Common symptoms in Homoeopathy. Are they really useless? Results from an empirical assessment of best cases data from a Bayesian perspective: An observational study

Background: In seeking the homoeopathic medicine that best fits their case, homoeopaths traditionally yield greater value to the most peculiar symptoms, while dismissing those they believe to be common, according to the number of medicines found in repertories' rubrics. Objectives: To assess the validity of the following homoeopathic usual practice notions: common symptoms are useless and prescription should rely on peculiar ones. Methods: Empirical data from the Best Chronic Homeopathic Cases (BCHC) and the COVID-19 cases collection databases were analysed by using a Bayesian approach. Symptoms' prevalence and likelihood ratio (LR) were calculated. Results from three very common symptoms in the databases - fatigue, anxiety and headache - were compared against a current homoeopathic repertory (Complete Repertory). Results: From the BCHC and COVID-19 collection databases, 794 and 1,424 cases, 2,663 and 35 symptoms, and 101 and 76 medicines were assessed, respectively. Symptoms' prevalence distributes in a highly asymmetrical fashion. In the BCHC set, half of the symptoms were observed in one case each, making statistical analysis unfeasible. 14 symptoms were seen in ≥ 10% of patients. The two assessed databases rendered similar LRs for the very common symptoms analysed, while suggesting that most medicines are misplaced as indications in the Repertory. Conclusion: When a Bayesian approach is applied, the common symptoms were found to be useful, and highly peculiar ones may be unreliable in the process of finding the patient's medicine.