Utility of Nontraditional Data Sources for Early Detection of Influenza

Shari Barlow, Jonathan Temte, Yenlik Zheteyeva, Ashley Fowlkes, Carrie Reed, Derek Cummings


ObjectiveThis session will provide an overview of the current systemsfor influenza surveillance; review the role of schools in influenzatransmission; discuss relationships between school closures, schoolabsenteeism, and influenza transmission; and explore the usefulnessof school absenteeism and unplanned school closure monitoring forearly detection of influenza in schools and broader communities.IntroductionInfluenza surveillance is conducted through a complex networkof laboratory and epidemiologic systems essential for estimatingpopulation burden of disease, selecting influenza vaccine viruses,and detecting novel influenza viruses with pandemic potential (1).Influenza surveillance faces numerous challenges, such as constantlychanging influenza viruses, substantial variability in the number ofaffected people and the severity of disease, nonspecific symptoms,and need for laboratory testing to confirm diagnosis. Exploringadditional components that provide morbidity information mayenhance current influenza surveillance.School-aged children have the highest influenza incidence ratesamong all age groups. Due to the close interaction of children inschools and subsequent introduction of influenza into households,it is recognized that schools can serve as amplification points ofinfluenza transmission in communities. For this reason, pandemicpreparedness recommendations include possible pre-emptive schoolclosures, before transmission is widespread within a school system orbroader community, to slow influenza transmission until appropriatevaccines become available. During seasonal influenza epidemics,school closures are usually reactive, implemented in response tohigh absenteeism of students and staff after the disease is alreadywidespread in the community. Reactive closures are often too late toreduce influenza transmission and are ineffective.To enhance timely influenza detection, a variety of nontraditionaldata sources have been explored. School absenteeism was suggestedby several research groups to improve school-based influenzasurveillance. A study conducted in Japan demonstrated that influenza-associated absenteeism can predict influenza outbreaks with highsensitivity and specificity (2). Another study found the use of all-causes absenteeism to be too nonspecific for utility in influenzasurveillance (3). Creation of school-based early warning systemsfor pandemic influenza remains an interest, and further studies areneeded. The panel will discuss how school-based surveillance cancomplement existing influenza surveillance systems.

Full Text:


DOI: https://doi.org/10.5210/ojphi.v9i1.7636

Online Journal of Public Health Informatics * ISSN 1947-2579 * http://ojphi.org