hould we put more effort in proving or in improving homeopathy? Now we have plenty of RCTs indicating an effect incompatible with the placebo-hypothesis, it appears that sceptics find ways to deny this evidence. They state that the quality is not good, but the quality of conventional trials is worse! They state that RCT evidence should not be taken serious because the mechanism of action is implausible. This is circular reasoning: RCT is the gold standard to prove that the theoretical impossibility that homeopathy can work should be questioned.

Others state that there is not enough evidence for homeopathy in separate indications. This can be argued, but that is not the point; eventually we would want to know the chance of effect in the individual patient. Dividing the whole population in separate indications is a surrogate for personalised medicine. Neither homeopathy, nor conventional medicine can provide RCT proof on an individual level.

Homeopathy is, in fact, working in the direction of personalised medicine by consensus about symptoms indicating specific medicines; these symptoms are prognostic factors. Homeopathy is a prototype for personalised medicine since two centuries. For scientific development of personalised homeopathy we need prognosis research, more specific prognostic factor research.

There are several for research to improve homeopathy, qualitative or quantitative methods, or a combination of both. Our profession is already using consensus methods for qualitative evaluation of individual cases. Consensus methods can be formalised to get better validity. An important point is causal relationship between medicine and effect. Even in individual cases we can establish some certainty that an effect is due to the prescribed medicine. The problem is comparable to establishing a causal relationship in case of side effects of medicines. For this purpose we can use the Naranjo algorithm.

Cases where the homeopathic medicine had a probable effect can be used for case analysis (qualitative), but also for Likelihood Ratio (LR) calculations or multivariate analysis. But the first step is recording of our cases in computer databases. You can read more about this by clicking here.