Traditional vaccine clinical development is an undertaking involving meticulous, multiple studies in multiple populations at risk of infection and disease over multiple years. SARS-CoV-2 and COVID-19 vaccine development is following this traditional development pathway, and accelerated Phase I-II-III clinical programs are being applied. This is not the first time vaccines have been manufactured and tested quickly to meet a public health crisis. Selected statistical concepts pertaining to vaccine efficacy and safety, relevant during the design and implementation of such clinical development programs, will be discussed.

Clinical safety and immunogenicity data in vaccines are commonly dichotomized into responder and nonresponder binary data. This article expands upon Fagerland, Lyndersen, and Laake to further study Newcombe's hybrid score and Fisher's conditional exact test using imbalanced randomization in vaccine clinical studies. The intent of this article is to study a potential statistical issue in clinical study design and analysis; hence, no reference to any particular product is made.