Abstract
Background Homeopathic repertories are essential tools in remedy diagnosis, helping practitioners match patient symptoms with those produced by remedies. However, repertories often need to be revised due to omissions, misinterpretations, and incomplete representation of remedy symptoms. Despite their importance, the sensitivity of repertories – their ability to correctly identify remedies based on corresponding rubrics – has never been systematically estimated. Addressing this gap is crucial to ensuring repertories' accuracy, reliability and validity in homeopathic practice.
Methods We adopted the sensitivity formula used in medical diagnostics, where true positives indicate correct remedy identification and false negatives represent failures. This method was applied to Kent's repertory for Allium cepa using symptoms from Hering's Guiding Symptoms of our Materia Medica. We extracted the rubrics and identified the non-representing rubrics and omissions. We created a Python script that generated combinations of rubrics based on Allen's ‘three-legged stool rule’. We calculated the sensitivity as the ratio of true positives to total combinations.
Results Of the 525 symptoms of Allium cepa, we extracted 364 rubrics from Kent's repertory, with 161 symptoms omitted. Among the extracted rubrics, 111 failed to represent Allium cepa. The Python script generated 23,979,550 combinations, of which 21,050,260 (87.78%) were false negatives, and 2,929,290 (12.2%) were true positives.
Conclusion The sensitivity of Kent's repertory for Allium cepa was estimated as 12.2%. The method can thus effectively estimate the sensitivity for given remedies in a homeopathic repertory. Applying this method to other remedies would enhance a repertory's diagnostic accuracy and could lead to the development of artificial intelligence-driven tools for repertorial analysis.
Keywords
homeopathy - sensitivity and specificity - clinical decision-making - symptom assessment - programming languages - diagnostic techniques and procedures
