Homeopathic medicine is used in diverse countries and regions of the world, and remains popular despite scepticism amongst a majority of scientists concerning its physiological mechanism of action. Given its widespread use, homeopathy is severely under-researched: the total volume of clinical research, published in peer-review journals to the end of 2009, comprises a total of 142 randomised controlled trials (RCTs),¹ a minuscule number compared with that in conventional medicine (estimated at c.500,000 trials). The clinical research literature in homeopathy attracts polarised interpretation by advocates and opponents of the subject. As a contribution to finding ways to move beyond the current impasse of controversy, this brief article summarises the nature of clinical research evidence in general, with specific comments about research in homeopathy and its challenges for the future.

Traditional hierarchy of clinical research evidence
The World Health Organization identifies the following categories of clinical research evidence:

The evaluation of evidence gives the greatest weight to systematic reviews. Such reviews, which might include statistical meta-analysis, are a synthesis of the findings from a number of different RCTs. In an RCT, the therapy under investigation is compared with a control group, which might be well-established treatment, no treatment, different doses of the same treatment, sham or placebo treatment, full-scale treatment, minimal treatment or alternative treatment. RCTs are considered by scientists as “the most rigorous way of determining whether a cause-effect relation exists between treatment and outcome”.² They may be defined as explanatory (tightly defined medical intervention, inclusion criteria for patient entry and clinical outcomes; placebo-controlled) or pragmatic (reflect ‘real world’ conditions in terms of treatment and patients; controlled by other than placebo, OTP). Each type of RCT addresses a different research question: an explanatory trial aims to find out whether a specific treatment is efficacious under ‘experimental’ conditions; a pragmatic trial aims to find out how effective a treatment is in everyday medical practice.³

Other forms of clinical research evidence are also available, but none of those is capable of enabling inferences about cause and effect. Quasi-experimental trials (category II) are those in which comparison of the test intervention is made with a similar, but not randomly selected, cohort of participants. Descriptive studies (category III) usually involve some form of systematic documentation of clinical practice in a particular cohort of patients; this type of study is neither randomised nor controlled in character.  


Clinical research evidence in homeopathy
From three of four systematic reviews of various groups of medical conditions and styles of homeopathy, there has been the qualified conclusion that homeopathic intervention differs from placebo.⁴ Based on findings from 20 condition-specific systematic reviews, there is a broadly equal mix of positive, negative and non-conclusive RCT evidence.⁵ From the original peer-reviewed literature, comprising 120 placebo-controlled RCTs and 22 OTP-controlled RCTs, 44% of trials were reported as positive, 8% as negative, and 48% as non-conclusive.¹ Replication of RCT findings in a given diagnostic area has been obtained infrequently. The most convincing replicated evidence (positive in nature) is in seasonal allergic rhinitis.^1,4

A number of individual reports of category II or category III research have focused on homeopathy for either a particular medical condition or a specified range of medical complaints. Their findings are invariably positive, for example in otitis media in children. ^5,6


Challenges for clinical research in homeopathy
Homeopathy trials are highly heterogeneous in intervention, diagnostic area, controls and outcomes, and so are potentially inappropriate or misleading to summarise in a review or meta-analysis.⁷ Focus on a given medical diagnosis or group of related diagnoses seems preferable.⁴

A key challenge in homeopathy research is how best to reflect individualised treatment of patients, a fundamental attribute of the therapy. As outlined above, explanatory RCTs require homogeneous groups of participants, which may not be consistent with the individualised treatment approach.^8,9 Various strategies to circumvent this concern have been adopted. These include: (a) the use of a simplified research model in which a single, standardised, homeopathic medicine is prescribed for a particular diagnosis; e.g.¹⁰ and (b) ‘double selection’, in which patients with a particular diagnosis and indications for one or a limited range of homeopathic medicines are included in the trial.e.g.¹¹ There is a need, in addition, to enhance the output of pragmatic RCTs, in which individualisation of treatment can more readily be preserved.

Biases inherent in clinical research of non-randomised design probably contribute to their invariably positive findings, and so caution is required in their interpretation. Nevertheless, the results obtained from such study design provide a secure foundation for subsequent RCT investigation.^12,13  


Conclusion
Application of conventional clinical research methods to homeopathy is feasible, and has resulted in a limited but distinct literature. In extending the development of research in this therapeutic area, an awareness of homeopathy-specific issues is required in study design and the interpretation of related data.

Robert T Mathie, PhD, Research Development Adviser at the British Homeopathic Association and Faculty of Homeopathy

Please click here for a brief biography of the author