- Email: [email protected]
Evidence for freedom from swine influenza in a remote area of Northern Vietnam
Acta Tropica 122 (2012) 160–163
Contents lists available at SciVerse ScienceDirect
Acta Tropica journal homepage: www.elsevier.com/locate/actatropica
Short communication
Evidence for freedom from swine influenza in a remote area of Northern Vietnam K. Trevennec a,b,∗ , V. Grosbois a , F. Roger a , Thu H. Ho c , C. Berthouly-Salazar d , V. Chevalier a a French Agricultural Research Center for International Development (CIRAD), Animal and Integrated Risk Management Research Unit, Baillarguet Campus (AGIRs), 34398 Montpellier Cedex 5, France b École Nationale Vétérinaire de Toulouse (ENVT), INP, Toulouse, France c National Institute of Veterinary Research (NIVR), Viet Nam d Centre for Excellence of Invasion Biology (CIB), Department of Botany and Zoology, Stellenbosch University, South Africa
a r t i c l e
i n f o
Article history: Received 9 September 2011 Received in revised form 7 November 2011 Accepted 23 November 2011 Available online 2 December 2011 Keywords: Epidemiology Influenza type A Backyard Swine Vietnam Cross-species transmission Avian
a b s t r a c t Swine influenza is responsible for one of the most prevalent disease affecting the swine industry worldwide. Epidemiological surveys rarely focus on remote areas, because traditional farming systems characterized by locally consumed production and low pig densities are considered as having little influence on the emergence, re-emergence, persistence or spread of swine influenza viruses. In addition, routine disease investigations in remote areas are often neglected due to logistic and economical constraints. A bank of swine sera collected in 2005 in the ethnic minorities households of Ha Giang province (Northern Vietnam) located adjacent to the Chinese border was analyzed to estimate the seroprevalence of swine influenza (SI) and to identify potential risk factors for infection. The results suggest that this specific agro-ecological system is free from SI and is not favourable to SI spread either through pig-to-pig transmission, or through poultry-to-pig transmission. © 2011 Elsevier B.V. All rights reserved.
1. Introduction Swine populations are the natural reservoir of H1N1, H3N2 and H1N2 influenza subtypes (Brown, 2000). These 3 subtypes are reported in China (Liu et al., 2011; Vijaykrishna et al., 2011) and in Thailand (Chutinimitkul et al., 2008; Takemae et al., 2008). The avian-origin H9N2 and H5N1 highly pathogenic avian influenza (HPAI) have been occasionally isolated in swine, in China (Cong et al., 2008; Yu et al., 2008), and during outbreaks in Indonesia (Nidom et al., 2010). Since pigs are receptive to both avian and human influenza viruses, they play a crucial role as a mixing vessel in which strains may combine, which raises the risk of emergence of a pandemic influenza virus strain (Brown, 2000; Castrucci et al., 1993; Webster et al., 1977). Routine disease investigations in remote areas are often neglected due to logistic and economical constraints. A FrenchVietnamese research project, called Biodiva (www.biodiva.org.vn), which aimed to assess and promote genetic diversity in domestic and wild animals, collected a sero-bank in the remote area of Northern Vietnam located adjacent to the Chinese-Vietnamese border. In the framework of a global project on animal influenza viruses in
∗ Corresponding author at: French Agricultural Research Center for International Development (CIRAD), Animal and Integrated Risk Management Research Unit, Baillarguet Campus, 34398 Montpellier Cedex 5, France. 0001-706X/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.actatropica.2011.11.012
Vietnam, serological analyses were conducted in this low animal density area on both backyard poultry and pigs, during H5N1 outbreaks in 2005–2006. The seroprevalence of H5N1 was estimated at 7.2% [CI 1.4; 10.5] in chickens (Trevennec et al., 2011a). The aim of the present study is to focus on the swine compartment, in order to estimate the seroprevalence of swine influenza (SI) and to estimate the risk of pig infection with the avian-origin H5N1.
2. Material and methods 2.1. Study design The serological study was performed on sera of pigs bred in Northern Vietnam, which had been collected from April 2005 to August 2006. The study area was the Ha Giang province (22◦ 08–23◦ 19 N; 104◦ 33 –105◦ 33 E), sharing a 274 km border with China (Yunnan province). Ha Giang’s topography is characterized by sharp altitudinal variation. According to the government statistics office, the size of the swine population in the Ha Giang province in 2005 was around 330 000, corresponding to a pig density of 42 head/km2 (General Statistics Office of Vietnam, 2005), which represents a low density in comparison to those found in the highproducing areas of the Red River Delta or of the Mekong Delta (Fig. 1). The sampling design is detailed in the previous article on avian influenza (Trevennec et al., 2011a). Sampled villages are
K. Trevennec et al. / Acta Tropica 122 (2012) 160–163
161
is infected, the prior distribution is built from a maximal possible value and a most likely value. Based on personal communications provided by the test producer, SE and SP most likely values were estimated at 90% and 98%, respectively (Pourquier, 2010). Minimum possible values of SE and SP were chosen as to be weakly informative, thus 50%. The ELISA-A kit evaluation for pig sera is presently in progress and will be published subsequently after completion. Since preliminary results indicate a low SE (unpublished data), we also tested a second option based on a most likely value for SE at 70%. Regarding Pr, we set the most likely value at 5%, which ranges within the estimated prevalence of SI in low pig raising density in China (Liu et al., 2011) or in Northern Vietnam (Trevennec et al., 2011b). The calculator was used for binomial sampling, since the population size, estimated at 330 000 pigs within the province (General Statistics Office of Vietnam, 2005), was large relative to the sample size (Johnson et al., 2004). The posterior distributions of the model parameters are obtained through the Markov Chain MonteCarlo method with 5000 burin-in iterations followed by 20 000 iterations. A population is considered “disease-free” when the proportion of true positive animals is less than a selected threshold value. Various classical scenario (Table 1), corresponding to sporadic cases or endemic situation with a low seroprevalence were chosen. Thus, we focused on the posterior probability that the serological prevalence in the sampled population does not exceed cut-off values of 1%, 2%, 5% and 10%, given the number of positive tests in the sample (Johnson et al., 2004). 3. Results A total of 519 pig sera collected in 158 villages were analyzed (Fig. 2). On average 3.3 pigs from distinct farms (1 pig per farm) were sampled in each village, with a minimum of 1 and a maximum of 7 per village. Only 2 pigs had a positive ELISA-A test. The posterior estimation of seroprevalence was 0.43% [95% CI 0.02; 1.9] for scenario S1–S4 and 0.46% [95% CI 0.03; 1.9] for scenario S5–S8 (Table 1). The probability of the swine population being free from SI antibodies varied from 61%, for a prior SE of 70% and a cut-off value of 1%, to 100% for cut-off values equal to or higher than 5%. Fig. 1. Pig density by province in Vietnam (General Statistics Office of Vietnam, 2005).
mapped in Fig. 2. The pig sera were tested against influenza type A with an enzyme-like immunosorbant assay (ELISA-A) competition test IDVET©. This commercial kit is designed to specifically detect antibodies directed against the NP protein antigen that is highly conserved among all subtypes of influenza type A viruses. 2.2. Freedom of disease analysis for swine In order to account for the uncertainty in test performance and expected prevalence, a Bayesian statistical approach was adopted to demonstrate freedom from disease in swine, using Bayesfreecalc1© available at the UC Davis website (http://www.epi.ucdavis.edu/diagnostictests/). This software program assesses the probability that the prevalence in the sampled population exceeds a given cut-off value given the number of positive reactors, the total number of tested animals and prior knowledge on the performances of the test used and on the expected serological prevalence if the focal population were infected (Johnson et al., 2004). The uncertainty on input parameters is expressed in the form of prior distributions. Prior distributions are built for the sensitivity (SE) and the specificity (SP) from a minimum possible value and a most likely value. For the expected prevalence (Pr) under the hypothesis that the study population
4. Discussion Our findings indicate that the serological prevalence in the Ha Giang province in 2005–2006 was clearly less than 5% and probably less than 1%. When the seroprevalence is under a defined threshold, the disease is considered under control. The detection of sporadic cases will be attributed to isolated and probably independent events (i.e. animal introduction from usual area) without spread. The use of various cut-off values allows testing different scenario and may be a useful tool for health planners. However, this result has to be interpreted cautiously. Firstly, information on the performances of the ELISA-A test on swine is lacking. Further studies are needed to evaluate this test for swine. Secondly, this result contrasts with the estimated seroprevalence of SI published in the literature and the analysis of meta-data on swine influenza infection in Chinese swine populations, where seroprevalence of swine influenza A is estimated at 13.6% in low pig raising area (Liu et al., 2011) and 19.93% in the whole Southwest China, including the Yunnan province (Liu et al., 2011) that shares a border with our study area. The farming system of the study area also presents characteristics that are frequently described as risk factors of SI: farrow-to-finish breeding with mixing animals of different ages, and uncontrolled contacts with other-species (Olsen et al., 2006). Nevertheless, the probable absence of SI antibodies in the Ha Giang pig population may be explained by its agro-ecological
162
K. Trevennec et al. / Acta Tropica 122 (2012) 160–163
Table 1 Probability of freedom from swine influenza according to various scenario of prior information and cut-off values. Scenario
Prior information SE
Cut-off SP
Output P freedom
Pr
Min
Mode
Min
Mode
Max
Mode
High SE
S1 S2 S3 S4
50%
90%
50%
98%
20%
5%
1% 2% 5% 10%
64% 93% 100% 100%
Low SE
S5 S6 S7 S8
50%
70%
50%
98%
20%
5%
1% 2% 5% 10%
61% 93% 100% 100%
system characterized by the absence of well-described risk factors such as high pig and farm densities, high animal turnover or external source of breeding (Olsen et al., 2006). In the Ha Giang province, the pig density is one of the lowest in Vietnam, with only 42.7 pig/km2 (the national pig density in 2005 is 82.7 pig/km2 ) (General Statistics Office of Vietnam, 2005), farms are small, with on average 4–5 pigs and villages are scattered. As a consequence, exposure from neighbouring farms and neighbouring villages is
probably limited and local spread through aerosols is unlikely to occur. Furthermore, pig production is mainly devoted to family consumption so that the local pig trade is poorly developed. The average number of purchased pigs is less than 2 pigs/farmer/year and most of them are provided by family members, farmers of the same village or farmers from the same commune. Thus, we suggest that the lack of pig trade may protect pig farms from the introduction and spread of SI.
Fig. 2. Sampled villages (n = 158) and visited communes in Han Giang province, Northern Vietnam.
K. Trevennec et al. / Acta Tropica 122 (2012) 160–163
H5N1 outbreaks and a seroprevalence of 7.2% [CI 1.4; 10.5] were reported within poultry in the Ha Giang region. It can be concluded that even when exposed to H5N1 virus through possible direct close contact with infected backyard chickens, pigs are not likely to be infected by this avian influenza virus. These conclusions are consistent with results observed in a similar smallholders agro-ecosystem of Indonesia (Santhia et al., 2009). The same conclusion has also been reached in more intensive farming systems in Vietnam and Thailand where no evidence of avian-to-pig transmission of H5N1 have been observed during outbreaks in the poultry populations (Parchariyanon, 2006; Trevennec et al., 2011b). 5. Conclusion The present study in remote areas of Northern Vietnam suggests that this specific agro-ecological system is not favourable to SI spread either through pig-to-pig transmission or through poultryto-pig transmission. Low animal densities and the lack of pig trade may explain this low seroprevalence. However, pig production in Vietnam is developing quickly: facing an increase of pig trade and a potential intensification of the pig production, the epidemiological situation may evolve rapidly. A global description of the animal influenza situation and determinants in various farming systems will help to better understand epidemiological mechanisms in domestic animals. The present study is integrated within the framework of a global study of animal influenza in Vietnam, including specific surveys in the highproducing provinces of the Red River Delta. Acknowledgments This study was a part of the BIODIVA project and was jointly funded by the French Ministry of Foreign Affairs and the National Institute of Animal Husbandry of Vietnam. This study was partially support by the GRIPAVI FSP project (MAEE), and Vent d’Est scholarship (MAEE). We would like to also thank IdVet® lab for providing the ELISA-A kits and Marie Gély from CIRAD for her precious help on mapping skills. References Brown, I.H., 2000. The epidemiology and evolution of influenza viruses in pigs. Veterinary Microbiology 74, 29–46.
163
Castrucci, M.R., Donatelli, I., Sidoli, L., Barigazzi, G., Kawaoka, Y., Webster, R.G., 1993. Genetic reassortment between avian and human influenza A viruses in Italian pigs. Virology 193, 503–506. Chutinimitkul, S., Thippamom, N., Damrongwatanapokin, S., Payungporn, S., Thanawongnuwech, R., Amonsin, A., Boonsuk, P., Sreta, D., Bunpong, N., Tantilertcharoen, R., Chamnanpood, P., Parchariyanon, S., Theamboonlers, A., Poovorawan, Y., 2008. Genetic characterization of H1N1, H1N2 and H3N2 swine influenza virus in Thailand. Archives of Virology 153, 1049–1056. Cong, Y.L., Wang, C.F., Yan, C.M., Peng, J.S., Jiang, Z.L., Liu, J.H., 2008. Swine infection with H9N2 influenza viruses in China in 2004. Virus Genes 36, 461–469. General Statistics Office of Vietnam, 2005. Number of Pigs by Province, ed. GSO, http://www.gso.gov.vn/default en.aspx?tabid= 13/08/2010 469&idmid=3&ItemID=10092. Johnson, W.O., Su, C.-L., Gardner, I.A., Christensen, R., 2004. Sample size calculations for surveys to substantiate freedom of populations from infectious agents. Biometrics 60, 165–171. Liu, W., Wei, M.T., Tong, Y., Tang, F., Zhang, L., Fang, L., Yang, H., Cao, W.C., 2011. Seroprevalence and genetic characteristics of five subtypes of influenza A viruses in the Chinese pig population: a pooled data analysis. Veterinary Journal 187, 200–206. Nidom, C.A., Takano, R., Yamada, S., Sakai-Tagawa, Y., Daulay, S., Aswadi, D., Suzuki, T., Suzuki, Y., Shinya, K., Iwatsuki-Horimoto, K., Muramoto, Y., Kawaoka, Y., 2010. Influenza A (H5N1) viruses from pigs, Indonesia. Emerging Infectious Diseases 16, 1515–1523. Olsen, C., Brown, I., Easterday, B., Van Reeth, K., 2006. Swine Influenza. Diseases of Swine, 9th ed. Blackwell Publishing, Ames, IA, pp. 469–482. Parchariyanon, S., 2006. Investigation of influenza A virus infection in pigs from 5 reported AIV outbreak provinces in 2004. Journal of Thailand Veterinary Medical Association 57, 16–25. Pourquier, P., 2010. Sensitivity and Specificity of FluAc Idvet ELISA Test on Swine, pers. commun. Santhia, K., Ramy, A., Jayaningsih, P., Samaan, G., Putra, A.A., Dibia, N., Sulaimin, C., Joni, G., Leung, C.Y., Sriyal, J., Peiris, M., Wandra, T., Kandun, N., 2009. Avian influenza A H5N1 infections in Bali Province, Indonesia: a behavioral, virological and seroepidemiological study. Influenza and Other Respiratory Viruses 3, 81–89. Takemae, N., Parchariyanon, S., Damrongwatanapokin, S., Uchida, Y., Ruttanapumma, R., Watanabe, C., Yamaguchi, S., Saito, T., 2008. Genetic diversity of swine influenza viruses isolated from pigs during 2000 to 2005 in Thailand. Influenza and Other Respiratory Viruses 2, 181–189. Trevennec, K., Chevalier, V., Grosbois, V., Garcia, J.M., Thu, H.H., Berthouly–Salazar, C., Peiris, J.S.M., Roger, F., 2011a. Looking for avian influenza in remote areas. A case study in Northern Vietnam. Acta Tropica 120, 160–166. Trevennec, K., Mortier, F., Lyazrhi, F., Huong, H.T., Chevalier, V., Roger, F., 2011b. Swine influenza in Vietnam: preliminary results of epidemiological studies. Influenza and Other Respiratory Viruses 5, 71–73. Vijaykrishna, D., Smith, G.J.D., Pybus, O.G., Zhu, H., Bhatt, S., Poon, L.L.M., Riley, S., Bahl, J., Ma, S.K., Cheung, C.L., 2011. Long-term evolution and transmission dynamics of swine influenza A virus. Nature 473, 519–522. Webster, R.G., Hinshaw, V.S., Bean, W.J., Turner Jr., B., Shortridge, K.F., 1977. Influenza viruses from avian and porcine sources and their possible role in the origin of human pandemic strains. Developments in Biological Standardization 39, 461–468. Yu, H., Hua, R.H., Wei, T.C., Zhou, Y.J., Tian, Z.J., Li, G.X., Liu, T.Q., Tong, G.Z., 2008. Isolation and genetic characterization of avian origin H9N2 influenza viruses from pigs in China. Veterinary Microbiology 131, 82–92.
Contents lists available at SciVerse ScienceDirect
Acta Tropica journal homepage: www.elsevier.com/locate/actatropica
Short communication
Evidence for freedom from swine influenza in a remote area of Northern Vietnam K. Trevennec a,b,∗ , V. Grosbois a , F. Roger a , Thu H. Ho c , C. Berthouly-Salazar d , V. Chevalier a a French Agricultural Research Center for International Development (CIRAD), Animal and Integrated Risk Management Research Unit, Baillarguet Campus (AGIRs), 34398 Montpellier Cedex 5, France b École Nationale Vétérinaire de Toulouse (ENVT), INP, Toulouse, France c National Institute of Veterinary Research (NIVR), Viet Nam d Centre for Excellence of Invasion Biology (CIB), Department of Botany and Zoology, Stellenbosch University, South Africa
a r t i c l e
i n f o
Article history: Received 9 September 2011 Received in revised form 7 November 2011 Accepted 23 November 2011 Available online 2 December 2011 Keywords: Epidemiology Influenza type A Backyard Swine Vietnam Cross-species transmission Avian
a b s t r a c t Swine influenza is responsible for one of the most prevalent disease affecting the swine industry worldwide. Epidemiological surveys rarely focus on remote areas, because traditional farming systems characterized by locally consumed production and low pig densities are considered as having little influence on the emergence, re-emergence, persistence or spread of swine influenza viruses. In addition, routine disease investigations in remote areas are often neglected due to logistic and economical constraints. A bank of swine sera collected in 2005 in the ethnic minorities households of Ha Giang province (Northern Vietnam) located adjacent to the Chinese border was analyzed to estimate the seroprevalence of swine influenza (SI) and to identify potential risk factors for infection. The results suggest that this specific agro-ecological system is free from SI and is not favourable to SI spread either through pig-to-pig transmission, or through poultry-to-pig transmission. © 2011 Elsevier B.V. All rights reserved.
1. Introduction Swine populations are the natural reservoir of H1N1, H3N2 and H1N2 influenza subtypes (Brown, 2000). These 3 subtypes are reported in China (Liu et al., 2011; Vijaykrishna et al., 2011) and in Thailand (Chutinimitkul et al., 2008; Takemae et al., 2008). The avian-origin H9N2 and H5N1 highly pathogenic avian influenza (HPAI) have been occasionally isolated in swine, in China (Cong et al., 2008; Yu et al., 2008), and during outbreaks in Indonesia (Nidom et al., 2010). Since pigs are receptive to both avian and human influenza viruses, they play a crucial role as a mixing vessel in which strains may combine, which raises the risk of emergence of a pandemic influenza virus strain (Brown, 2000; Castrucci et al., 1993; Webster et al., 1977). Routine disease investigations in remote areas are often neglected due to logistic and economical constraints. A FrenchVietnamese research project, called Biodiva (www.biodiva.org.vn), which aimed to assess and promote genetic diversity in domestic and wild animals, collected a sero-bank in the remote area of Northern Vietnam located adjacent to the Chinese-Vietnamese border. In the framework of a global project on animal influenza viruses in
∗ Corresponding author at: French Agricultural Research Center for International Development (CIRAD), Animal and Integrated Risk Management Research Unit, Baillarguet Campus, 34398 Montpellier Cedex 5, France. 0001-706X/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.actatropica.2011.11.012
Vietnam, serological analyses were conducted in this low animal density area on both backyard poultry and pigs, during H5N1 outbreaks in 2005–2006. The seroprevalence of H5N1 was estimated at 7.2% [CI 1.4; 10.5] in chickens (Trevennec et al., 2011a). The aim of the present study is to focus on the swine compartment, in order to estimate the seroprevalence of swine influenza (SI) and to estimate the risk of pig infection with the avian-origin H5N1.
2. Material and methods 2.1. Study design The serological study was performed on sera of pigs bred in Northern Vietnam, which had been collected from April 2005 to August 2006. The study area was the Ha Giang province (22◦ 08–23◦ 19 N; 104◦ 33 –105◦ 33 E), sharing a 274 km border with China (Yunnan province). Ha Giang’s topography is characterized by sharp altitudinal variation. According to the government statistics office, the size of the swine population in the Ha Giang province in 2005 was around 330 000, corresponding to a pig density of 42 head/km2 (General Statistics Office of Vietnam, 2005), which represents a low density in comparison to those found in the highproducing areas of the Red River Delta or of the Mekong Delta (Fig. 1). The sampling design is detailed in the previous article on avian influenza (Trevennec et al., 2011a). Sampled villages are
K. Trevennec et al. / Acta Tropica 122 (2012) 160–163
161
is infected, the prior distribution is built from a maximal possible value and a most likely value. Based on personal communications provided by the test producer, SE and SP most likely values were estimated at 90% and 98%, respectively (Pourquier, 2010). Minimum possible values of SE and SP were chosen as to be weakly informative, thus 50%. The ELISA-A kit evaluation for pig sera is presently in progress and will be published subsequently after completion. Since preliminary results indicate a low SE (unpublished data), we also tested a second option based on a most likely value for SE at 70%. Regarding Pr, we set the most likely value at 5%, which ranges within the estimated prevalence of SI in low pig raising density in China (Liu et al., 2011) or in Northern Vietnam (Trevennec et al., 2011b). The calculator was used for binomial sampling, since the population size, estimated at 330 000 pigs within the province (General Statistics Office of Vietnam, 2005), was large relative to the sample size (Johnson et al., 2004). The posterior distributions of the model parameters are obtained through the Markov Chain MonteCarlo method with 5000 burin-in iterations followed by 20 000 iterations. A population is considered “disease-free” when the proportion of true positive animals is less than a selected threshold value. Various classical scenario (Table 1), corresponding to sporadic cases or endemic situation with a low seroprevalence were chosen. Thus, we focused on the posterior probability that the serological prevalence in the sampled population does not exceed cut-off values of 1%, 2%, 5% and 10%, given the number of positive tests in the sample (Johnson et al., 2004). 3. Results A total of 519 pig sera collected in 158 villages were analyzed (Fig. 2). On average 3.3 pigs from distinct farms (1 pig per farm) were sampled in each village, with a minimum of 1 and a maximum of 7 per village. Only 2 pigs had a positive ELISA-A test. The posterior estimation of seroprevalence was 0.43% [95% CI 0.02; 1.9] for scenario S1–S4 and 0.46% [95% CI 0.03; 1.9] for scenario S5–S8 (Table 1). The probability of the swine population being free from SI antibodies varied from 61%, for a prior SE of 70% and a cut-off value of 1%, to 100% for cut-off values equal to or higher than 5%. Fig. 1. Pig density by province in Vietnam (General Statistics Office of Vietnam, 2005).
mapped in Fig. 2. The pig sera were tested against influenza type A with an enzyme-like immunosorbant assay (ELISA-A) competition test IDVET©. This commercial kit is designed to specifically detect antibodies directed against the NP protein antigen that is highly conserved among all subtypes of influenza type A viruses. 2.2. Freedom of disease analysis for swine In order to account for the uncertainty in test performance and expected prevalence, a Bayesian statistical approach was adopted to demonstrate freedom from disease in swine, using Bayesfreecalc1© available at the UC Davis website (http://www.epi.ucdavis.edu/diagnostictests/). This software program assesses the probability that the prevalence in the sampled population exceeds a given cut-off value given the number of positive reactors, the total number of tested animals and prior knowledge on the performances of the test used and on the expected serological prevalence if the focal population were infected (Johnson et al., 2004). The uncertainty on input parameters is expressed in the form of prior distributions. Prior distributions are built for the sensitivity (SE) and the specificity (SP) from a minimum possible value and a most likely value. For the expected prevalence (Pr) under the hypothesis that the study population
4. Discussion Our findings indicate that the serological prevalence in the Ha Giang province in 2005–2006 was clearly less than 5% and probably less than 1%. When the seroprevalence is under a defined threshold, the disease is considered under control. The detection of sporadic cases will be attributed to isolated and probably independent events (i.e. animal introduction from usual area) without spread. The use of various cut-off values allows testing different scenario and may be a useful tool for health planners. However, this result has to be interpreted cautiously. Firstly, information on the performances of the ELISA-A test on swine is lacking. Further studies are needed to evaluate this test for swine. Secondly, this result contrasts with the estimated seroprevalence of SI published in the literature and the analysis of meta-data on swine influenza infection in Chinese swine populations, where seroprevalence of swine influenza A is estimated at 13.6% in low pig raising area (Liu et al., 2011) and 19.93% in the whole Southwest China, including the Yunnan province (Liu et al., 2011) that shares a border with our study area. The farming system of the study area also presents characteristics that are frequently described as risk factors of SI: farrow-to-finish breeding with mixing animals of different ages, and uncontrolled contacts with other-species (Olsen et al., 2006). Nevertheless, the probable absence of SI antibodies in the Ha Giang pig population may be explained by its agro-ecological
162
K. Trevennec et al. / Acta Tropica 122 (2012) 160–163
Table 1 Probability of freedom from swine influenza according to various scenario of prior information and cut-off values. Scenario
Prior information SE
Cut-off SP
Output P freedom
Pr
Min
Mode
Min
Mode
Max
Mode
High SE
S1 S2 S3 S4
50%
90%
50%
98%
20%
5%
1% 2% 5% 10%
64% 93% 100% 100%
Low SE
S5 S6 S7 S8
50%
70%
50%
98%
20%
5%
1% 2% 5% 10%
61% 93% 100% 100%
system characterized by the absence of well-described risk factors such as high pig and farm densities, high animal turnover or external source of breeding (Olsen et al., 2006). In the Ha Giang province, the pig density is one of the lowest in Vietnam, with only 42.7 pig/km2 (the national pig density in 2005 is 82.7 pig/km2 ) (General Statistics Office of Vietnam, 2005), farms are small, with on average 4–5 pigs and villages are scattered. As a consequence, exposure from neighbouring farms and neighbouring villages is
probably limited and local spread through aerosols is unlikely to occur. Furthermore, pig production is mainly devoted to family consumption so that the local pig trade is poorly developed. The average number of purchased pigs is less than 2 pigs/farmer/year and most of them are provided by family members, farmers of the same village or farmers from the same commune. Thus, we suggest that the lack of pig trade may protect pig farms from the introduction and spread of SI.
Fig. 2. Sampled villages (n = 158) and visited communes in Han Giang province, Northern Vietnam.
K. Trevennec et al. / Acta Tropica 122 (2012) 160–163
H5N1 outbreaks and a seroprevalence of 7.2% [CI 1.4; 10.5] were reported within poultry in the Ha Giang region. It can be concluded that even when exposed to H5N1 virus through possible direct close contact with infected backyard chickens, pigs are not likely to be infected by this avian influenza virus. These conclusions are consistent with results observed in a similar smallholders agro-ecosystem of Indonesia (Santhia et al., 2009). The same conclusion has also been reached in more intensive farming systems in Vietnam and Thailand where no evidence of avian-to-pig transmission of H5N1 have been observed during outbreaks in the poultry populations (Parchariyanon, 2006; Trevennec et al., 2011b). 5. Conclusion The present study in remote areas of Northern Vietnam suggests that this specific agro-ecological system is not favourable to SI spread either through pig-to-pig transmission or through poultryto-pig transmission. Low animal densities and the lack of pig trade may explain this low seroprevalence. However, pig production in Vietnam is developing quickly: facing an increase of pig trade and a potential intensification of the pig production, the epidemiological situation may evolve rapidly. A global description of the animal influenza situation and determinants in various farming systems will help to better understand epidemiological mechanisms in domestic animals. The present study is integrated within the framework of a global study of animal influenza in Vietnam, including specific surveys in the highproducing provinces of the Red River Delta. Acknowledgments This study was a part of the BIODIVA project and was jointly funded by the French Ministry of Foreign Affairs and the National Institute of Animal Husbandry of Vietnam. This study was partially support by the GRIPAVI FSP project (MAEE), and Vent d’Est scholarship (MAEE). We would like to also thank IdVet® lab for providing the ELISA-A kits and Marie Gély from CIRAD for her precious help on mapping skills. References Brown, I.H., 2000. The epidemiology and evolution of influenza viruses in pigs. Veterinary Microbiology 74, 29–46.
163
Castrucci, M.R., Donatelli, I., Sidoli, L., Barigazzi, G., Kawaoka, Y., Webster, R.G., 1993. Genetic reassortment between avian and human influenza A viruses in Italian pigs. Virology 193, 503–506. Chutinimitkul, S., Thippamom, N., Damrongwatanapokin, S., Payungporn, S., Thanawongnuwech, R., Amonsin, A., Boonsuk, P., Sreta, D., Bunpong, N., Tantilertcharoen, R., Chamnanpood, P., Parchariyanon, S., Theamboonlers, A., Poovorawan, Y., 2008. Genetic characterization of H1N1, H1N2 and H3N2 swine influenza virus in Thailand. Archives of Virology 153, 1049–1056. Cong, Y.L., Wang, C.F., Yan, C.M., Peng, J.S., Jiang, Z.L., Liu, J.H., 2008. Swine infection with H9N2 influenza viruses in China in 2004. Virus Genes 36, 461–469. General Statistics Office of Vietnam, 2005. Number of Pigs by Province, ed. GSO, http://www.gso.gov.vn/default en.aspx?tabid= 13/08/2010 469&idmid=3&ItemID=10092. Johnson, W.O., Su, C.-L., Gardner, I.A., Christensen, R., 2004. Sample size calculations for surveys to substantiate freedom of populations from infectious agents. Biometrics 60, 165–171. Liu, W., Wei, M.T., Tong, Y., Tang, F., Zhang, L., Fang, L., Yang, H., Cao, W.C., 2011. Seroprevalence and genetic characteristics of five subtypes of influenza A viruses in the Chinese pig population: a pooled data analysis. Veterinary Journal 187, 200–206. Nidom, C.A., Takano, R., Yamada, S., Sakai-Tagawa, Y., Daulay, S., Aswadi, D., Suzuki, T., Suzuki, Y., Shinya, K., Iwatsuki-Horimoto, K., Muramoto, Y., Kawaoka, Y., 2010. Influenza A (H5N1) viruses from pigs, Indonesia. Emerging Infectious Diseases 16, 1515–1523. Olsen, C., Brown, I., Easterday, B., Van Reeth, K., 2006. Swine Influenza. Diseases of Swine, 9th ed. Blackwell Publishing, Ames, IA, pp. 469–482. Parchariyanon, S., 2006. Investigation of influenza A virus infection in pigs from 5 reported AIV outbreak provinces in 2004. Journal of Thailand Veterinary Medical Association 57, 16–25. Pourquier, P., 2010. Sensitivity and Specificity of FluAc Idvet ELISA Test on Swine, pers. commun. Santhia, K., Ramy, A., Jayaningsih, P., Samaan, G., Putra, A.A., Dibia, N., Sulaimin, C., Joni, G., Leung, C.Y., Sriyal, J., Peiris, M., Wandra, T., Kandun, N., 2009. Avian influenza A H5N1 infections in Bali Province, Indonesia: a behavioral, virological and seroepidemiological study. Influenza and Other Respiratory Viruses 3, 81–89. Takemae, N., Parchariyanon, S., Damrongwatanapokin, S., Uchida, Y., Ruttanapumma, R., Watanabe, C., Yamaguchi, S., Saito, T., 2008. Genetic diversity of swine influenza viruses isolated from pigs during 2000 to 2005 in Thailand. Influenza and Other Respiratory Viruses 2, 181–189. Trevennec, K., Chevalier, V., Grosbois, V., Garcia, J.M., Thu, H.H., Berthouly–Salazar, C., Peiris, J.S.M., Roger, F., 2011a. Looking for avian influenza in remote areas. A case study in Northern Vietnam. Acta Tropica 120, 160–166. Trevennec, K., Mortier, F., Lyazrhi, F., Huong, H.T., Chevalier, V., Roger, F., 2011b. Swine influenza in Vietnam: preliminary results of epidemiological studies. Influenza and Other Respiratory Viruses 5, 71–73. Vijaykrishna, D., Smith, G.J.D., Pybus, O.G., Zhu, H., Bhatt, S., Poon, L.L.M., Riley, S., Bahl, J., Ma, S.K., Cheung, C.L., 2011. Long-term evolution and transmission dynamics of swine influenza A virus. Nature 473, 519–522. Webster, R.G., Hinshaw, V.S., Bean, W.J., Turner Jr., B., Shortridge, K.F., 1977. Influenza viruses from avian and porcine sources and their possible role in the origin of human pandemic strains. Developments in Biological Standardization 39, 461–468. Yu, H., Hua, R.H., Wei, T.C., Zhou, Y.J., Tian, Z.J., Li, G.X., Liu, T.Q., Tong, G.Z., 2008. Isolation and genetic characterization of avian origin H9N2 influenza viruses from pigs in China. Veterinary Microbiology 131, 82–92.