Abstract
BACKGROUND: Our understanding of the epidemiology of influenza is limited in tropical regions, which in turn has hampered identifying optimal region-specific policy to diminish disease burden. Influenza-like illness (ILI) is a clinical diagnosis that can be used as a surrogate for influenza. This study aimed to define the incidence and seasonality of ILI and to assess its association with climatic variables and school calendar in an urban community in the tropical region of Salvador, Brazil. METHODS: Between 2009 and 2013, we conducted enhanced community-based surveillance for acute febrile illnesses (AFI) among patients ≥ 5 years of age in a slum community emergency unit in Salvador, Brazil. ILI was defined as a measured temperature of ≥ 37.8 °C or reported fever in a patient with cough or sore throat for ≤ 7 days, and negative test results for dengue and leptospirosis. Seasonality was analyzed with a harmonic regression model. Negative binomial regression models were used to correlate ILI incidence with rainfall, temperature, relative humidity and the number of days per month that schools were in session while controlling for seasonality. RESULTS: There were 2,651 (45.6% of 5,817 AFI patients) ILI cases with a mean annual incidence of 60 cases/1,000 population (95% CI 58-62). Risk of ILI was highest among 5-9 year olds with an annual incidence of 105 cases/1,000 population in 2009. ILI had a clear seasonal pattern with peaks between the 35-40th week of the year. ILI peaks were higher and earlier in 5-9 year olds compared with > 19 year olds. No association was seen between ILI and precipitation, relative humidity or temperature. There was a significant association between the incidence of ILI in children 5-9 years of age and number of scheduled school days per month. CONCLUSIONS: We identified a significant burden of ILI with distinct seasonality in the Brazilian tropics and highest rates among young school-age children. Seasonal peaks of ILI in children 5-9 years of age were positively associated with the number of school days, indicating that children may play a role in the timing of seasonal influenza transmission.