The importance of age-specific data in routine syndromic surveillance
AbstractObjectiveTo investigate whether aberration detection methods for syndromicsurveillance would be more useful if data were stratified by age band.IntroductionWhen monitoring public health incidents using syndromicsurveillance systems, Public Health England (PHE) uses the ageof the presenting patient as a key indicator to further assess theseverity, impact of the incident, and to provide intelligence on thelikely cause. However the age distribution of cases is usually notconsidered until after unusual activity has been identified in the all-ages population data. We assessed whether monitoring specific agegroups contemporaneously could improve the timeliness, specificityand sensitivity of public health surveillance.MethodsFirst, we examined a wide range of health indicators from the PHEsyndromic surveillance systems to identify for further study thosewith the greatest seasonal variation in the age distribution of cases.Secondly, we examined the identified indicators to ascertain whetherany age bands consistently lagged behind other age bands. Finally,we applied outbreak detection methods retrospectively to age specificdata, identifying periods of increased activity that were only detectedor detected earlier when age-specific surveillance was used.ResultsSeasonal increases in respiratory indicators occurred first inyounger age groups, with increases in children under 5 providingearly warning of subsequent increases occurring in older age groups.Also, we found age specific indicators improved the specificity ofsurveillance using indicators relating to respiratory and eye problems;identifying unusual activity that was less apparent in the all-agespopulation.ConclusionsRoutine surveillance of respiratory indicators in young childrenwould have provided early warning of increases in older age groups,where the burden on health care usage, e.g. hospital admissions, isgreatest. Furthermore this cross-correlation between ages occurredconsistently even though the age distribution of the burden ofrespiratory cases varied between seasons. Age specific surveillancecan improve sensitivity of outbreak detection although all-agesurveillance remains more powerful when case numbers are low.
How to Cite
Morbey, R., Elliot, A. J., & Smith, G. E. (2017). The importance of age-specific data in routine syndromic surveillance. Online Journal of Public Health Informatics, 9(1). https://doi.org/10.5210/ojphi.v9i1.7602
Novel algorithms, statistical or mathematical methods