Methodological limitations in assessing influenza vaccine effectiveness in school children: A call for rigorous clinical epidemiology

We read with interest the article by Kajiume et al. reporting a 10-year investigation of influenza vaccine effectiveness in elementary and middle schools. 1) While the longitudinal data on immunization rates among Japanese school children offer useful descriptive context, we have serious concerns about the study's methodology and the authors' conclusion that infant influenza vaccination increases the subsequent risk of influenza infection through "original antigenic sin." The study design has critical flaws that preclude any such causal inference.