Using capture-recapture methods to estimate influenza hospitalization incidence rates

Area of Science:

• Epidemiology
• Public Health
• Biostatistics

Background:

• Accurate disease incidence and burden estimates are crucial for effective public health policy.
• Traditional single-source data often underestimate disease prevalence.
• The capture-recapture (C-R) method offers a way to improve estimates by integrating multiple data streams.

Purpose of the Study:

• To estimate the population size of influenza cases using the Petersen C-R method.
• To calculate the adult influenza hospitalization burden.
• To compare the effectiveness of multi-source C-R methods versus single-source data for disease burden estimation.

Main Methods:

• Utilized clinical virology test results and influenza vaccine effectiveness study data from 2016-2019.
• Applied the Petersen C-R method to estimate influenza case population size.
• Employed a Centers for Disease Control and Prevention (CDC) multiplier method to determine hospitalization burden.

Main Results:

• Combined databases identified 343 (clinical) and 313 (research) influenza cases.
• Capture completeness for influenza hospitalizations was estimated at 76%.
• Annual influenza hospitalization incidence averaged 307-309 cases per 100,000 adults, varying by age and season.

Conclusions:

• Multi-database capture-recapture methods, combined with multiplier adjustments, enhance population estimates for influenza disease burden.
• These improved estimates are more reliable than those derived from single data sources.
• This approach supports more informed public health policy decisions regarding influenza.