APPLICATION OF BAYESIAN THEOREM IN VALIDATION OF RUBRICS FROM KENT'S REPERTORY

Monika Kesarwani; Arun Bhargav Jadhav; Jyoti Deepak Patil

Volume : IX, Issue : III, March - 2020

APPLICATION OF BAYESIAN THEOREM IN VALIDATION OF RUBRICS FROM KENT'S REPERTORY

Monika Kesarwani, Arun Bhargav Jadhav, Jyoti Deepak Patil

Abstract :

The reliability issues in the grading system of Kents repertory may be overcome by implementing the Bayesian statistical approach to calculate the likelihood ratio (LR) with 95% confidence interval (CI) and probability. Hence the values obtained for 7 analysed ruics from a retrospective study were compared with the already set cut off values for validation. Prevalence of a symptom within the targeted population was (1.5 – 32.4%) and the ruics size varying from 60- 226 in the Kent�s repertory. Out of the 85 relevant outcomes obtained, 5.9% results showed probable occurrence with the existing entries while 30.6 % suggested the downgrading of the existing typeface. 63.5% results are still doubtful out of which 26% entries are likely to be weak indications for Roman typeface (LR>1, p>0.70) and 5.6% medicines which otherwise doubtful are present in the population (symptom present with cured medicine) but absent in the repertory.