Estimate the excess numbers of influenza-like illness cases attributed to influenza in Beijing by Serfling regression model

Abstract

Objective To estimate the excess numbers of influenza-like illness cases attributed to influenza by Serfling regression model during 2007 to 2011 in Beijing.Methods We obtained weekly numbers of influenza-like illness cases in Grade 2 or 3 hospitals from a city level surveillance system,“Early warning system for infectious disease in medical institutions in Beijing”.The excess numbers of influenzalike illness cases attributed to influenza were assessed by Serfling regression model.Results There were 6,3,19 and 8 epidemic weeks in season 2007/2008,2008/2009,2009/2010 and 2010/2011,respectively,including 52 nd-4 th week in season 2007/2008,51st- 1st week in season 2008/2009,37 th-38 th and 42 nd-6th week in season 2009/2010,and 48th-3rd week in season 2010/2011.The excess numbers of influenza-like illnesscases attributed to influenza were 41905 (95％ CI:21541-62265),17936 (95％ CI:7747-28124),226150 (95％ CI:162531-289768) and 40083 (95％ CI:13013-67154),and accounted for 44.75％(95％ CI:23.00-66.49％),37.93％ (95％ CI:16.38-58.47％),50.01％ (95％ CI:35.94-64.08％ ) and 27.89％ (95％ CI:8.05-46.73％ ) of all influenza-like illness cases in season 2007/2008,2008/2009,2009/2010 and 2010/2011,respectively.Conclusion There were hundreds of thousands of excess influenza-like illness cases and outpatient services attributed to influenza pandemic every year estimated by this model. Key words: Influenza; Influenza-like illness cases; Disease burden; Serfling regression model