Methodological Perspectives on Variability in Homeopathic Practice: Detecting the Signals by Reducing the Noise

Background: The Clificol® COVID-19 Support Project is an innovative international data collection project aimed at tackling some of the core questions in homeopathy, including the key aspect of treatment individualization. We observed high levels of variability in both rubric selection and remedy prescriptions. This presentation elaborates further on both wanted and unwanted variability in homeopathy from a methodological perspective. Methods: Two types of variability can be distinguished: wanted variability and unwanted variability. Sources of wanted variability in homeopathic prescribing are explored and identified. Unwanted variability is referred to as “noise”. The topic of noise is explored further from a methodological perspective. Results: The nature and possible sources of variability in homeopathic practice are illustrated. The topic of noise, as distinct from bias, in homeopathic practice is explained. Noise-reducing measures are proposed and illustrated, including (1) the use of questionnaires, (2) the use of symptoms that can be ranked, such as polar symptoms, and (3) the grouping of similar symptoms for analytical purposes. Standardizing the data collection process by using questionnaires was found to be the single most effective noise-reducing measure. The possibilities and limitations of standardizing data collection, including the role of algorithms, are explained. Conclusion: Noise in homeopathic practice is a poorly understood topic, despite the fact that it can seriously hamper the ability to learn lessons from homeopaths’ individual and collective experiences. Partly standardizing data collection is an important and under-utilized tool for improving the success of homeopathy. Elaborating on the nature and sources of both wanted and unwanted variability and on how the latter can be reduced has important implications for the future of homeopathy practice.