# Evaluating the Impact of Different Methods on the Timing and Duration of RSV Epidemics: Analysis of Surveillance Data From the GERi (Global Epidemiology of RSV in Hospitalized and Community Care) Study

## TL;DR

This study compares different methods for estimating RSV epidemic timing and duration using global surveillance data to highlight the impact of method choice on results.

## Contribution

The study evaluates eight RSV epidemic estimation methods across multiple countries, revealing significant variability in outcomes based on method selection.

## Key findings

- The median duration of RSV epidemics varied by over 10 weeks depending on the estimation method used.
- The 3% positivity rate method identified the longest epidemics with the highest capture rate.
- The 10% positivity rate, MEM, and 75% AAP methods indicated the shortest epidemics with the lowest capture rate.

## Abstract

We previously reviewed methods for estimating the timing of respiratory syncytial virus (RSV) epidemics. This study examines the impact of various estimation methods on determining the start, end, duration, and capture rate of RSV epidemics.

We applied eight estimation methods to RSV surveillance data from the Global Epidemiology of RSV (GERi) study, covering Chile, New Zealand, Singapore, South Africa, Spain, and the United States: 3% and 10% positivity rate, moving epidemic method (MEM), mean positivity, 1.2% total detections, mean and 60% mean number, and 75% average annual percentage (AAP). We compared the median start, end, duration, and capture rate of RSV epidemics obtained from these methods.

Within countries, the median duration of RSV epidemics varied by over 10 weeks, and the median capture rates ranged from > 95 to < 60%, depending on the estimation method. Generally, the 3% positivity rate method identified the longest RSV epidemics (earliest median start and latest end, and highest capture rate). The 10% positivity rate, MEM, and 75% AAP methods indicated the shortest RSV epidemics with the lowest capture rate. The remaining four methods produced intermediate results.

These findings underscore the importance of selecting estimation methods suited to the surveillance system and the intended use, whether for outbreak alert, planning, or targeted interventions.