Reconciling heterogeneous dengue virus infection risk estimates from different study designs

Area of Science:

• Epidemiology
• Infectious Disease Modeling
• Public Health

Background:

• Accurate estimation of the force of infection (FOI) is vital for understanding pathogen transmission and population immunity.
• For dengue, FOI is particularly important as prior exposure increases the risk of severe disease.
• Existing methods for measuring FOI include longitudinal serology (seroincidence), cross-sectional serology (seroprevalence), and age-of-case data.

Purpose of the Study:

• To assess the agreement between FOI estimates derived from different data sources: seroincidence, seroprevalence, and reported cases.
• To identify factors contributing to discrepancies in FOI estimates across these methods.
• To develop improved inference models for reconciling FOI and susceptibility estimates.

Main Methods:

• Utilized 26 years of data from cohort studies and hospital-attended dengue cases in Kamphaeng Phet, Thailand.
• Compared FOI estimates obtained from longitudinal serology, cross-sectional serology, and age-of-case data.
• Conducted extensive simulations and theoretical analysis to investigate sources of incongruence, including antibody kinetics, assay noise, and FOI heterogeneity.

Main Results:

• FOI estimates from the three sources were highly inconsistent.
• Seroincidence-derived FOI estimates were 1.75 to 4.05 times higher than case-derived FOI.
• Seroprevalence-derived FOI showed moderate correlation with case-derived FOI but yielded slightly lower estimates.

Conclusions:

• Discrepancies in FOI estimates arise from failure to account for antibody kinetics, assay noise, and age-related FOI heterogeneity.
• Integrating these factors into standard inference models successfully reconciled FOI and susceptibility estimates.
• Comparing inferences across multiple data types is essential for robust epidemiological insights, especially for diseases like dengue.