Abstract
Background The critical task of researchers conducting double-blinded, randomized, placebo-controlled
homeopathic pathogenetic trials is to segregate the signals from the noises. The noises
are signs and symptoms due to factors other than the trial drug; signals are signs
and symptoms due to the trial drug. Unfortunately, the existing tools (criteria for
a causal association of symptoms only with the tested medicine, qualitative pathogenetic
index, quantitative pathogenetic index, pathogenic index) have limitations in analyzing
the symptoms of the placebo group as a comparator, resulting in inadequate segregation
of the noises. Hence, the Jaccard similarity index and the Noise index are proposed
for analyzing the symptoms of the placebo group as a comparator.
Methods The Jaccard similarity index is the ratio of the number of common elements among
the placebo and intervention groups to the aggregated number of elements in these
groups. The Noise index is the ratio of common elements among the placebo and intervention
group to the total elements of the intervention group. Homeopathic pathogenetic trials
of Plumbum metallicum, Piper methysticum and Hepatitis C nosode were selected for experimenting with the computation of the Jaccard similarity index
and the Noise index.
Results Jaccard similarity index calculations show that 8% of Plumbum metallicum's elements, 10.7% of Piper methysticum's elements, and 19.3% of Hepatitis C nosode's elements were similar to the placebo group when elements of both the groups (intervention
and placebo) were aggregated. Noise index calculations show that 10.7% of Plumbum metallicum's elements, 13.9% of Piper methysticum's elements and 25.7% of Hepatitis C nosode's elements were similar to those of the placebo group.
Conclusion The Jaccard similarity index and the Noise index might be considered an additional
approach for analyzing the symptoms of the placebo group as a comparator, resulting
in better noise segregation in homeopathic pathogenetic trials.
Keywords
Jaccard similarity index - Homeopathic pathogenetic trials - Noise index
