Epidemiology and risk factor analysis of avian influenza virus

• Κύριος συγγραφέας: Χατζηπροδρομίδου, Ιωάννα
• Άλλοι συγγραφείς: Βανταράκης, Απόστολος
• Μορφή: Thesis
• Γλώσσα: English
• Έκδοση: 2019
• Θέματα: Avian influenza virus; Greece; Epidemiology; Risk factor analysis; Boruta algorithm; Logistic regression; Outbreak; Ιός της γρίπης των πτηνών; Ελλάδα; Επιδημιολογία; Ανάλυση παραγόντων κινδύνου; Αλγόριθμος Boruta; Λογιστική παλλινδρόμηση; Επιδημική έκρηξη; 614.518

Background: The avian influenza virus may cause severe outbreaks in the poultry population and may occasionally infect humans exposed to infected poultry. The relative occurrence of the disease varies upon the infectious agent and the subtype, the composition of the population exposed and prior exposure or lack of exposure of the population, as well as the temporal and spatial frame. According to the World Health Organization and the World Organization for Animal Health, notifications are required for all highly pathogenic subtypes, as well as for H5 and H7 low pathogenic avian influenza subtypes. Aim of the Phd thesis: In the first part of this dissertation, a systematic review was conducted to investigate avian influenza outbreaks and to explore their distribution upon avian influenza subtypes, country, avian species and other relating details, as no comprehensive epidemiological analysis of global avian influenza outbreaks from 2010 to 2016 existed. Another scope of this work was to evaluate and compare the epidemiological data recorded by the surveillance programmes applied in Greece, during the decade 2002-2011, by both medical and veterinary authorities. Finally, this dissertation aimed to quantify the relationship between HPAI occurrences in wild/aquatic birds and poultry species in Greece and the Anseriformes population, the Charadriiformes population, precipitation, temperature, poultry population density, human population density, distance to the closest highway and water collection and number of water bird mid-winter census counts, applying both multivariate and uni and multi-variate models through the logistic regression method and Boruta algorithm method. Results: Based on data collated from four databases (Scopus, Web of Science Core Correlation, PubMed and SpringerLink electronic journal) and a global electronic reporting system (ProMED mail), one hundred and seventy-three avian influenza virus outbreaks were identified and included in this review, alongside 198 ProMED mail reports. The results showed that most of the reported outbreaks occurred in 2016 (22.2%). These outbreaks were in China (13.6%) and referred to commercial poultry farms (56.1%). The most common subtype reported in these outbreaks was H5N1 (38.2%), while almost 82.5% of the subtypes were highly pathogenic avian influenza viruses. There were differences noticed between ProMED mail and the scientific literature screened. The Sentinel surveillance system was used to analyze influenza virus cases in humans, while a passive surveillance system was initiated by proper directive in 2005 onwards. Data retrieved by competent officials showed that most of human cases were both of type A and B, whereas type A samples were identified of subtype H1N1 and H3N2. Among the animal specimens tested for this specific study period, only 35 proved to be positive in 2006. All except for one positive animal case were identified as of subtype Highly Pathogenic Avian Influenza Virus H5N1, except for one, which was of subtype Low Pathogenic Avian Influenza Virus H6N2. Almost 45% of humans were of pediatric population and 100% of animals were wild bird species. The results of the logistic regression model, based on raw risk factors’ data, indicated that the wild/aquatic birds’ possibility of infection is mainly determined by water birds mid-winter census (WMC) counts, as well as it is positively associated with WMC counts and distance to closest transportation route. Concerning poultry outbreaks, human population density, coexistence with mammals and distance to water collection proved to be statistically associated, too. Based on the log transformed risk factors data concerning both aquatic/wild birds and combined (aquatic/wild and poultry species) datasets, WMC, Anseriformes and Charadriiformes had statistically significant correlation with the number of HPAI outbreaks, while i) coexistence with mammals proved to be statistically significant in case of aquatic/wild birds dataset and ii) human population density and temperature seem to play a statistically significant role only in case of combined dataset. The Boruta algorithm method indicated that all ten risk factors were important for both species. The spatial pattern of the outbreaks was investigated and a map indicating the spatial hot- /cold-spots of HPAI occurrences in Greece was generated. Conclusions: Avian influenza virus has been proved to be able to contaminate all types of avian species, including commercial poultry farms, wild birds, backyard domestic animals, live poultry, game birds and mixed poultry. The study focused on wet markets, slaughterhouses, wild habitats, zoos and natural parks, in both developed and developing countries. The impact of avian influenza virus seems disproportionate and could potentially burden the already existing disparities in the public health domain. Therefore, collaboration between all the involved health sectors is more than necessary. No correlation of influenza types between humans and animals was observed, in the case of Greece. The study demonstrates that environmental, weather and anthropogenic factors may contribute to the avian influenza virus disease spread among animals and humans. These findings are considered as highly important for the global surveillance of HPAI occurrences.