A model for immunological correlates of protection

Immunological assays measure characteristics of the immune system, such as antibody levels, specific to certain diseases. High assay values are often associated with protection from disease. A question of interest is how the relationship between assay values and subsequent development of disease should be quantitatively modelled. Existing approaches successfully model the relationship for high assay values, where the probability of developing disease is low. However at low assay values, the probability of developing disease is more closely associated with factors such as disease prevalence rates and an individual's chance of exposure to infection; these are less well captured by existing models. This paper presents a model that accommodates both assay values and factors independent of assay values, enabling protection from disease to be modelled over the whole range of assay values and proposing a method for predicting the efficacy of a vaccine from the assays of vaccinees and non-vaccinees.