INTRODUCTION: Understanding the seasonality of influenza can help inform prevention and clinical treatment strategies. The aim of this manuscript is to describe the trends and epidemiology of outpatient influenza in South Africa prior to the influenza A(H1N1) pandemic. METHODS: Throughout each year, participating healthcare practitioners sent throat swabs from patients with influenza-like illness (ILI) to the National Institute for Communicable Diseases for influenza testing by immunofluorescence and viral culture through the Viral Watch influenza surveillance program. RESULTS: From 1984 to 2004, participating sites were restricted to 1 province and the annual number of specimens ranged from 91 to 534. In 2005 the program was expanded. By 2008 the program included all 9 provinces; 1276 specimens were submitted that year. The mean week of onset was the first week of June and the mean peak was the first week of July. The duration of the season ranged from 6 to 18 weeks with a mean of 10 weeks. The mean annual influenza detection rate was 28% (range, 23%-41%). Influenza A(H3N2) predominated in 14 (56%) of the 25 years, seasonal influenza A(H1N1) in 7 (28%), and influenza B in 2 (8%), and in 2 years multiple types cocirculated. CONCLUSIONS: The program has provided valuable data on the timing of the influenza season each year that can be useful to direct the timing of vaccination and assist clinicians in deciding whether to prescribe empirical antiviral therapy.
INTRODUCTION: Understanding the seasonality of influenza can help inform prevention and clinical treatment strategies. The aim of this manuscript is to describe the trends and epidemiology of outpatient influenza in South Africa prior to the influenza A(H1N1) pandemic. METHODS: Throughout each year, participating healthcare practitioners sent throat swabs from patients with influenza-like illness (ILI) to the National Institute for Communicable Diseases for influenza testing by immunofluorescence and viral culture through the Viral Watch influenza surveillance program. RESULTS: From 1984 to 2004, participating sites were restricted to 1 province and the annual number of specimens ranged from 91 to 534. In 2005 the program was expanded. By 2008 the program included all 9 provinces; 1276 specimens were submitted that year. The mean week of onset was the first week of June and the mean peak was the first week of July. The duration of the season ranged from 6 to 18 weeks with a mean of 10 weeks. The mean annual influenza detection rate was 28% (range, 23%-41%). Influenza A(H3N2) predominated in 14 (56%) of the 25 years, seasonal influenza A(H1N1) in 7 (28%), and influenza B in 2 (8%), and in 2 years multiple types cocirculated. CONCLUSIONS: The program has provided valuable data on the timing of the influenza season each year that can be useful to direct the timing of vaccination and assist clinicians in deciding whether to prescribe empirical antiviral therapy.