#drop-caps-5c0de5f804119 {padding:10px !important;}#drop-caps-5c0de5f804119 {font-size:25px !important;}

tatistics are behind every presciption of a homeopathic medicine. Two important statistical notions are mentioned here: prevalence and variance. To start with prevalence: A symptom is an indication for a specific homeopathic medicine if that symptom occurs frequently in patients that respond well to that specific medicine. The word ‘frequently’ is very important; it makes a difference if a common symptom like headache occurs once in every 10 patients or seven times in every 10 patients.

This is expressed by the statistical term prevalence: once in every 10 patients means prevalence=10%, 7 times in every 10 patients means prevalence=70%. Suppose the prevalence of the symptom headache is 10% in the population responding well to, say, Sulphur, and 40% in the remainder of your practice population. Intuitively you feel that in that case headache is not a symptom that indicates Sulphur. You would prefer a medicine with prevalence of headache of 70%.

This statistical notion is expressed by Bayes’ theorem:

Posterior odds= LR x prior odds
LR=likelihood ratio=(prevalence in target population)/(prevalence in the remainder of the popualtion)
Odds=chance/(1-chance); chance=odds/(1+odds)

in the example above the likelihood ratio of the symptom headache for Sulphur is 10/40=0.25. The likelihood ratio of headache in a medicine with a prevalence of headache of 70% would be 70/40=1.75.

Bayes’ formula indicates that the presence of a symptom increases the chance that a medicine will work if LR>1. So Headache is no indication for Sulphur, but indeed for a medicine that is frequently (>40%) related to headache.

Bayes’ theorem, and thereby prevalence, are in the core of homeopathic prescribing. In our present repertories the absolute occurrence is reason for adding entries, while it should be prevalence and LR>1.


Human observations and symptoms in living systems are no steady items. Two observers interpret the same symptom differently and most symptoms are not constantly the same. If you takeĀ  measurements of, say, blood glucose in a larger number of people, these measurements could be represented by the following graph.

If we look at the symptom ‘Desire for salt’ in a population responding well to the homeopathic medicine Natrium muriaticum (Nat-m.) you will see variance. Some of these patients will have a ‘low’ desire for salt (less than the average in the whole population), some will have an extreme desire for salt, but the average patient responding well to Nat-m. will have a greater desire for salt than average in the whole population. Does the fact that a few patients have an aversion to salt mean that ‘Aversion to salt’ is an indication for Nat-m.? No, this is simply due to statistical variance!

Statistical variance and prevalence are hitherto not considered in the homeopathic materia medica and repertories and that leads to misleading entries in the repertories. This should be mended by prognostic factor research.

Start typing and press Enter to search