Sentinel surveillance for influenza among severe acute respiratory infection and acute febrile illness inpatients at three hospitals in Ghana

BACKGROUND Influenza epidemiology in Africa is generally not well understood. Using syndrome definitions to screen patients for laboratory confirmation of infection is an established means to effectively conduct influenza surveillance. METHODS To compare influenza-related epidemiologic data, from October 2010 through March 2013, we enrolled hospitalized severe acute respiratory infection (SARI; fever with respiratory symptoms) and acute febrile illness (AFI; fever without respiratory or other localizing symptoms) patients from three referral hospitals in Ghana. Demographic and epidemiologic data were obtained from enrolled patients after which nasopharyngeal and oropharyngeal swabs were collected, and processed by molecular methods for the presence of influenza viruses. RESULTS Of 730 SARI patients, 59 (8%) were influenza positive; of 543 AFI patients, 34 (6%) were positive for influenza. Both SARI and AFI surveillance yielded influenza A(H3N2) (3% versus 1%), A(H1N1)pdm09 (2% versus 1%), and influenza B (3% versus 4%) in similar proportions. Data from both syndromes show year-round influenza transmission but with increased caseloads associated with the rainy seasons. CONCLUSIONS As an appreciable percentage of influenza cases (37%) presented without defined respiratory symptoms, and thus met the AFI but not the SARI definition, it is important to consider broader screening criteria (i.e., AFI) to identify all laboratory-confirmed influenza. The identified influenza transmission seasonality has important implications for the timing of related public health interventions.