Quantifying differences in the epidemic curves from three influenza surveillance systems: A nonlinear regression analysis

E. G. Thomas*, J. M. McCaw, H. A. Kelly, K. A. Grant, J. McVernon

*Corresponding author for this work

    Research output: Contribution to journalArticlepeer-review

    6 Citations (Scopus)

    Abstract

    Influenza surveillance enables systematic collection of data on spatially and demographically heterogeneous epidemics. Different data collection mechanisms record different aspects of the underlying epidemic with varying bias and noise. We aimed to characterize key differences in weekly incidence data from three influenza surveillance systems in Melbourne, Australia, from 2009 to 2012: laboratory-confirmed influenza notified to the Victorian Department of Health, influenza-like illness (ILI) reported through the Victorian General Practice Sentinel Surveillance scheme, and ILI cases presenting to the Melbourne Medical Deputising Service. Using nonlinear regression, we found that after adjusting for the effects of geographical region and age group, characteristics of the epidemic curve (including season length, timing of peak incidence and constant baseline activity) varied across the systems. We conclude that unmeasured factors endogenous to each surveillance system cause differences in the disease patterns recorded. Future research, particularly data synthesis studies, could benefit from accounting for these differences.

    Original languageEnglish
    Pages (from-to)427-439
    Number of pages13
    JournalEpidemiology and Infection
    Volume143
    Issue number2
    DOIs
    Publication statusPublished - 12 Jan 2015

    Fingerprint

    Dive into the research topics of 'Quantifying differences in the epidemic curves from three influenza surveillance systems: A nonlinear regression analysis'. Together they form a unique fingerprint.

    Cite this