639 Am. J. Trop. Med. Hyg., 62(5), 2000, pp. 639–643 Copyright 2000 by The American Society of Tropical Medicine and Hygiene EPIDEMIOLOGICAL AND VIROLOGICAL INFLUENZA SURVEY IN DAKAR, SENEGAL: 1996–1998 ANNICK DOSSEH, KADER NDIAYE, ANDRE SPIEGEL, MONIQUE SAGNA, AND CHRISTIAN MATHIOT De ´partement de Virologie and Unite ´ d’Epide ´miologie, Institut Pasteur de Dakar, Se ´ne ´gal Abstract. An influenza survey was conducted in seven sentinel sites in Dakar, Senegal from June 1996 to De- cember 1998. Throat or nasal swab cultures were randomly collected from 804 patients suffering from influenza-like symptoms. Influenza viruses were isolated at a similar proportion in adults and in children (P = 0.29). Strains of influenza B viruses were isolated from sporadic cases in 1997, whereas type A virus was associated with an isolated peak. Proportions of influenza virus isolation varied from 17.5% to 40.0% between 1996 and 1998 during the peak period (July/September) of acute respiratory infection in Dakar. Rainfall, humidity, and temperatures rose during the same period. Influenza in Dakar seems to be an-all-age groups respiratory infection characterized by high transmission during the hot and rainy season. The antigenic similarity of the A(H3N2) and B viruses to those circulating elsewhere in the world at the same time was confirmed. However, the A(H1N1) strains were found to be more closely related to an Asiatic strain which had not been isolated outside Asia previously. Consequently, the strain close to the A(H1N1)/ Wuhan/371/95 strain isolated in Dakar was included in the composition of the 1998/1999 influenza vaccine. This reinforces the importance of setting up a national influenza control strategy in tropical regions. INTRODUCTION Influenza is an acute, highly contagious respiratory infec- tion that has the potential to cause very serious illness. Typ- ical clinical symptoms are characterized by a short incuba- tion period, high attack rates, and the progression of the disease through the population. 1 Due to the major antigenic changes occurring with the influenza A virus, epidemics oc- cur every year on a seasonal basis. Influenza B is responsible for less frequent outbreaks. 2 Influenza epidemiology varies according to geographical location. In the temperate regions, influenza is epidemic and occurs during winter, whereas in tropical regions influenza is endemic and more than one pe- riod of activity may occur in any given year. 3 Because of significant mortality, morbidity, and the loss of productive human-hours associated with influenza, strategies for pre- vention and control have been set up in temperate regions. Influenza surveillance activity is also maintained by the global network of laboratories (108 National Influenza Cen- ters in 73 countries) sponsored by the World Health Orga- nization (WHO). These laboratories participate by circulat- ing current influenza viral isolates and identifications to the biannual recommendations for the composition of vaccines. Senegal is the only West African country involved in this network. At the National Influenza Laboratory (Institut Pas- teur) the isolation of the influenza type B virus in December 1995, after a 4-year absence of detection, prompted us to conduct a more vigorous epidemiological influenza survey in Senegal partially in order to contribute to a better under- standing of the epidemiology of influenza in tropical re- gions. 4 MATERIALS AND METHODS Description of the study site. Three health units located in the town center and 4 in the suburbs of the Dakar region of Senegal were selected as sentinel sites for the study (Fig- ure 1). A large part of the population in the city is served by these health units. Of the patients taking medical advice in the 3 sites located in the town center, most were children under 15 years old, whereas the 4 suburban sites served pa- tients from all age groups. All the sentinel sites had the ad- vantage of drawing patients from different neighborhoods in the area. The average weekly number of patients for each unit was 174. Study subjects. Subjects included all patients who pre- sented influenza-like symptoms from June 1996 to Decem- ber 1998. These symptoms included a combination of fever, sore throat, and nasal discharge within the past 4 days. A standardized questionnaire was used to obtain demographic data, medical history, and clinical features for each patient. Throat or nasal swab cultures were collected at the same time. Data collection. To assess the magnitude of influenza-like illness, data on the number of acute respiratory infections (ARI) were recorded weekly by physicians in 6 of the 7 sentinel sites. ARI definitions included all acute infections involving the respiratory tract, except chronic and allergic infections, tuberculosis, and asthma. The proportion of ARI was calculated by dividing the number of ARI cases by the total number of patients. Proportions for Week 4 were cal- culated as a mean of 4 weeks. Data concerning rainfall, tem- perature, and relative humidity in Dakar were recorded monthly from the national climatology office in order to study the influence of climatological factors on the spread of influenza in Dakar. Identification and characterization of viral isolates. Specimens collected in a tryptose phosphate broth were in- oculated on Madin Darby Canine Kidney (MDCK) cell lines in a medium which contained trypsin. 5 Supernatant cultures at 34°C were subjected to hemagglutination tests with goose red blood cells (0.5%) as soon as a cytopathic effect (CPE) was evident or after a 7-day incubation period in the absence of CPE. Supernatants with a negative hemagglutination re- action were reinoculated into fresh cultures and tested as be- fore. Supernatants with a positive hemagglutination reaction were further characterized by hemagglutination inhibition as- say using antiserum to type B and A subtypes (H1N1, H3N2). Antigenic characterization and sequencing of the isolated vi- ruses were performed at WHO Collaborating Centers for Ref- erence and Research on Influenza at the London Medical Re- search Council and also at the Institut Pasteur in Paris).