High added value of a population-based participatory surveillance system for community acute gastrointestinal, respiratory and influenza-like illnesses in Sweden, 2013–2014 using the web

• Published in: Epidemiology and infection 2017-04, Vol.145 (6), p.1193-1202
• Main Authors: PINI, A., MERK, H., CARNAHAN, A., GALANIS, I., VAN STRATEN, E., DANIS, K., EDELSTEIN, M., WALLENSTEN, A.
• Format: Article
• Language: English
• Subjects: Acute gastrointestinal illness; acute respiratory illness; Adolescent; Adult; Age groups; Aged; Aged, 80 and over; Child; Child, Preschool; Community involvement; Community Participation; Confidence intervals; Correlation coefficient; Correlation coefficients; Cough; Data collection; Data correlation; Education; Educational attainment; Epidemiological Monitoring; Exports; Female; Fever; Gastroenteritis - epidemiology; Gastrointestinal diseases; Health Services Research; Health surveillance; Humans; Illnesses; Infant; Infant, Newborn; Infectious diseases; Influenza; influenza-like illness; Information seeking behavior; Internet; Male; Middle Aged; Original Papers; Participation; Physicians; Population; Population-based studies; population-based study; Public health; Queries; Questionnaires; Respiratory diseases; Respiratory Tract Infections - epidemiology; Searching; Signs and symptoms; Sociodemographics; Surveillance; surveillance system; Surveillance systems; Surveys and Questionnaires; Sweden - epidemiology; Vomiting; Websites; Young Adult
• ISSN: 0950-2688; 1469-4409; 1469-4409
• DOI: 10.1017/S0950268816003290

In 2013–2014, the Public Health Agency of Sweden developed a web-based participatory surveillance system, Hӓlsorapport, based on a random sample of individuals reporting symptoms weekly online, to estimate the community incidence of self-reported acute gastrointestinal (AGI), acute respiratory (ARI) and influenza-like (ILI) illnesses and their severity. We evaluated Hӓlsorapport's acceptability, completeness, representativeness and its data correlation with other surveillance data. We calculated response proportions and Spearman correlation coefficients (r) between (i) incidence of illnesses in Hӓlsorapport and (ii) proportions of specific search terms to medical-advice website and reasons for calling a medical advice hotline. Of 34 748 invitees, 3245 (9·3%) joined the cohort. Participants answered 81% (139 013) of the weekly questionnaires and 90% (16 351) of follow-up questionnaires. AGI incidence correlated with searches on winter-vomiting disease [r = 0·81, 95% confidence interval (CI) 0·69–0·89], and ARI incidence correlated with searches on cough (r = 0·77, 95% CI 0·62–0·86). ILI incidence correlated with the web query-based estimated incidence of ILI patients consulting physicians (r = 0·63, 95% CI 0·42–0·77). The high response to different questionnaires and the correlation with other syndromic surveillance systems suggest that Hӓlsorapport offers a reasonable representation of AGI, ARI and ILI patterns in the community and can complement traditional and syndromic surveillance systems to estimate their burden in the community.