Epidemiologic and Economic Analyses on Highly Pathogenic Avian Influenza H5N1 in Nigeria and Egypt

Abstract

Influenza A viruses have caused several devastating outbreaks in poultry with some zoonotic infections and deaths in humans. To control the viruses and their continuous circulation in poultry population, an understanding of the epidemiology of the viruses is needed. An evaluation of the situation of avian influenza A H5N1 in two African countries most affected by this virus was conducted. To that end, two aims were pursued: (i) to identify the factors that played important roles in the dissemination and circulation of the virus, and (ii) to explore the potentials of the virus from becoming endemic in the poultry populations in Africa. The analyses partly described the extent of H5N1outbreaks in Nigeria in at-risk and non-infected premises and live bird markets in the country. All samples (tracheal and cloacal swabs, parenchymatous tissues and sera) were analysed using virus isolation, reverse-transcription polymerase chain reaction and serology. Primary data on influenza A H5N1 in human as recorded by the Egyptian government was collected and analysed using different variables and risk factors for HPAI H5N1 virus infection in poultry farms in Nigeria during 2006-2007 outbreaks in Nigeria were explored using the conditional logistic regression models. Factors that supported infections in new premises included: receiving visitors on farm premises, purchase of live poultry/products, farm workers that live outside the premises. Because biosecurity is important to reduce risks of infection in poultry farms, the effects and costs associated with the implementation or neglect of biosecurity, as well as the feasibility of implementation in the household poultry in Egypt were evaluated. Risks of particular importance for the household poultry were categorised into people-related, environmental-related and other birds/animal-related risks. Biosecurity measures were compatible in the household poultry and it is 8.45 times better to implement biosecurity than to do nothing against HPAI H5N1. The financial risk analysis was robust and withstood sensitivity analysis. The change that may occur in the course of the household poultry project was accommodated profitably. The effect of road network and its contributions to the epidemics in Nigeria were evaluated to determine whether HPAI H5N1 exhibit properties of network theory. Spatial aggregation of cases (disease clusters), links among similar ‘nodes’ (assortativity), simultaneous activation of similar nodes (synchronicity), epidemic flows moving from highly to poorly connected nodes (directionality), and a Pareto analyses pattern were observed including synchronicity and directionality of spread properties. An exploratory analysis was conducted on the probability of human infection through the oral contacts with A-H5N1 contaminated meat. Detailed quantitative risk assessment using predictive microbiology-process risk model revealed that up to 15,159 humans may have contracted HPAI H5N1 in Africa, with 1,964 deaths. These figures are higher than those officially reported. The model has limitations but bias associated with official epidemiological statistics which are often prone to underreporting, lack of awareness especially in the rural communities and censoring effects may also account for this discrepancy. Finally, continuous monitoring of influenza viruses in Africa and implementation of measures that should reduce the burden of infection in Africa is necessary.