Avian influenza transmission risk along live poultry trading networks in Bangladesh

Abstract

Live animal markets are known hotspots of zoonotic disease emergence. To mitigate those risks, we need to understand how networks shaped by trading practices influence disease spread. Yet, those practices are rarely recorded in high-risk settings. Through a large cross-sectional study, we assessed the potential impact of live poultry trading networks’ structures on avian influenza transmission dynamics in Bangladesh. Networks promoted mixing between chickens sourced from different farming systems and geographical locations, fostering co-circulation of viral strains of diverse origins in markets. Viral transmission models suggested that the observed rise in viral prevalence from farms to markets was unlikely explained by intra-market transmission alone, but substantially influenced by transmission occurring in upstream network nodes. Disease control interventions should therefore alter the entire network structures. However, as networks differed between chicken types and city supplied, standardised interventions are unlikely to be effective, and should be tailored to local structural characteristics.

H9N2 avian influenza virus dispersal along Bangladeshi poultry trading networks

“…We used the six pairwise combinations of chicken types (sunali, deshi, and broiler) sold in Dhaka and Chattogram as the six discrete traits. The GLM predictors included (i) the number of sequences associated with each of the six categories as a proxy for sampling effort (both origin and destination), (ii) the weekly estimated sales at LBMs for all six categories (both origin and destination) as detailed in Moyen (2019) and Moyen et al. (2021) , (iii) a binary chicken type similarity index of the six discrete traits (i.e.…”

“…same chicken type = 1 and different chicken type = 0), (iv) an equivalent city similarity index, and (v) the extent of overlap between production areas from which birds are sourced (1, Pianka index) between each discrete trait. These production areas are the set of Upazilas (sub-district) where farms supplying markets within a city are located, which were previously computed from reconstructed transaction networks based on traders’ interviews as described in Moyen et al. (2021) .…”

“…Here, poultry is either sold directly to end-users or may be traded further between market vendors before sale ( Moyen et al. 2018 , 2021 ; Moyen 2019 ; Høg et al. 2021 ).…”

“…Recent studies show that poultry trading practices vary substantially across the network ( Moyen et al. 2018 , 2021 ; Moyen 2019 ). The two largest cities in Bangladesh, Dhaka and Chattogram, contain numerous LBMs where AIV infections have been consistently reported and receive poultry from largely non-overlapping regions of Bangladesh (henceforth, ‘production areas’) ( Moyen et al.…”

Mapping of dressed and processed poultry products in Bangladesh: identifying the food safety risks for policy intervention

“…Most poultry, including exotic broiler chickens, are supplied to consumers in Bangladesh via live bird markets (LBMs) (Moyen et al 2021) where birds are slaughtered and processed on demand at the point of sale. Bangladesh's rst commercial company-owned slaughterhouse became operational in 1998.…”