The Challenge of Building a Bayesian Homeopathic Materia Medica and Repertory from Best Cases' Data

Abstract It is essential to homeopathy that medicines' indicating symptoms and signs are accurately determined. In recent times, it has become apparent that the traditional assessment of indicating symptoms based on their absolute occurrence criteria (attributing a symptom to a medicine when it has been observed in several cases of responders) is flawed. A Bayesian approach using various study methods has been increasingly applied in this field of research. In this publication, the feasibility of building a homeopathic Materia Medica and a Repertory based on best cases' data is explored. Cases' data were extracted from the ‘Best Chronic Homeopathic Cases’ database. A common unification language was agreed, with clearly defined inclusion criteria of symptoms and medicines for analysis. Count, prevalence, likelihood ratio (LR), and cumulative binomial probability were calculated for each symptom–medicine combination, with adjustment for outliers. A Nat-m Bayesian Materia Medica was developed using defined criteria of statistical significance for validating the indicating symptoms. A Bayesian Homeopathic Repertory application (app) was developed. A total of 794 cases, 2,663 symptoms and 101 different medicines were extracted from the database. A total of 794 cases, 587 symptoms and 41 medicines were included for analysis, resulting in a matrix of 24,067 LRs. A total of 102 Nat-m symptoms were validated as indicating the medicine and 8 symptoms as contraindicating it. It is possible to detect and validate accurately the indicating symptoms of homeopathic medicines with a Bayesian assessment of data from the best clinical cases. It is also possible to develop a Bayesian Homeopathic Repertory and a Bayesian Homeopathic Materia Medica from these data. Skilled homeopaths with different cultural backgrounds, experience and preferences are encouraged to engage in this path of research.