 Kaziro 1984 Moderate quality
Toxicology Lead poisoning A103 Padilha 2011 Low quality

Homeopathy
Please cite this article in press as: Mathie RT, et al., Model validity of randomised placebo-controlled trials of non-individualised homeopathic treatment,
Homeopathy (2017), http://dx.doi.org/10.1016/j.homp.2017.07.003

sufficient homeopathic input (domain III). In fact, only
three trials were judged ‘Yes’e‘Yes’e‘Unclear’ in the first
three domains. One of those (A48 Balzarini11) did not
require any consensus discussion: its deficiency was that
treatment involved homeopathic Belladonna and X-ray,
which was not justified in the paper. In the other two papers
(A79 Hofmeyr18; A135 Wiesenauer34), there was insuffi-
cient information to permit judgment of ‘Yes’ or ‘No’
for domain III: neither paper gave adequate references
to assure us that the authors or practitioners were aware
of a relevant literature to inform their selection of the
homeopathic treatment used. Explicit assurance on this
point is the purpose of domain III. Importantly, this is
one of the two domains where an ‘Unclear’ vote does not
preclude ‘B1*’ classification (‘acceptable MVHT’).5,6
The Balzarini and Wiesenauer papers are both in that
‘B1*’ category; the Hofmeyr paper included other
concerns, leading to a lower MVHT classification.
Perceived deficiencies in the current trials regarding
domains I (rationale) and II (principles) may be explained
by the nature of non-individualised homeopathy in the
RCT context. Both for single and for combination/com-
plex medicines, two questions often arose: (a) Was/were
the medicine/s used a reasonable choice for a given symp-
tom or set of symptoms? (b) Was every trial participant
likely to manifest the symptom/s matched to those medi-
cine/s? The clinical approach in the current trials does not
require accurate matching between the symptoms caused
by a homeopathic medicine and those displayed by a pa-
tient, as in individualised homeopathy; instead, a pre-
selected medicine is applied to the typical symptoms of
a clinical condition. In the normal practice of ‘clinical ho-
meopathy’, a medicine is prescribed on the basis of a pa-
tient’s manifest symptoms; the choice of a medicine
based on typical symptoms for an RCT of non-
individualised homeopathy may therefore not equate to
that normal practice. If a trial investigates a homeopathic
product that does not match symptoms manifested by all
participants, it can be expected to produce sub-optimal re-
sults: it is important that this possibility is understood by
consumers using such products in self-treatment and by
medical practitioners not trained in homeopathy. The
problem of sub-optimal matching of the medicine to the
patient is more explicit in RCTs of non-individualised ho-
meopathy than in those of individualised homeopathy.
Interestingly, the main outcome measure was typically
deemed able to reflect the effect expected of the given ho-
meopathic intervention (domain IV), as was the case also
for the individualised trials.5
We classified just nine of the 26 eligible trials (34%) as
‘acceptable MVHT’ overall, a lesser proportion than in the
corresponding assessment of individualised homeopathy
trials (59%).5 Moreover, these 26 are the non-
individualised trials (from 75 in total) judged to be without
overt risk of bias: the MVHT of the other 49 remains un-
known at present. The MVHT of 10 of the 26 trials was
classified as ‘uncertain’, again highlighting the need for
original authors to report sufficient details in their publica-
tions to enable a complete and unequivocal appraisal.
Please cite this article in press as: Mathie RT, et al., Model validity of randomised placebo-controlled trials of non-individualised homeopathic treatment,
Homeopathy (2017), http://dx.doi.org/10.1016/j.homp.2017.07.003
Model validity of randomised placebo-controlled trials
RT Mathie et al
7
Although 19 trials were judged as not overtly deficient
(i.e. ‘acceptable’ or ‘uncertain’) in MVHT, just one of
those (A272 Colau12) has been identified ‘reliable evi-
dence’9 and thus able to be designated ‘high quality’ over-
all. Two trials rated as ‘reliable evidence’ by risk-of-bias
assessment9 (A103 Padilha25; A120 Singer29) were down-
graded to ‘low quality’ overall due to inadequate MVHT,
showing that the best conventionally viewed evidence is
not necessarily also the best evidence from a homeopathy
perspective. The remaining 56 of the 75 trials that were
eligible for the overarching systematic review (i.e. the
seven clearly MV-deficient RCTs plus the 49 that were
ineligible for this MV assessment due to their high risk of
bias) can, at best, be designated ‘low quality’ overall. Sig-
nificant quality issues also pervade the RCT literature in
conventional medicine,38 though, to our knowledge, MV
has not been addressed in that research context.
In relative terms, RCTs of individualised homeopathy
seem to possess a somewhat higher standard of MVHT,
though only three such trials have been designated ‘high
quality’ overall.6 In meta-analysis, data from these three
high-quality trials suggested that the medicines prescribed
in individualised homeopathy have small effects that are
greater than placebo effects. Our meta-analysis results for
the current non-individualised trials are reported in a sepa-
rate communication9: the pooled data reported for ‘reliable
evidence’ should now be viewed from the perspective of
the present MVHT findings, in which a singular trial
(A272 Colau12) comprises the high-quality evidence over-
all. The present paper also reveals the clinical research cat-
egories that contain the least deficient quality of evidence:
Allergy & Asthma; Ear Nose & Throat; Musculoskeletal;
Obstetrics & Gynaecology; Respiratory Infection.
Conclusions
Of the 26 RCTs of non-individualised homeopathic
treatment that were previously judged ‘not at high risk of
bias’, nine have also now been rated as ‘acceptable
MVHT’. One of those nine trials had been rated ‘reliable
evidence’ and is thus the sole trial that can be designated
‘high quality’ overall (‘acceptable MVHT’ and ‘reliable
evidence’). That singular paper, a study on menopausal
syndrome, is currently the only reported RCT that repre-
sents best therapeutic practice as well as unbiased evidence
in non-individualised homeopathy. Together with minimis-
ing risk of bias, new placebo-controlled RCTs of non-
individualised homeopathy must aim to maximise
MVHT, and with sufficient clarity of reporting to assist
its independent assessment in systematic review.
Sources of funding
None.
Declaration of interest
Three of the authors provide, or have provided in recent
years, consultancy services to Boiron (RvH, PF), Deutsche
Homeopathy
 Model validity of randomised placebo-controlled trials
RT Mathie et al
8
Hom€ oopathie-Union (RvH, RTM, PF) and Heel GmbH
(RvH), each of which manufacture some of the products
studied in this review.
Authors’contributions
All authors contributed to the conception and design of
the study, to interpreting the data, and to drafting the
article, revising it critically for important intellectual con-
tent. Each author contributed substantially to the original
data acquisition, and approved the submitted version of
the article.
Supplementary data
Supplementary data (Appendices 1e4) related to this
article can be found at http://dx.doi.org/10.1016/j.homp.
2017.07.003.
References
1 Verhoef MJ, Lewith G, Ritenbaugh C, Boon H, Fleishman S, Leis A.
Complementary and alternative whole systems research: beyond
identification of inadequacies of the RCT. Complement Ther Med
2005; 13: 206e212.
2 Bornh€oft G, Maxion-Bergemann S, Wolf U, et al. Checklist for the
qualitative evaluation of clinical studies with particular focus on
external validity and model validity. BMC Med Res Methodol
2006; 6: 56.
3 Mathie RT, Roniger H, Van Wassenhoven M, et al. Method for
appraising model validity of randomised controlled trials of homeo-
pathic treatment: multi-rater concordance study. BMC Med Res
Methodol 2012; 12: 49.
4 Mathie RT, Lloyd SM, Legg LA, et al. Randomised placebo-
controlled trials of individualised homeopathic treatment: system-
atic review and meta-analysis. Syst Rev 2014; 3: 142.
5 Mathie RT, Van Wassenhoven M, Jacobs J, et al. Model validity of
randomised placebo-controlled trials of individualised homeopathic
treatment. Homeopathy 2015; 104: 164e169.
6 Mathie RT, Van Wassenhoven M, Jacobs J, et al. Model validity and
risk of bias in randomised placebo-controlled trials of individual-
ised homeopathic treatment. Complement Ther Med 2016; 25:
120e125.
7 Mathie RT, Hacke D, Clausen J, Nicolai T, Riley DS, Fisher P.
Randomised controlled trials of homeopathy in humans: character-
ising the research journal literature for systematic review. Homeop-
athy 2013; 102: 3e24.
8 Swayne J. International dictionary of homeopathy. Edinburgh:
Churchill Livingstone, 2000.
9 Mathie RT, Ramparsad N, Legg LA, et al. Randomised, double-
blind, placebo-controlled trials of non-individualised homeo-
pathic treatment: systematic review and meta-analysis. Syst
Rev 2017; 6: 63.
10 Baker DG, Myers SP, Howden I, Brooks L. The effects of homeo-
pathic Argentum nitricum on test anxiety. Complement Ther Med
2003; 11: 65e71 [Systematic review programme number A47].
11 Balzarini A, Felisi E, Martini A, De Conno F. Efficacy of homeo-
pathic treatment of skin reactions during radiotherapy for breast
cancer: a randomized, double-blind clinical trial. Br Homeopath J
2000; 89: 8e12 [A48].
12 Colau JC, Vincent S, Marijnen P, Allaert FA. Efficacy of a non-
hormonal treatment, BRN-01, on menopausal hot flashes: a multi-
center, randomized, double-blind, placebo-controlled trial. Drugs
R D 2012; 12: 107e119 [A272].
Homeopathy
Please cite this article in press as: Mathie RT, et al., Model validity of randomised placebo-controlled trials of non-individualised homeopathic treatment,
Homeopathy (2017), http://dx.doi.org/10.1016/j.homp.2017.07.003
13 Cornu C, Joseph P, Gaillard S, et al. No effect of a homeopathic
combination of Arnica montana and Bryonia alba on bleeding,
inflammation, and ischaemia after aortic valve surgery. Br J Clin
Pharmacol 2010; 69: 136e142 [A61].
14 Diefenbach M, Schilken J, Steiner G, Becker HJ. Hom€ oopathische
Therapie bei Erkrankungen der Atemwege. Auswertung einer klini-
schen Studie bei 258 Patienten [Homeopathic therapy in respiratory
tract diseases. Evaluation of a clinical study in 258 patients]. Z All-
gemeinmed 1997; 73: 308e314 [A62].
15 Ferley JP, Zmirou D, D’Adhemar D, Balducci F. A controlled eval-
uation of a homeopathic preparation in the treatment of influenza
like syndromes. Br J Clin Pharmacol 1989; 27: 329e335 [A64].
16 Frass M, Dielacher C, Linkesch M, et al. Influence of potassium di-
chromate on tracheal secretions in critically ill patients. Chest 2005;
127: 936e941 [A67].
17 GRECHO (Groupe de Recherches et d’Essais Cliniques en
Hom
eopathie). Evaluation de deux produits hom
eopathiques sur
la reprise du transit apres chirurgie digestive e Un essai contr^ ol
e
multicentrique [Evaluation of the effects of two homeopathic prep-
arations on the resumption of intestinal peristalsis after digestive
tract surgery e a multicentre controlled trial]. Presse Med 1989;
18: 59e62 [A75].
18 Hofmeyr GJ, Piccioni V, Blauhof P. Postpartum homeopathic
Arnica montana: a potency-finding pilot study. Br J Clin Pract
1990; 44: 619e621 [A79].
19 Kaziro GS. Metronidazole (Flagyl) and Arnica montana in the pre-
vention of post-surgical complications, a comparative placebo
controlled clinical trial. Br J Oral Maxillofac Surg 1984; 22:
42e49 [A83].
20 Leaman AM, Gorman D. Cantharis in the early treatment of minor
burns. Arch Emerg Med 1989; 6: 259e261 [A92].
21 Lewith G, Watkins AD, Hyland ME. Use of ultramolecular po-
tencies of allergen to treat asthmatic people allergic to house dust
mite: double blind randomized controlled clinical trial. Br Med J
2002; 324: 520e523 [A93].
22 Malapane E, Solomon EM, Pellow J. Efficacy of a homeopathic
complex on acute viral tonsillitis. J Altern Complement Med
2014; 20: 868e873 [A292].
23 Naidoo P, Pellow J. A randomized placebo-controlled pilot study of
Cat saliva 9cH and Histaminum 9cH in cat allergic adults. Homeop-
athy 2013; 102: 123e129 [A290].
24 Oberbaum M, Yaniv I, Ben-Gal Y, et al. A randomized,
controlled clinical trial of the homeopathic medication TRAU-
MEEL SÒ in the treatment of chemotherapy-induced stomatitis
in children undergoing stem cell transplantation. Cancer 2001;
92: 684e690 [A100].
25 Padilha RQ, Riera R, Atallah AN.
Homeopathic Plumbum metalli-
cum for lead poisoning: a randomized clinical trial. Homeopathy
2011; 100: 116e121 [A103].
26 Papp R, Schuback G, Beck E, et al. OscillococcinumR in patients
with influenza-like syndromes: a placebo-controlled double-blind
evaluation. Br Homoeopath J 1998; 87: 69e76 [A104].
27 Paris A, Gonnet N, Chaussard C, et al. Effect of homeopathy on
analgesic intake following knee ligament reconstruction: a phase
III monocentre randomized placebo controlled study. Br J Clin
Pharmacol 2008; 65: 180e187 [A105].
28 Reilly D, Taylor MA, Beattie NGM, et al. Is evidence for homeop-
athy reproducible? Lancet 1994; 344: 1601e1606 [A112].
29 Singer SR, Amit-Kohn M, Weiss S, et al. Traumeel SÒ for pain re-
lief following hallux valgus surgery: a randomized controlled trial.
BMC Clin Pharmacol 2010; 10: 9 [A120].
30 Taylor MA, Reilly D, Llewellyn-Jones RH, McSharry C,
Aitchison TC. Randomised controlled trial of homeopathy versus
placebo in perennial allergic rhinitis with overview of four trial se-
ries. Br Med J 2000; 321: 471e476 [A123].
31 Tveiten D, Bruseth S, Borchgrevink CF, Løhne K. Effekt av Arnica
D 30 ved hard fysisk anstrengelse. En doppeltblind randomisert

undersøkelse under Oslo maraton 1990 [Effect of Arnica D30 on
hard physical exertion. A double-blind randomized trial during
the 1990 Oslo Marathon]. Tidsskr Nor Laegeforen 1991; 111:
3630e3631 [A125].
32 Tveiten D, Bruseth S, Borchgrevink CF, Norseth J. Effects of the ho-
meopathic remedy Arnica D30 on marathon runners: a randomized,
double-blind study during the 1995 Oslo Marathon. Complement
Ther Med 1998; 6: 71e74 [A126].
33 Vickers AJ, Fisher P, Smith C, Wyllie SE, Rees R. Homeopathic
Arnica 30X is ineffective for muscle soreness after long-distance
running: a randomized, double-blind, placebo-controlled trial.
Clin J Pain 1998; 14: 227e231 [A128].
34 Wiesenauer M, L€udtke R. The treatment of pollinosis with Galphi-
mia glauca D4 e a randomized placebo-controlled double-blind
clinical trial. Phytomedicine 1995; 2: 3e6 [A135].
Please cite this article in press as: Mathie RT, et al., Model validity of randomised placebo-controlled trials of non-individualised homeopathic treatment,
Homeopathy (2017), http://dx.doi.org/10.1016/j.homp.2017.07.003
Model validity of randomised placebo-controlled trials
RT Mathie et al
9
35 Zabolotnyi DI, Kneis KC, Richardson A, et al. Efficacy of a complex
homeopathic medication (Sinfrontal) in patients with acute maxillary
sinusitis: a prospective, randomized, double-blind, placebo-controlled,
multicenter clinical trial. Explore (NY) 2007; 3: 98e109 [A137].
36 Higgins JPT, Altman DG (eds). Chapter 8: Assessing risk of bias in
included studies. In: Higgins JPT, Green S (eds). Cochrane Hand-
book for Systematic Reviews of Interventions; Version 5.1.0. The
Cochrane Collaboration, 2011.
37 Sch€unemann HJ, Oxman AD, Vist GE, et al. Chapter 12: Interpret-
ing results and drawing conclusions. In: Higgins JPT, Green S (eds).
Cochrane Handbook for Systematic Reviews of Interventions;
Version 5.1.0. The Cochrane Collaboration, 2011.
38 El Dib RP, Atallah AN,
Andriolo RB. Mapping the Cochrane evi-
dence for decision making in health care. J Eval Clin Pract 2007;
13: 689e692.
Homeopathy