Surveillance system of severe influenza cases admitted to the regional ICU,2009-2015


  • Esra Morvan Regional office, Sante publique France, Orléans, France
  • Joanna Parra Regional office, Sante publique France, Orléans, France
  • Gerard Roy Regional office, Sante publique France, Orléans, France
  • Dominique Jeannel Regional office, Sante publique France, Orléans, France



ObjectiveThe study aimed at: i) analyses the regional characteristics and riskfactors of severe influenza, taking into account dominant circulatingvirus(es) ii) estimate the regional completeness of the surveillancesystem.IntroductionEvery year, circulating influenza viruses generate a significantnumber of deaths. During the 2009 pandemic influenza A(H1N1),a national non mandatory surveillance system of severe influenzacases admitted to intensive care units(ICU) was set up in France.This surveillance is regionally driven by the regional offices (CIRE)of Santé publique France, the French Public Health Agency. Thisreport provides epidemiologic analysis of the recorded data sincethe implementation of surveillance in the Centre-Val de Loire regionover seasons 2009-10 to 2015-16 in regard of influenza epidemicsdynamics.MethodsSurveillance was carried out each year from October to April.Descriptive and analytic analyses were conducted to comparepopulation characteristics, pre-existing risk factors and the clinicaldata according to influenza season and dominant circulatinginfluenza virus(es). Logistic regressions were performed to identifyfactors associated with an increased risk of acute respiratory distresssyndrome (ARDS) or death. Two capture-recapture analyses wereperformed to establish the completeness of the surveillance systemin the region. The first one was realized on all cases, using two datasources (hospital records/surveillance data) and the second one, onlyon deaths, using three data sources(additional source: medical deathcertificates).ResultsFrom 2009-10 to 2015-16, the outbreak of influenza epidemicswas started more and more late. The number of severe influenzacases reported in the Loire Valley varied from 19 in 2010-11 to 75 in2014-15. Overall, the most affected population was adults, from 41%in 2011-12 to 83% in 2009-10. However seniors (more than 65 yearsold) represented an important part of patients during three epidemics:50% in 2011-12 and around 45% during the two last seasons;during these epidemics, men, (60%-68%), were more affected thanwomen. Patients’ pre-existing risk factors were mainly: being olderthan 65 years old and suffering of cardiac or pulmonary diseases.The comparison by dominant viruses over the seasons revealed thatwhen A(H1N1) virus prevailed, severe influenza occurred mainlyin adults patients with any type of pre-existing risk factors whereaswhen A(H3N2) virus prevailed, seniors with pre-existing pulmonarydisease were the most affected. More than a third of patientsdeclared an ARDS. The overall observed lethality was close to 16%.ARDS occurred more frequently in patients who were middle-aged(45-64 years), immunocompromised or infected with A(H1N1).Pre-existing pulmonary disease was a protective factor. Risk factorsassociated with death were being older than 65 years, male and havingdeclared an ARDS. The completeness of this surveillance system wasestimated by capture-recapture methods at 59% for severe influenzacases and 40% for death cases.ConclusionsThe epidemiology of severe influenza and epidemics dynamics inthe Centre-Val de Loire follow the national trends. Every season ischaracterized by the same dominant virus at national and regionallevels in intensive care units. Influenza epidemics 2009-10 and2014-15 were particularly long and severe, the first dominatedby the A(H1N1)pdm09 virus and the second by the A(H3N2).Our study has demonstrated that the populations at risk of severeinfluenza differ according to the circulating virus(es). Accordingto the obtained estimations, the completeness of the surveillancesystem, based on voluntary report by physicians, can be consideredas satisfactory. Regarding influenza deaths relatively low percentageof completeness may be explained by the fact that two sources arehospital based whereas the third one, medical death certificates,includes all influenzadeaths with no information on the death place.Many patients were not vaccinated or their status was unknown. Mostcases admitted to ICU presented pre-existing risk factors includedin eligibility criteria in influenza vaccination policies. This studyoutlines the importance of vaccination as the first prevention measure.




How to Cite

Morvan, E., Parra, J., Roy, G., & Jeannel, D. (2017). Surveillance system of severe influenza cases admitted to the regional ICU,2009-2015. Online Journal of Public Health Informatics, 9(1).



Population Health Analysis of Patient-Centered, Health System Outcomes