PATTERNS OF AVIAN INFLUENZA VIRUSES CIRCULATION IN POULTRY SOLD AT LIVE BIRD MARKETS AND FACTORS ASSOCIATED WITH THE RISK OF AVIAN INFLUENZA VIRUSES IN BANGLADES

Abstract

H5N1 virus has caused repeated outbreaks in the poultry sector in Bangladesh. More than 550 reported HPAI H5N1 outbreaks in poultry and eight human cases, including one death, have been reported since 2007. In Bangladesh, multiple poultry species including domestic chickens, ducks, geese, pigeon and quail are reared together in backyard farms and sold in live bird market (LBM), facilitating avian influenza viral transmission between species. Though many epidemiological studies from Bangladesh have been conducted to detect avian influenza viruses (AIVs) in poultry, long term surveillance data on avian influenza prevalence are more useful and effective for local preparedness and control activities. We have been conducting LBM based surveillance since 2007 to evaluate AIVs circulation among poultry reared in different systems in Bangladesh. Here, we described the pattern of avian viruses circulation among domestic poultry from August 2007 to December 2016. A pair wise Spearman correlation coefficient was calculated to identify the role of climatic factors on the occurrence of AI. Another study was conducted to identify the association between biosecurity practices and environmental contamination with AI viruses. Over the surveillance period (2007-2016), Avian influenza (AI) A viral RNA was detected in 6% (95% CI: 5.7-6.8) waterfowl, 3% (95% CI: 2.9-3.9) commercial chicken, 2% (95% CI: 1.2-2.5) backyard chicken and 29% (95% CI: 27.2-31.3) environmental samples. Subtypes H5N1, H9N2, H11N3, H4N6 and H1N1 viruses were commonly detected. We detected 5 clades for H5N1 viruses; clade 2.2, 2.3.2, 2.3.4, 2.3.2.1 and 2.3.2.1a. Among the all detected clades, clade 2.3.2.1 was the predominant one, circulating since 2011. Among the waterfowl, ducks were more likely to be positive for AI A viruses compared to geese (OR 3.6, 95% CI: 2.3-5.7). The domestic waterfowl which were sampled during winter season were more likely to be tested positive for influenza A viruses compared to the waterfowl sampled during summer and monsoon (OR 2.4, 95% CI: 1.9-3.1). Among commercial chicken, Cobb type broiler chicken were more likely to be positive for AI A viruses than broiler, layer and breeder (OR 9.8, 95% CI: 3.1-31.1). Environmental samples collected from urban LBMs were more positive for AI A/H5 than the rural or peri-urban LBMs specimens (OR 2.2, 95% CI: 1.7-2.8).2 In waterfowl, there was a single annual peak of AIV occurrence that mostly observed in between December and January across different years. No clear seasonal pattern of AIV occurrence was observed in commercial and backyard chicken. However, there were two periodic annual signals of AIV occurrence were pronounced in environmental samples (September and February). In waterfowl, AIV circulation was negatively correlated with the climate variables (monthly average temperature, average humidity, total precipitation and average wind speed). In commercial chicken, AIVs circulation was positively correlated with humidity and precipitation, but negatively correlated with temperature and wind speed. In backyard chicken, AIV circulation was negatively correlated with all climate variables except wind speed. Poultry shops that slaughtered poultry within their shops (APR 1.6, CI: 1.1-2.3) and/or shops with unsold poultry from the previous day (APR 1.9, CI: 1.3-2.8) and/or absence of a weekly rest day (APR 1.2, CI: 1.1-1.4) and/or keep sick and healthy poultry together (APR 1.2, CI: 1.1-1.4) were more likely to harbor detectable AIV RNA. The findings of the surveillance suggest that avian influenza A viruses, including H5 and H9 subtypes, circulate year-round in poultry in Bangladesh. AI seasonality study suggest that circulation of AI was seasonal only in waterfowl. Avian influenza viruses were circulating in commercial chicken throughout the year indicating AI is endemic among commercial chicken in Bangladesh. Existing cleaning and disinfection practices were not significantly associated with decreased environmental contamination with AIVs. LBM based active surveillance provided valuable information about the status of AIVs in poultry in Bangladesh. Findings from molecular analysis will help public health policy makers develop and update candidate vaccine viruses for pandemic preparedness. Identification of certain risky biosecurity practices such as poultry slaughtering practices, management of leftover poultry, weekly rest day, cleaning and disinfection is necessary for improvement. Overall, this research will improve our understanding about AIVs dynamics and also help us develop intervention strategies for preventing and controlling AI in poultry and humans in Bangladesh