ObjectiveTo improve understanding of the relative burden of differentcausative respiratory pathogens on respiratory syndromic indicatorsmonitored using syndromic surveillance systems in England.IntroductionPublic Health England (PHE) uses syndromic surveillance systemsto monitor for seasonal increases in respiratory illness. Respiratoryillnesses create a considerable burden on health care services andtherefore identifying the timing and intensity of peaks of activity isimportant for public health decision-making. Furthermore, identifyingthe incidence of specific respiratory pathogens circulating in thecommunity is essential for targeting public health interventionse.g. vaccination. Syndromic surveillance can provide early warningof increases, but cannot explicitly identify the pathogens responsiblefor such increases.PHE uses a range of general and specific respiratory syndromicindicators in their syndromic surveillance systems, e.g. “allrespiratory disease”, “influenza-like illness”, “bronchitis” and“cough”. Previous research has shown that “influenza-like illness”is associated with influenza circulating in the community1whilst“cough” and “bronchitis” syndromic indicators in children under 5are associated with respiratory syncytial virus (RSV)2, 3. However, therelative burden of other pathogens, e.g. rhinovirus and parainfluenzais less well understood. We have sought to further understand therelationship between specific pathogens and syndromic indicators andto improve estimates of disease burden. Therefore, we modelled theassociation between pathogen incidence, using laboratory reports andhealth care presentations, using syndromic data.MethodsWe used positive laboratory reports for the following pathogens as aproxy for community incidence in England: human metapneumovirus(HMPV), RSV, coronavirus, influenza strains, invasivehaemophilusinfluenzae, invasivestreptococcus pneumoniae, mycoplasmapneumoniae, parainfluenza and rhinovirus. Organisms were chosenthat were found to be important in previous work2and were availablefrom routine laboratory testing. Syndromic data included consultationswith family doctors (called General Practitioners or GPs), calls to anational telephone helpline “NHS 111” and attendances at emergencydepartments (EDs). Associations between laboratory reports andsyndromic data were examined over four winter seasons (weeks40 to 20), between 2011 and 2015. Multiple linear regression was usedto model correlations and to estimate the proportion of syndromicconsultations associated with specific pathogens. Finally, burdenestimates were used to infer the proportion of patients affected byspecific pathogens that would be diagnosed with different symptoms.ResultsInfluenza and RSV exhibited the greatest seasonal variation andwere responsible for the strongest associated burden on generalrespiratory infections. However, associations were found with theother pathogens and the burden ofstreptococcus pneumoniaewasimportant in adult age groups (25 years and over).The model estimates suggested that only a small proportion ofpatients with influenza receive a specific diagnosis that is coded toan “influenza-like illness” syndromic indicator, (6% for both GPin-hours consultations and for emergency department attendances),compared to a more general respiratory diagnosis. Also, patients withinfluenza calling NHS 111 were more likely to receive a diagnosisof fever or cough than cold/flu. Despite these findings, the specificsyndromic indicators remained more sensitive to changes in influenzaincidence than the general indicators.ConclusionsThe majority of patients affected by a seasonal respiratory pathogenare likely to receive a non-specific respiratory diagnosis. Therefore,estimates of community burden using more specific syndromicindicators such as “influenza-like illness” are likely to be a severeunderestimate. However, these specific indicators remain importantfor detecting changes in incidence and providing added intelligenceon likely causative pathogens.Specific syndromic indicators were associated with multiplepathogens and we were unable to identify indicators that were goodmarkers for pathogens other than influenza or RSV. However, futurework focusing on differences between ages and the relative levels ofa range of pathogens may be able to provide estimates for the mix ofpathogens present in the community in real-time.
Identifying and Communicating the Importance of the Variable Nature of SyS Data