Explanations on the methodology of placebo-controlled randomized clinical homeopathy trials and their meta-analyses

to the document "Scientific assessment of the motion V-01"

Homeopathic treatment can be individualized or non-individualized: Individualized homeopathy involves a comprehensive interview with the patients regarding the type of symptoms of their disease, other symptoms, etc., in order to select a suitable "individualized" homeopathic treatment. Non-individualized homeopathy follows other principles and requires less time.

In a randomized, placebo-controlled clinical homeopathy trial, the patient is randomly assigned to a treatment or control group, according to specific procedures. The random treatment allocation is intended to distribute other factors, which, in addition to the treatment, could influence the study outcome (e.g. the duration and severity of the disease to be treated) equally between both groups.

The treatment group receives a homeopathic remedy, the control group receives a placebo substance which looks and tastes as similar as possible to the homeopathic remedy etc. The purpose of the placebo treatment is that the patients should not know which treatment they are receiving, in order that any potential psychological factors which might influence the study results (e.g. the expectation of an improvement) are distributed equally between both groups, as well as other treatment components (e.g. interview with the patient, clinical examination, explanation and advice). For this reason, not only the patients should be "blinded" to whether they are allocated to the homeopathy or the placebo group, but if possible also the physicians, therapists and other people involved in data collection (double blind trial).

Clinical study outcomes are e.g. the number of improved patients or the extent of improvement (according to predefined criteria) in the two groups. The treatment effect of the homeopathic remedy is the difference between the two groups, i.e. the treatment outcome in the homeopathy group minus the outcome in the placebo group, according to specific calculation procedures such as subtraction, division, etc. The outcome comprises a measure of the size of the difference (effect size) and a measure for the statistical significance. A difference or treatment effect is said to be statistically significant if the likelihood of the difference occurring by chance is less than 5% (p<0.05). Smaller as well as larger percentage values are possible. The usual 5% are used in the present text.

In systematic reviews, the outcomes of all retrievable studies on a specific research issue (in our case "randomized, placebo-controlled clinical trials on homeopathic treatment for any disease in humans") are compiled and assessed.

Meta-Analyses are systematic reviews in which the results of the individual trials – large and small trials, trials with statistically significant and non-significant results – are summarized into one quantitative average treatment effect, again with effect size and statistical significance.

In the primary analysis the summarized outcome for all trials included in the meta-analysis is calculated and presented.