Estimating incidence of co-infection from emergency department serosurvey data

Date: Monday 20 June 2022, 5.30PM
Location: University of Warwick
University of Warwick CV4 7AL
Mathematical Science Building room MB2.23
Map of the campus available at
Local Group Meeting

Share this event

Our understanding of pathogens and disease transmission has improved dramatically over the past 100 years, but coinfection, how different pathogens interact with each other, remains a challenge. Cross-sectional serological studies including multiple pathogens offer a crucial insight into this problem.  We used data from three cross-sectional serological surveys (in 2003, 2007 and 2013) in a Baltimore emergency department to predict the prevalence for HIV, hepatitis C virus (HCV) and herpes simplex virus, type 2 (HSV2), in a fourth survey in 2016. We developed a mathematical model to make this prediction and used Bayesian techniques to estimate the incidence of infection and co-infection in each age and ethnic group in each year. Overall we found a much stronger age cohort effect than a time effect, so that, while incidence at a given age may decrease over time, individuals born at similar times experience a more constant force of infection over time. These results emphasise the importance of age-cohort counselling and early intervention while people are young. Our approach adds value to data such as these by providing age- and time-specific incidence estimates which could not be obtained any other way, and allows forecasting to enable future public health planning.


For more details see