Epidemiology of influenza in Ethiopia: findings from influenza sentinel surveillance and respiratory infection outbreak investigations, 2009–2015

Influenza viruses are enveloped viruses of the family Orthomyxoviridae that contain a segmented RNA genome. Influenza is an acute viral disease of the respiratory tract which is characterized by some or all of the following sign and symptoms: fever, headache, myalgia, prostration, coryza, sore-throat and cough [1]. Of the three types of influenza viruses, A, B, and C, the first two are associated with significant seasonal morbidity and mortality [2, 3]. Influenza A is more prevalent and leads to greater mortality in humans than influenza B [4, 5].Type A influenza has subtypes that are determined by the surface antigens hemagglutinin (H) and neuraminidase (N) [6].

Globally, an estimated 3 to 5 million cases of severe influenza illness [7] and 291 243–645 832 seasonal influenza-associated respiratory deaths occur annually [8]. In Africa, influenza-related incidence data remain inadequate and outcome data are still very limited despite of some surveillance information [9].

In temperate regions, seasonal influenza typically occurs every year in the late fall or winter [10] while the seasonality is less clearly defined in tropical and subtropical regions [11]. Seasonality patterns of influenza in eastern Africa including Ethiopia, have not been clearly established [12]. However, some evidences from different studies indicated that Influenza A (H1N1)pdm2009, Seasonal Influenza A(H3N2) and Influenza B are circulating in different countries of sub-Saharan Africa [13‐15].

To address this gap and generate evidence in Ethiopia, severe acute respiratory infections (SARI) and influenza like illness (ILI) sentinel surveillance was established in November 2008 [16]. The Ethiopian Public Health Institute (EPHI) has adapted and prepared an influenza sentinel surveillance implementation manual [17]. The establishment of influenza sentinel surveillance is aimed to detect new influenza strains capable of or having the potential to cause a pandemic, determine the characteristics of influenza, characterize and monitor trends in illnesses and deaths attributable to severe acute respiratory infections, determine the proportions of confirmed cases of influenza among SARI in-patients and/or among ILI out-patients and finally to provide information on the contribution of influenza to the burden of respiratory diseases in order to prioritize resources and plan appropriate public health interventions in Ethiopia.

From January 1, 2009 through December 31, 2015, a total of 4962 patients which fulfilled ILI and SARI case definitions were enrolled and throat samples collected at designated influenza sentinel site health facilities and from respiratory infection outbreak investigations in different parts of the country (Fig. 1). The mean age of the sampled cases was 15 years (Range: 1 month to 82 years) and the median age was 11 years (Inter quartile range 2–25 years). Of the total, 4799 (96.7%) were tested for influenza by RT-PCR and 988 (20.6%) were found to be positive for different influenza viruses. Among the positive cases 349 (35.3%) were caused by seasonal influenza A (H3N2), 321(32.5%) by influenza A (H1N1) pdm2009 and the remaining 318 (32%) by Influenza B. The mean age of the influenza positive cases was 19 years (Range: 2 months to 72 years) and the median age was 17 years (Inter quartile range 8–26 years). The positivity rate was 5% in persons under 2 years, 10.3% in 2–4 years, 29.5% in 5–14 years, 26.3 in 15–49 years, 21.1 in 50–64 years and 16.7% in person greater than 64 years (Table 3). Of the total positive cases 568 (57.5%) were females with a positivity rate of 21.3% and 420 (42.5%) were males with a positivity rate of 19.7%. Among total 4962 specimens collected, 3369 (67.9%) were ILI and 1381 (27.8%) were SARI cases while the remaining 212 (4.3%) were sampled during different respiratory disease outbreaks.

In this article we have presented seven years (2009–2015) of virological and epidemiological data from influenza sentinel surveillance in Ethiopia as well as results from respiratory disease outbreak investigations carried out in the country over the same period. During this time 4799 samples (representing 96.7% of all samples collected) were tested. The majority of samples which were not tested were obtained at the very start of the surveillance activities. Due to lack of viral transport media at the time, these early samples were being collected in absolute ethanol. Samples had mostly evaporated before any routine testing could begin. A total of 988 influenza positive samples were found over the seven-year period, giving a total influenza positivity rate of 20.6%. The findings indicated that seasonal Influenza A(H3N2), Pandemic Influenza A(H1N1) and Influenza B viruses were circulating in Ethiopia. Seasonal influenza A (H3N2) accounted for more than one-third (35.3%) of all influenza positive cases, followed by Influenza A(H1N1)pdm2009 (32.5%) and influenza B (32.2%).

The data from 2012 to 2015 indicated that influenza positive cases and positivity rate were predominantly identified in November (190 cases, 37.5%). Seasonal influenza A(H3N2) and Influenza A(H1N1)pdm2009 peaked in the period from September through January, with November being the month with the highest number of influenza positive cases. Similar transmission patterns have been reported for seasonal influenza A (H3N2) and influenza A (H1N1)pdm2009 viruses, supporting the idea that the spread of the Influenza A (H1N1)pdm2009 virus was similar to that of seasonal influenza A viruses [21]. However influenza B was predominantly identified in the period from November to April with a peak (67 cases, 22.1%) during March.

The influenza positivity rate was highest from September to February, mainly in November and December which are cool and dry months in Ethiopia [20, 22]. Since the data presented is predominantly from Addis Ababa (Fig. 1), the type of weather conditions described (cool and dry) is applicable for this locality during these months. The months from September to November normally correspond to the autumn season in temperate areas of the northern hemisphere. In such places the late autumn season coincides with when there is increased circulation of seasonal influenza viruses [2].

We found that the crude influenza positivity rate among ILI patients was 26.8% (21.9%, 28.7% and 25.3% for Akaki, Kolfe and Shiromeda health centers respectively). Compared to the influenza positivity rate from ILI samples in Niger (12%) and Gabon (11.4%), the influenza positivity rate among ILI samples was high in Ethiopia [23, 24]. It was very similar to the findings in Kenya, 26.7% [25] but lower than in Rwanda, 29.2% [26]. The influenza positivity rate for SARI samples was 3.1% which was low compared to findings in South Africa and China which reported 8% and 6% respectively [27, 28]. Our data showed that relatively the influenza positivity rate (27.9%) from outbreak sites was high.

The finding from ILI sentinel sites revealed that influenza positivity rate was high among children aged 5–14 years (33.2%) followed by individuals 15–44 years (26.8%) and among those older than 44 years (21.9%). Amongst SARI patients the influenza positive cases were very few, only 39 (3.1%). The majority of influenza positive cases from the SARI surveillance were identified among under-five children. Since there is no routine screening of other respiratory pathogens in either the ILI or SARI surveillance, it is not clear if the low influenza positivity rate in SARI cases might mean that most cases are being caused by pathogens other than influenza. Conversely it may indicate that the quality of SARI surveillance in Ethiopia might need significant improvement.

The primary limitation and challenges experienced by the Ethiopian Influenza sentinel surveillance system were the weak follow up of influenza-positive cases to determine the outcome of infection with influenza virus. The surveillance system has also enrolled only few SARI patients whereas the proportion of SARI as compared with total admission cases was high in all hospitals. The surveillance system is also capturing only influenza.

In Ethiopia, seasonal Influenza A (H3N2), Influenza A (H1N1)pdm2009 and Influenza B are circulating. Both influenza positivity rate and number of presenting cases were predominantly observed in November showing a distinct seasonality. The surveillance has also revealed that influenza A (H3N2) and Influenza A (H1N1)pdm2009 were predominantly circulating from September through January and peaked in November while influenza B was predominantly detected from November to April with a peak in March. Hence, influenza is found one of the public health problems in Ethiopia. Introduction of vaccine and influenza antivirus is important to prevent and treat infected cases for future.

Quality influenza surveillance systems are needed to enable countries to better understand influenza epidemiology, including disease incidence and severity, and help them implement appropriate prevention strategies. A system has to be established to make a follow up on SARI cases which turned out to be positive for influenza so as to be able to determine outcome of influenza infections in all SARI cases. We also propose the introduction of laboratory diagnosis of other respiratory pathogens, especially on influenza negative samples so as to determine the proportion of other respiratory pathogens causing ILI or SARI. Finally, we recommend the inclusion of more influenza sentinel sites from different geographical and ecological areas of the country to generate more data which would be more representative of the influenza virus activity across the country so that more appropriate and comprehensive influenza control strategies are designed and implemented.