Estimation of the Likelihood Ratio of Different Symptoms for Six Homeopathic Medicines: Prognostic Factor Research

Abstract

Abstract Background Arsenicum album, Causticum, Nux vomica, Pulsatilla nigricans, Rhus toxicodendron and Sulphur are frequently prescribed homeopathic medicines; however, their symptoms, as mentioned in different homeopathic literature works, have rarely been investigated systematically. Likelihood ratio (LR), based on Bayesian statistics, may reflect a better estimation of the strengths of symptoms than the existing entries in the homeopathic literature. Methods A prospective, longitudinal, analytical patient outcome study was conducted in the outpatient departments of D. N. De Homeopathic Medical College and Hospital, Kolkata, on 1,954 patients over 21 months. The outcomes were recorded at each follow-up using the Outcome Related to Impact on Daily Living (ORIDL) +4 to −4 scale. The average period of treatment for each participant was 3 months. The LRs of four symptoms for each of the six selected medicines were calculated. Results One hundred and two different remedies were prescribed. The prevalence, LR + , and LR − , with respective 95% confidence intervals, of different symptoms were reported. The study found that the following symptoms had particularly high LR+ scores: “intense sympathy for the suffering of others” (Causticum, LR+ = 12.0); “dyspepsia from business anxiety” (Nux vomica, LR+ = 27.4); “burning pain relieved by heat” (Arsenicum album, LR+ = 29.6); “envy” (Pulsatilla nigricans, LR+ = 13.2); “desire for milk” (Rhus toxicodendron, LR+ = 7.5); “very selfish, no regard for others” (Sulphur, LR+ = 20.6). The findings corroborated well with the presentation of the symptoms in different homeopathic materia medica and repertories. ORIDL scores of +2 or greater were identified most prominently for Pulsatilla nigricans (n = 138) and Sulphur (n = 119). Conclusion There was adequate evidence to attribute all the assessed symptoms to the medicines investigated. Further studies with a larger population are warranted to tackle the possible confirmation bias.