Study of the Determinants and Patterns of HPAI H5N1 in North Vietnam Methodology, progress and initial results GRIPAVI PROJECT S.Desvaux, N. Gaidet, P.T.T.Hoa, N.T.Dung
GRIPAVI project in Vietnam: Vietnam Observatory Context H5N1 outbreaks declared since 2003 (poultry + transmission to Humans) Vaccination adopted to face the non clinical expression on waterfowls High humain and poultry density in the Delta area More than 200 millions poultry in diverse production systems Complex poultry commodity chain, many actors Resident and migratory wild birds population limited
Our question «What are the determinants and the patterns of introduction, diffusion and persistence of the highly pathogenic avian influenza virus H5N1 in the Red River Delta in North Vietnam and the consequences in terms of surveillance and control?» Use the results of epidemiological studies and models to propose an efficient and targeted surveillance model for HPAI in Vietnam
Our hypothesis Both persistence and regular introductions occur Low circulation may happen on some domestic poultry populations (long cycle) Some wild birds species sharing the same habitat may be part of the ecological cycle Environment may play a role in the transmission
Our approach Description of the system Commodity chain study Risk analysis: risk of introduction from China Classical epidemiology: Retrospective studies Prospective study Better understanding of vaccination efficiency Virtual laboratory: Multi-Agent Based modeling in coll with IRD
Retrospective analysis in the RRD Objective: identify factors related to the introduction or the persistence of the virus Descriptive analysis and statistical analysis at different scales 5 districts = 39 cases in 2005/2007 North = 5603 communes = 549 outbreaks Only description Delta = 3075 communes = 394 outbreaks Description + variables testing Cases control at village and farm levels
Retrospective analysis in the variables tested RRD Delta = 3075 communes V. Extracted from GIS: Densité moyenne des routes nationales Densité moyenne des routes nationales + chemin de fer Densité moyenne des routes (toutes les routes) Distance moyenne aux routes nationales Distance moyenne aux routes nationales + chemin de fer Distance moyenne aux routes (toutes les routes) Densité moyenne des rivières principales Densité moyenne des rivières Distance moyenne aux rivières principales Distance moyenne aux rivières Durée d inondation moyenne V. Extraction from MODIS picture interpretation: Pourcentage des pixels sans interprétation possible Pourcentage des pixels occupés par une culture de riz annuelle Pourcentage des pixels occupés par deux cultures de riz annuelles Pourcentage des pixels occupés par de la foret Pourcentage des pixels occupés par de l eau permanente. Other data Human density 5 districts = communes = 803 villages 18 case villages / 39 control villages 19 case farms / 18 matched control farms Idem Delta at commune level + vaccination coverage + pop at risk + soil typology V. Extracted from GIS: Idem Delta V. Extraction from SPOT picture interpretation: % pixels occupés par de l eau % pixels occupés par de la foret % pixels occupés par du riz % pixels occupés par des habitations % pixels occupés par de l eau ou du riz Durée d inondation moyenne V. collected on the field Village level Prsence-distance market / presence trader / presence hatchery Soil occupation Human pop / no HH with poultry No farm / type farm arm level Housing system / vaccination status / taille / proximité trader / habitude de vente
Retrospective analysis in the Delta Objective: identify factors related to the introduction or the persistence Initial results: of the virus W3: 1,6 3,6% Average number of outbreak / commune Descriptive W5: 1,27 1,1% analysis and statistical analysis on Risk at commune level 5 districts W6: 1,05 0,7% Some areas concentrate outbreaks but at commune level, very few reoccurrence impact of the control measures? Some provinces with very good barriers to introduction?? Perspectives: Descriptive spatial analysis still needs to be performed + test variables extracted from GIS and satellite image
Sol occupation typology: a way to explore environmental factors to be extented on more communes to test the risk of getting HPAI
Case-control study in BG Case control study testing different variables related to the human and natural environment and to the poultry production system: data collected (20 villages 40 non matched controls, 19 farms - matched controls), analysis on going Case Initial definition: results: Case definition arm level applied on the lists from different sources: 44% of the cases said to be vaccinated arm having Size of reported the farms mortality seems to whatever be a risk factor: the percentage higher size or for farm cases with (47% laboratory of the confirmation cases > 800 heads reported. ; 100% control <800) Case definition Proximity applied with traders: on the field 40 % for cases confirmation: farms are located <200 m from trader house (25% for control) Acute disease Village level Average mortality density over 10 of the % road within > in 1 day cases Average + nervous distance signs to on the ducks main river < cases Perspectives: rapid test + (usually not reported on the list) Regression model to be designed
Prospective study 2 provinces: Bac Giang, Ha Tay, 4 districts 9 communes 84 villages Around 700-900 flocks
Wild birds?? Eau Environnent?? aune sauvage?? Domestic poultry Long cycle Short cycle Support solide?? Véto Basse-cour?? Peri-domestic wild birds Elevage cycle long?? Pig Elevage cycle court Prospective study (sero/viro monitoring) Identify Ecloseriespotential seasonal patterns for the incidence of H5N1 in the different compartments + try to link these patterns to seasonal cycles and other human interventions. Test hypothesis of the persistence of the virus in the population of poultry throughout a year against a regular reintroduction using bio molecular Marché Petit collecteur Gros collecteur analysis of the strains collected. Identify the local factors for the diffusion intra and intercompartments. ormulate hypothesis related to the virological ecology (comparing the strains detected on the different compartments) Consomateur Chine
Prospective Protocol Methodology for study: longitudinal methodology study Domestic Domestic poultry poultry Repeated Repeated population-based population-based cross-sectional cross-sectional surveys surveys on on on domestic domestic poultry poultry (around (around 1000 1000 1000 birds birds birds / survey) survey) / OK (4357 birds sampled) requency: - 4 to 4 to 6 surveys 6 over over 1 year 1 year - Unit = flock or village Virology - and and serology (exploration of of ELISA against NS1 NS1 protein) - sampling + questionnaire Sampling - Random of of highly selection connected for each markets survey / Nested stratification case-control by cycle study study type: at at farm long farm level / short level within the the study study population (sampling frame of the farms reviewed for each C) Estimation - requency: of of the 4 the surveys risk risk of of getting over 1 AI year AI for for a subset a of of the the study study population - Virology followed and serology over over the the testing different in cross-sectional progress surveys Outbreak Sampling investigation of communal markets OK (17 market visits, 287 pooled samples = around 1250 birds) Nested case-control study at farm level within the study population: NO Outbreak investigation: NO Wild birds: separated presentation
Database
Prospective study: variables General: localisation (adm unit, no GPS) / date arm: Housing system / Water source DOC source / trader OUT Vaccination status for HPAI and NCD Breed / age / production type (breeder, layer, broiler, backyard) Village / Commune Number of HH Poultry pop Presence of trader / live birds market / hatchery Soil occupation Vaccination «coverage» Morbidity / mortality events Market: Species / age Birds origin (localisation) Seller (farmer / trader) prices
Prospective study: methodology for laboratory testing Screening for domestic sampling Serology: HI test (pos if >4log2) Virology: RT-PCR type A Validation serology Elisa Id Vet + HI tests for different sub-types Confirmation virology All positive and doubtful results: egg inoculation + RT-PCR for H5 if inoculation negative 4357 samples Negative for H5 with HI test 1104 samples +ve for H5 with HI test ELISA Type A Id Vet Estimation of the seroprevalence for type A virus Negative samples Positive samples Elisa H5 and H7 HI for H4, H6, H3,H9 -Control the correlation between H5 HI positive samples and ELISA H5 positive results. -Identify other main circulating sub-type Newcastle for chickens (1840 tests): Elisa indirect VETNDV/I
Prospective study: initial serological results % birds >=4log2 50,00 50,00 40,00 40,00 30,00 30,00 20,00 20,00 10,00 10,00 0,00 0,00 % birds >= 4log2 Evolution of the seropositivity per campaign and species 1 2 3 1 2 3 Campaign Campaign A_layer_breeder B_layer_breeder chickens ducks Percentage of birds with antibodies >=4log2 in each farm: In average, the farms with positive birds have around 40% of the birds positive (ranging from 1 birds to all birds sampled being positive). Elisa type A: 47% seropositivity on a sample fron C1 and C2: which sub-type? (not H7)
Commodity chain survey Protocol for longitudinal study Objectives: - describe the networks for poultry trading - identify different networks structures and test their possible influence on the risk of HPAI outbreaks - identify highly connected nodes for sampling: NO Initial results of the Commodity chain survey: - Interviews conducted on 240 farms, 60 backyard farmers group meetings, 60 traders and markets - HACCP applied to the chain (C.Lebas) - Models for poultry flow + virus dynamic propagation (Master student) - Descriptive statistics on going (typology): seasonal variation; trading habit, at risk practice (S.Desvaux) - irst matrix built to apply SNA (R.Duboz / S.Desvaux)
3 databases: farm / backyard / trader
Paramètres de contact Dynamique des flux de produits Circulation de produits dans le réseau 1 2 i 1 2 i 1 1 2 2 i i Transmission du virus dans une métapopulation (1 compartiment=1 unité épidémiologique) Dynamique de propagation du virus 1 2 1 2 1 2 1 2 1 1 2 2 Modeling virus propagation (Ariane Payne, Stéphanie Desvaux, Karine Chalvet-Monfray, Jean-rançois Renard, Dominique J. Bicout)
Résultat: structure du réseau Schéma des différentes voies de commercialisation des produits. ilière avicole semi commerciale ( échelle communale) erme x 2 1-x 2 1- p (1-p) (1 -x d ) Marchéé z (1-p)x d 1-z p px c Marchand collecteur y consommateur p(1-x c ) x 1 1-x 1 1-y 21 Grossistes/d étaillants
Vaccination effectiveness study A population of ducks (super egg breed) and chickens (Luong Phong breed) followed in the field for serological answer (T0 =before vaccination, W3= 3 weeks post vaccination, W6 = 6 weeks after first injection or 3 weeks after second injection = 3 months after first injection, M4, M6 etc. Initial population made of 6 breeding farms (90 birds): we expect to get at least 50% of the birds covered for the full period. Started in March 09 and should be completed by January 2010
Risk of introduction from China To be starting from November Obj: Describe the flow of poultry coming from China (old hens specially) Quantify this flow? Estimate the risk of introduction of HPAI in the domestic poultry population due to the (illegal) import of infected poultry from China
Wild Birds study
Outline Vietnam context Research question Wild bird species studying Sampling campaigns Sampling methods Initial results Ongoing samples analysis
Vietnam context Endemic Avian Influenza (H5N1) in poultry population Duck scavenging in rice field Egret, Heron share the wetlands habitat with ducks Chicken scavenging in villages Peri-domestic wild bird species available in villages and rice fields
Research question Can peri-domestic wild bird species be infected with HPAI H5N1 in a context of virus circulation? What is sero-prevalence of antibodies against HPAI H5N1 Avian Influenza virus in peri-domestic wild bird species? What is prevalence of Influenza A virus and HPAI H5N1 Avian Influenza infection in peri-domestic wild bird species? Do the AI viruses circulating within wild waterbird, peridomestic wild bird and domestic birds have phylogenetic relations?
Wild Protocol bird for species longitudinal studying Scaly-breasted Munia (Lonchura punctulata) Chinese Pond Heron (Ardeola bacchus) Tree Sparrow (Passer montanus) little Egret (Egretta garzetta)
Wild bird sampling campaigns Protocol for longitudinal study Wild bird species Sparrow Aug 08 Sep 08 Oct 08 Nov 08 Dec 08 Jan 09 eb 09 Mar 09 April 09 Availability x x x x x x x Sampling x x x Munia Study areas: Bac Giang, Ha Tay (mainly in Bac Giang) Availability x x x x x Sampling x x Heron/Egret Availability x x x x x x x x Sampling x x x x x Breeding season May 09 x x x Campaign C1 C1 C2 C2 C2 C3 C3 C3 Jun 09 July 09
Sampling methods - Wild bird species sampled following season of birds abundance - Number of birds/species/campaign: 100-200 birds -Blood, trachea and cloacal swab samples were collected for each bird -Pooled trachea and cloacal swabs for Heron/Egret and pooled 5 trachea/cloacal swabs for Munia/ Sparrow - eces collection in roosting side every two months
Protocol Initial for longitudinal results study * Number of wild bird samples Campaign\ bird species Munia Sparrow Heron/Egret Total Campaign 1 205 153 28 386 Campaign 2 203-7 210 Campaign 3-200 130 330 Total 408 353 165 926 * ELISA test results - 456 sera (34 Herons/Egrets, 306 Munias and 116 Sparrows) were tested by Indexx Elisa. - All tested serum give negative results - Initial results show that healthy Munia, Sparrow and Heron/Egret collected in Vietnam have no antibodies to Influenza A
Ongoing and future samples Protocol for longitudinal study analysis eces samples from roosting site will be inoculated to increase the chance of virus isolation Tracheal and cloacal swabs: no testing if all serological samples negative? Newcastle serology?
Thank you for your attention!