Comparison of regression models on estimation of vaccine efficacy in anti-leprosy vaccination trial-a large prospective vaccination trial

• Main Author: Venmani, A.
• Format: Conference Proceeding
• Language: English
• Subjects: Demographics; Effectiveness; Identification methods; Immunization; Poisson density functions; Regression analysis; Regression models; Statistical analysis; Vaccines
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/1.5112333

This study aims to suggest regression models for improving estimation of vaccine efficacy instead of using conventional methods and to identify other influencing factors in evaluating vaccine efficacy among different vaccines and compare estimation of vaccine efficacy between conventional method, logistic, Poisson and Cox regression model. In a prospective cohort study 1,71,400 individuals were enrolled and they were confirmed as healthy with routine medical checkups. The participants were followed up through two subsequent surveys with two to three year interval. The demographic and clinical characteristics of the individuals have been used for constructing the regression models. The odds ratios were obtained from the regression models which enabled us to estimate the adjusted relative risk and also vaccine efficacy. Estimated relative risks obtained from Logistic regression and Poisson regression models give similar results in second resurvey. By comparing the estimated vaccine efficacy using conventional methods with regression models the Vaccines ICRC and BCG+Kml have maximum protection and Logistic and Poisson regression models provides appropriate results than the Cox regression model.