Research results published in the Lancet in 2005¹ on the lack of effectiveness of homeopathy are less conclusive than reported. Two new studies find that the conclusions of this research are not reliable.

Firstly, this is the opinion of R. Lüdtke and A. L. B. Rutten² who have made a thorough investigation into the set-up of the 21 placebo-controlled clinical trials investigated by Shang et al. By analysing the exact data used by Shang, the authors unearthed new information which undermines the main conclusion drawn by Shang and co-authors, that ‘…the clinical effects of homeopathy are placebo effects’.

This broader look at the meta-analyses by Shang et al can be summarised as follows:
Lüdtke and Rutten took the exact documentation used in the Shang study and newly arranged so called ‘subsets’ of the criteria used for evaluating trials. From that, differing results were determined. Among such criteria were the number of analysed patients (sample size), treated conditions, quality of statistical analysis, and the type of homeopathic treatment such as classical homeopathy with a single remedy according to individual patient evaluation, fixed homeopathic treatment by the same single agent, and treatment using complex homeopathic products. Attention was also paid to homeopathic dilutions. They were divided in those of ultra molecular range and those of lower dilution levels.  

What Lüdtke and Rutten now found were significant effects in favour of homeopathy, for instance when low dilution treatment was compared. Significant positive data were obtained when a problematic muscle soreness trial³ was omitted or when the statistical threshold level in the ‘sample size’ was changed.

In looking at varied subsets in Shang’s 21 high quality trials which compared homeopathic medicines with placebo, new meta-analyses data could be obtained. Consequently, Lüdtke and Rutten also found several possibilities of how evaluators could – intentionally or unintentionally – influence and interpret the study result by determining for instance threshold levels and other evaluation cut-off points. They also underlined that meta-analytical results would suffer from the high heterogeneity in the study material, even if every study by itself could be considered of highest quality. Additionally, the selection criteria for the study material in a meta-analysis can depend on the intentions and beliefs of the analyst resulting in considerable areas of subjectivity.

As a bottom line, using placebo-controlled trials in homeopathy would intrinsically be problematic because of the inevitable heterogeneity of their results. As a consequence, the scientific hypothesis that ‘homeopathic effects are placebo’ would not make sense. From the authors’ point of view, the clinical trial hypothesis in homeopathy should be more specific and include both a specific definition of the homeopathic intervention and a clear definition of the disease/health condition.

In the second study⁴ Rutten and Stolper find that a reanalysis of Shang's post-publication data also did not support the conclusion that homeopathy is a placebo effect. It found that the conclusion that homeopathy is and that conventional is not a placebo effect was not based on comparative analysis.

Due to the fact that Shang and colleagues drew conclusions from a meta set-up of highly heterogenic trials these conclusions must now be considered much less reliable and definite than they have been reported and discussed. 

Max Daege
ECHAMP President

---

¹ Shang A, Huwiler-Müntener K, Nartey L, Jüni P, Dörig S, Sterne JA, Are the clinical effects of homoeopathy placebo effects? Comparative study of placebo-controlled trials of homoeopathy and allopathy. Lancet, 2005 Aug 27-Sep 2;366(9487):726-32
² Lüdtke R, Rutten ALB. The conclusions on the effectiveness of homeopathy highly depend on the set of analyzed trials. J of Clin Epidemiology 2008
³ Vickers et al. Homeopathy for delayed onset muscle soreness – a randomised double blind, placebo-controlled trial. Brit J Sports Med 1997;31:304-7.
⁴ Rutten ALB, Stolper CF. The 2005 meta-analysis of homeopathy: the importance of post-publication data. Homeopathy 2008.  doi:10.1016/j.homp.2008.09.008