Decipher my Data – Investigating the association between school absence prevalence collected through scientific engagement with influenza surveillance data.
Authors:
RW Aldridge, AC Hayward, N Field, C Warren-Gash, and C Smith, from the research Department of Infection and Population Health at University College London, R Pebody, from the Respiratory Diseases Department, Public Health England, D Fleming, from Pate’s Grammar School, and S McCracken, on behalf of the Decipher my Data project and schools.
Abstract
Background:
School aged children are a key link in the transmission of influenza. Most cases have little or no interaction with health services and are therefore missed by the majority of existing surveillance systems. As part of a Public Engagement with Science project, this study aimed to establish an electronic system for the collection of routine school absence data and to determine if school absence prevalence was correlated with established healthcare surveillance measures for circulating influenza.
Methods:
We collected data for two influenza seasons in 2011/12 and 2012/13. The primary outcome was daily school absence prevalence (weighted to make it nationally representative) for children aged 11 to 16. School absence prevalence was triangulated graphically and through univariable linear regression to Royal College of General Practitioners (RCGP) influenza like illness (ILI) episode incidence rate, and as a counterfactual analysis, to national microbiological surveillance data on the proportion of samples positive for influenza (A+B), Rhinovirus, RSV and laboratory confirmed cases of Norovirus.
Results:
A website was created that enabled 27 schools to submit data over the two respiratory seasons. During the first season, levels of influenza measured by established surveillance were low. In the second season, a peak of school absence prevalence occurred in week 51, and week 52 in established influenza surveillance systems. Linear regression showed a strong association between the school absence prevalence and RCGP ILI, laboratory confirmed cases of influenza A & B, and some evidence for a linear association with Rhinovirus and Norovirus.
Interpretation:
This study provides some evidence for the validity of using routine school absence prevalence as a novel tool for influenza surveillance in children. It was not possible to establish conclusively whether school absence prevalence detected outbreaks of disease earlier than existing data, and future work should examine this further.
Funding:
Wellcome Trust, People Award, 096802.
