Inferring influenza infection attack rate from seroprevalence data

Area of Science:

• Epidemiology
• Infectious Disease Modeling
• Immunology

Background:

• Seroprevalence surveys are vital for estimating infection attack rates (IAR) during epidemics like influenza.
• Current methods often use arbitrary seropositivity thresholds and assume a 100% infection-seropositivity probability (ISP), which may not be robust.

Purpose of the Study:

• To investigate the impact of seropositivity thresholds and ISP on IAR estimates in influenza.
• To characterize the transmission dynamics and seroprevalence of the 2009 influenza A/H1N1 pandemic (pdmH1N1).

Main Methods:

• Utilized an age-structured transmission model to simulate pdmH1N1 spread in Hong Kong.
• Estimated ISP across different age groups at varying microneutralization (MN) titer thresholds (e.g., 1:20, 1:40).

Main Results:

• ISP varied significantly by age and threshold: 99% were MN1:20 seropositive, but only 34%-72% were MN1:40 seropositive.
• IAR estimates are sensitive to the chosen seropositivity threshold and ISP assumptions.
• Model parameters (reproductive number, generation time) aligned with existing consensus on pdmH1N1.

Conclusions:

• Conventional practices for influenza seroprevalence studies require reevaluation.
• Developing standardized methods for analyzing serologic data, including real-time ISP bias assessment, is essential for pandemic preparedness.
• These principles apply to seroprevalence studies for other infectious diseases.