Stochastic simulation of the spread of highly pathogenic avian influenza in Cuba

INTRODUCTION

Infection with high pathogenicity avian influenza viruses (HPAIVs) is a devastating disease of poultry listed by the World Organisation for Animal Health (WOAH) (11. WOAH. Infection with high pathogenicity avian influenza viruses [Internet]. Chapter10.4 in Terrestrial Animal Health Code; 2022 [citado 21 de noviembre de 2022]. Disponible en: https://www.woah.org/en/what-we-do/standards/codes-and-manuals/terrestrial-code-online-access/ ). Low pathogenic avian influenza virus (LPAIV) infections, likewise, are no longer reportable to WOAH since 2021 (22. OIE. Report of the Meeting of the OIE Terrestrial Animal Health Standards Commission [Internet]. 2021. Disponible en: https://www.woah.org/fileadmin/Home/eng/Internationa_Standard_Setting/docs/pdf/A_TAHSC_Feb_2021_Report_Part_A.pdf ) , although some strains of the H5 and H7 subtypes have the ability to mutate to HPAIV.

Waterfowl are often repeatedly exposed to and infected with avian influenza viruses with little or no signs of clinical disease. Since 2002, some HPAIVs have become established in certain wild bird populations, often associated with mortality (33. Sims LD, Domenech J, Benigno C, Kahn S, Kamata A, Lubroth J, et al. Origin and evolution of highly pathogenic H5N1 avian influenza in Asia. Veterinary Record. 2005;157(6):159-64. ,44. Ramey AM, Hill NJ, DeLiberto TJ, Gibbs SE, Camille Hopkins M, Lang AS, et al. Highly pathogenic avian influenza is an emerging disease threat to wild birds in North America. The Journal of Wildlife Management. 2022;86(2):e22171. ). This paradigm shift has been associated with large epidemics, some of them of panzootic magnitude such as those of the H5N1 and H5N8 subtypes from Asian origin (55. Napp S, Majó N, Sánchez‐Gónzalez R, Vergara‐Alert J. Emergence and spread of highly pathogenic avian influenza A (H5N8) in Europe in 2016‐2017. Transboundary and Emerging Diseases. 2018;65(5):1217-26. ).

Recently, a significant increase of HPAI outbreaks worldwide has affected domestic and wild birds, and some terrestrial and aquatic mammals. This reflects a change in the epidemiology and ecology of the virus, which poses a threat to animal health, public health, food security and biodiversity (66. Swayne DE, Sims L, Brown I, Harder T, Stegeman A, Delgado M, et al. Strategic challenges in the global control of High Pathogenicity Avian Influenza [Internet]. Paris: World Organisation for Animal Health; 2023. (90th General Session). Disponible en: https://www.woah.org/app/uploads/2023/05/a-90sg-8.pdf ,77. Shi J, Zeng X, Cui P, Yan C, Chen H. Alarming situation of emerging H5 and H7 avian influenza and effective control strategies. Emerging microbes & infections. 2023;12(1):2155072. ). This situation implies a renewed need to better understand transmission, spread and potential impact of HPAI to improve control and mitigate negative outcomes.

There is growing evidence that HPAIVs are endemic in the avian reservoir (88. Lycett SJ, Pohlmann A, Staubach C, Caliendo V, Woolhouse M, Beer M, et al. Genesis and spread of multiple reassortants during the 2016/2017 H5 avian influenza epidemic in Eurasia. Proceedings of the National Academy of Sciences. 2020;117(34):20814-25. ), adding complexity to their eradication, which allows an increased occurrence of infections. This may imply a higher risk of disease occurrence in countries connected with migratory flyways. Cuba is not only an important wintering area for migratory birds coming from North America, but also a very important staging area for birds moving to other Caribbean islands or birds migrating further south (99. Aguilar S, Manica LT, Acosta M, Castro R, Hernández Z, González A, et al. Spatio-temporal patterns of waterbird assemblages in Cuba’s south coast wetlands: conservation implications. Wetlands. 2019;1-13. ). This, together with the existence of important poultry populations, implies conditions for the occurrence of infections with HPAIVs (1010. de las Nieves Montano D, Irian Percedo M, Vioel Rodríguez S, Fonseca O, Centelles Y, Ley O, Abreu Y, Delgado B, Capdevila Y, Regis Santoro K, Quesada T. Influenza aviar. Oportunidades de mejora del sistema de vigilancia activa basado en riesgo en Cuba. Revista de Salud Animal. 2020 Dec;42(3):1-10.), as recently evidenced by the incursion of HPAI H5N1 in zoo birds (1111. WAHIS. Cuba - Influenza de tipo A de alta patogenicidad (Inf. por los virus de la) (aves que no sean de corral, incluyendo las silvestres) (2017-) [Internet]. Cuba; 2023 [citado 27 de noviembre de 2023]. Report No.: 10. Disponible en: https://wahis.woah.org/#/in-review/4895?fromPage=event-dashboard-url ).

Cuba has implemented an avian influenza surveillance program that includes both passive and active components according to the requirements of the Terrestrial Code (1212. WOAH. Animal health surveillance [Internet]. Chapter 1.4. in Terrestrial Animal Health Code; 2022. Disponible en: https://www.woah.org/en/what-we-do/standards/codes-and-manuals/terrestrial-code-online-access/?id=169&L=1&htmfile=chapitre_surveillance_general.htm ). This system has been continuously improved through the use of procedures such as risk-based surveillance and multi-criteria analysis (1010. de las Nieves Montano D, Irian Percedo M, Vioel Rodríguez S, Fonseca O, Centelles Y, Ley O, Abreu Y, Delgado B, Capdevila Y, Regis Santoro K, Quesada T. Influenza aviar. Oportunidades de mejora del sistema de vigilancia activa basado en riesgo en Cuba. Revista de Salud Animal. 2020 Dec;42(3):1-10.,1313. Ferrer E, Alfonso P, Ippoliti C, Abeledo M, Calistri P, Blanco P, et al. Development of an active risk-based surveillance strategy for avian influenza in Cuba. Preventive Veterinary Medicine. 1 de septiembre de 2014;116(1):161-7. ). However, the magnitude of the epidemic may be dependent on factors associated with transmission dynamics and biosecurity standards (1414. Do O, Ca M, Ab A, Ai A, Ma L, Fo K. Re-emergence of Highly Pathogenic Avian Influenza H5N1 in Nigeria, 2014-2016: Role of Social Network and Value Chain Forces in Interstate Transmission. Virologica Sinica [Internet]. agosto de 2020 [citado 8 de junio de 2021];35(4). Disponible en: https://pubmed.ncbi.nlm.nih.gov/32236814/ ) that have not been assessed. Since these factors vary from country to country, studies in different countries are warranted to support prioritization strategies.

Some studies indicate the benefits of modeling according to their conditions (1515. Flores U, Rivera A, Sanchez J. Uso del Modelo Norte Americano de Dispersión de Enfermedades - NAADSM, para la Simulación de Brotes de Fiebre Aftosa en el Perú, en Base al Movimiento de Bovinos - Perulactea [Internet]. 2014 [citado 24 de abril de 2018]. Disponible en: http://www.perulactea.com/2014/05/01/uso-del-modelo-norte-americano-de-dispersion-de-enfermedades-naadsm-para-la-simulacion-de-brotes-de-fiebre-aftosa-en-el-peru-en-base-al-movimiento-de-bovinos/ -1818. Stevenson MA, Sanson RL, Stern MW, O’Leary BD, Sujau M, Moles-Benfell N, et al. InterSpread Plus: a spatial and stochastic simulation model of disease in animal populations. Preventive veterinary medicine. 2013;109(1-2):10-24. ). Lewis et al. (1515. Flores U, Rivera A, Sanchez J. Uso del Modelo Norte Americano de Dispersión de Enfermedades - NAADSM, para la Simulación de Brotes de Fiebre Aftosa en el Perú, en Base al Movimiento de Bovinos - Perulactea [Internet]. 2014 [citado 24 de abril de 2018]. Disponible en: http://www.perulactea.com/2014/05/01/uso-del-modelo-norte-americano-de-dispersion-de-enfermedades-naadsm-para-la-simulacion-de-brotes-de-fiebre-aftosa-en-el-peru-en-base-al-movimiento-de-bovinos/ ), based on HPAI simulation; recommend the most effective policy for HPAI control in their region of interest. Given Cuba's AI-free status, modeling could be useful for the improvement of the established disease surveillance system and decisions on resource allocation for disease prevention and poultry industry expansion. This study aimed to estimate the magnitude of the consequences of HPAIV incursion in Cuba and the effectiveness of important control measures.

MATERIALS AND METHODS

RESULTS

It was found, in the best-case scenario by random selection, that the outbreak did not exceed the initially infected farms. Detection occurred within four days of infection and the first destruction within six days, with outbreak duration of 12 days. However, 87 826 dead animals were estimated, of which 37 888 were of the laying type.

In the most likely scenario, outbreak duration was 14 days, with an increase of one infected farm over the initial ones and an estimated loss of 199 913 poultry. Epidemic curves showed that the highest number of infected animals was within 95^th percentile.

The worst scenario involved 15 more farms than the best-case scenario, with persistent of farms in a latent state (Fig. 2), implying over a million of infected animals. On most farms, outbreak duration ranged from 20 to 26 days, almost two times longer than in the best scenario.

Statistical analyses of randomized scenarios, MANOVA, indicated that biosecurity and movement restriction were significant at the 99 % confidence level with respect to the dependent variables. (Annex 1). The interaction of biosecurity, movement restriction and detection speed had a 90 % confidence level with respect to the spread of HPAI (Annex 2). The least squares test for residuals revealed that biosecurity and movement restriction were significant at 95 % respectively. The model fitted the observed values with the predicted values, indicating a p-value of 0,0008804 (Annex 3). The fitted multiple linear regression model showed that low biosecurity and less movement restriction were the factors increasing infection with 99 % and 95 % confidence, respectively (Annex 4). Scenarios with low biosecurity had 57 % of the total positive cases.

For the high population size condition, the worst-case scenario, compared to the most favorable ones, implied a significant growth of the affected farms, highlighting the appearance of the latent state (Fig. 3).

Cases with a high population were tested and the detection rate for the worst case scenario was too slow at seven days (Fig. 4). It is evident that for rapid detection, infected and destroyed farms were not affected for the different biosafety levels. For the most likely scenario, the best scenario in terms of the number of depopulated and infected farms did not show significant differences.

Statistical analyses in high population size scenarios showed that biosecurity and movement restrictions were significant at 99 % and 90 % confidence level, respectively (Annex 5). Interactions between biosecurity and movement restrictions were significant at a 95 % confidence level (Annex 6). The least squares test showed that biosecurity and minor movement restriction had a 95 % confidence level (Annex 7). This model offered a goodness of fit of 0.7051(R2) according to the least squares test with a p-value = 0,00624. For the multiple binary regression, it was evident that movement restrictions was directly related with the increase of infected farms, which was significant at a 95 % confidence level (Annex 8).

There was a higher probability of dissemination on farm contiguity, making them priority regions for resilience development.

In the contiguous scenarios, the difference in losses between the worst-case scenario relative to the best-case scenario and the most likely scenario was evident (Fig. 5). Even for the worst scenario, there were so many infected animals that farms persisted in a dormant state. With respect to animal losses, the best scenario had 198,178 infected animals and 218,076 slaughtered animals. Similar results were obtained in the most likely scenario; however, the worst scenario had almost four times as many infected animals as the rest.

In scenarios close to migratory waterfowl settlements, the most likely had 2.65 times more slaughtered animals than the best scenario and it had almost two times less loses than the worst case. The outbreak for the worst scenario occurred in 27 days. In contrast, the other scenarios ranged from 10 to 12 days.

Comparison of the three main risk dimensions (high population size, farm contiguity and proximity to wetlands), for the worst-case scenario, showed the greatest impact for farm contiguity (Table 4).

Statistical analysis of farm contiguity revealed that biosecurity was significant with respect to infected farms at the 95 % confidence level and movement restrictions at the 90 % confidence level (Annex 9). Interactions between biosecurity and movement restrictions were significant for a 90 % confidence level with respect to infected farms (Annex 10). The least squares test showed that biosecurity was significant at a 95 % confidence level (Annex 11). This model showed a goodness of fit of 0,6389 (R2) with a p-value = 0.01868. Regression model indicated that low biosecurity was significant at the 95 % confidence level with respect to the percentage increase in infection. These results were consistent with those obtained in previous scenarios showing the implication for virus spread caused by low biosecurity in the farms. (Annex 12).

DISCUSSION

The current study provided the first insight into potential cases for of HPAIVs transmission in Cuba through NAADSM. It showed referent values for containment measures for the successful management of eventual emergencies due to the incursion of these viruses in commercial poultry farms. The main objective of contingency plans for exotic infectious diseases is to restore a disease-free status as quickly as possible (2323. Dalla Pozza M, Ceolin C, Marangon S. Emergency response following suspicion of an avian influenza outbreak. Zoonoses and public health. 2008;55(1):50-3. ). This is the case of Cuba for several transboundary diseases, including HPAI; therefore modeling can be an important contribution in this regard (2424. Clemmons EA, Alfson KJ, Dutton III JW. Transboundary animal diseases, an overview of 17 diseases with potential for global spread and serious consequences. Animals. 2021;11(7):2039. https://doi.org/10.3390/ani11072039 ).

The finding that biosecurity was one of the most influential factors in the impact of HPAI highlighted the importance of prevention. Biosecurity measures have been shown to be a major obstacle to the spread of HPAI on poultry farms (2525. Authority (EFSA) EFS. Animal health and welfare aspects of avian influenza and the risk of its introduction into the EU poultry holdings - Scientific opinion of the Panel on Animal Health and Welfare. EFSA Journal. 2008;6(6):715.,2626. So H, Bae Y, Mo I. Risk assessment of highly pathogenic avian influenza by comparison of biosecurity level in domestic poultry farms. Korean Journal of Poultry Science. 2019;46(4):313-25. ). Indeed, prevention and preparedness are "peacetime" phases of disaster management and risk reduction (2222. Health (OIE) WO for A. Guidelines on Disaster Management and Risk Reduction in Relation to Animal Health and Welfare and Veterinary Public Health. OIE Paris, France; 2016. ,2727. Bonbon E, Myers L. Good emergency management practices: a review of the core principles, with a focus on preparedness. Revue Scientifique et Technique (International Office of Epizootics). 2020;39(2):533-41. ). In particular, when facing infectious diseases with zoonotic/pandemic potential such as avian influenza, investing in prevention is not only cost-effective, but can also preserve human infections with their potential negative consequences (2828. Bernstein AS, Ando AW, Loch-Temzelides T, Vale MM, Li BV, Li H, et al. The costs and benefits of primary prevention of zoonotic pandemics. Science Advances. 2022;8(5):eabl4183. ,2929. Gilbert W, Adamson D, Donachie D, Hamilton K, Rushton J, 2023. A Cost-Benefit Analysis of Preparing National Veterinary Services for Transboundary Animal Disease Emergencies. Transboundary and Emerging Diseases, 2023. 1765243 ). Moreover, as the main biosecurity measures are not disease-specific, the overall risk of infection with other infectious agents is reduced.

The finding that laying included the highest losses could be related to its higher contact rates compared to other types of production and to the movement of flocks from farm to farm from hatching to pullets. This movement rate, coupled with regular egg collection, could increase the likelihood of infection from other farms. In addition, the proportion of laying hens is the most represented commercial type in the whole country (ONEI 2021). This fact allows to insist on the importance of adequate biosecurity and enhanced surveillance. A recent study identified that the most significant farm-level risk factor for HPAI in table-egg commercial farms is being located within an existing control zone (3030. Green AL, Branan M, Fields VL, Patyk K, Kolar SK, Beam A, et al. Investigation of risk factors for introduction of highly pathogenic avian influenza H5N1 virus onto table egg farms in the United States, 2022: a case-control study. Frontiers in Veterinary Science. 2023;10, 1229008: 1-14. https://doi.org/10.3389/fvets.2023.1229008.), while the supply chain is also identified as an important risk factor (3131. Moyen N, Hoque MA, Mahmud R, Hasan M, Sarkar S, Biswas PK, et al. Avian influenza transmission risk along live poultry trading networks in Bangladesh. Scientific reports. 2021;11(1):19962. ).

Although considering that all the breeding sites in Cuba may have similar probability of virus introduction (randomness), it would not be the most viable cost-effective surveillance (3232. Alfonso P, Miranda EF, Al E. Perfeccionamiento del sistema de vigilancia activa de la influenza aviar basado en el análisis de riesgo de su introducción al país. Anales de la Academia de Ciencias de Cuba [Internet]. 23 de diciembre de 2016 [citado 21 de mayo de 2018];6(3). Disponible en: http://www.revistaccuba.cu/index.php/acc/article/view/495 ). These scenarios helped to demonstrate that when facing the constant threat of HPAIV, the importance of reaching a high level of biosecurity to control or mitigate the risk of spread of these viruses should be emphasized. Similarly, Ssematimba et al. (2222. Health (OIE) WO for A. Guidelines on Disaster Management and Risk Reduction in Relation to Animal Health and Welfare and Veterinary Public Health. OIE Paris, France; 2016. ) report that poor biosecurity measures were the main reason for the continued spread of HPAI H7N7 in the Dutch poultry population, at a time when contingency plans had been put in place as a means to end the outbreak.

The lack of differences among affected farms according to population size highlights the importance of the interaction among risk factors, including biosecurity. In fact, Sharkey et al. (2323. Dalla Pozza M, Ceolin C, Marangon S. Emergency response following suspicion of an avian influenza outbreak. Zoonoses and public health. 2008;55(1):50-3. ) demonstrated the risk of infection in areas of high density produced on small and large scale, as well as Stevenson et al. (1717. Pepin KM, Spackman E, Brown JD, Pabilonia KL, Garber LP, Weaver JT, et al. Using quantitative disease dynamics as a tool for guiding response to avian influenza in poultry in the United States of America. Preventive veterinary medicine. 2014;113(4):376-97. ) and Green et al. (3535. Green AL, Branan M, Fields VL, Patyk K, Kolar SK, Beam A, et al. Investigation of risk factors for introduction of highly pathogenic avian influenza H5N1 virus onto table egg farms in the United States, 2022: a case-control study. Frontiers in Veterinary Science. 2023;10. https://doi.org/10.3389/fvets.2023.1229008 ). However, the cases of low containment effectiveness demonstrated by simulation can be explained by the persistence of farms in a dormant state.

The present study revealed the need to detect infections less than three days after they have occurred to reach adequate mitigation. Currently, high-throughput diagnostic techniques providing results in less than 24 hours are widely available, even in Cuba (3636. WOAH. Terrestrial Manual. Chapter 3.3.4. Avian Influenza (Infection with Avian Influenza Viruses [Internet]. 2021. Disponible en: https://www.woah.org/fileadmin/Home/eng/Health_standards/tahm/3.03.04_AI.pdf , 3737. Alfonso P, González CL, Becerra HC. Animal influenza in the Caribbean: a systematic review of research and surveillance. Revista de Salud Animal. 2020;42(3): 1- 15.). Nevertheless, a study based on outbreak data (3838. Hobbelen PH, Elbers AR, Werkman M, Koch G, Velkers FC, Stegeman A, et al. Estimating the introduction time of highly pathogenic avian influenza into poultry flocks. Scientific Reports. 2020;10(1):12388. ) estimates that the most likely time of HPAI introduction ranges from ≤14.8 to ≤ 18.8 days before disease detection, depending on the poultry species. In fact, incubation period for HPAI infection is set at 14 days according to the Code (1).

To ensure a timely response, stakeholders' awareness is crucial to promote self-reporting of unusual mortality or the presence of signs or lesions related to HPAI infection. Detection of the pathogen may depend on human capacity and reaction to the manifestation of HPAI clinical signs. Maintaining and assessing production and mortality records, and reporting when abnormalities are observed according to syndromic surveillance procedures, also provides early warning and are mandatory in some situations (3939. Hernández-Jover M, Phiri BJ, Stringer L, Martínez Avilés M. Developments in animal health surveillance. Front. Vet. Sci. 2021;7:637364. https://doi.org/10.3389/fvets.2020.637364 ,4040. Gonzales JL, Hobbelen PH, van der Spek AN, Vries EP, Elbers AR. Highly pathogenic avian influenza A H5 virus outbreaks in broiler farms in the Netherlands. Clinical signs, transmission and identification of reporting thresholds. bioRxiv. 2023;2023.01. 05.522008. ). On the other hand, knowledge and training are among the main components to control HPAIVs (4141. Sims LD, Swayne DE. Avian influenza control strategies. Chapter Animal influenza. Animal influenza. 2016;363-77. ).

Detection window is taken into account for contact tracing which, together with other important control measures such as movement restriction and stamping out, aims to prevent or reduce the size of an epidemic (4242. Klinkenberg D, Fraser C, Heesterbeek H. The effectiveness of contact tracing in emerging epidemics. PloS one. 2006;1(1):e12. ).

The model suggested that it is possible to contain the spread of HPAIVs for a detection within three days of infection, a first depopulation within six days and an outbreak duration of 12 days. Prompt identification of an infectious disease is crucial, but it is useful only if facilities can respond quickly and appropriately to the emergency situation in such a way that all necessary measures are implemented to contain and then progressively eliminate the infection (2323. Dalla Pozza M, Ceolin C, Marangon S. Emergency response following suspicion of an avian influenza outbreak. Zoonoses and public health. 2008;55(1):50-3. ).

Avian influenza virus control strategies in poultry vary depending on whether the target is prevention, management, or eradication (4141. Sims LD, Swayne DE. Avian influenza control strategies. Chapter Animal influenza. Animal influenza. 2016;363-77. ) . Even within the same target, the severity of control measures may vary from country to country depending on the scope of contingency plans and the expected consequences of the disease. In fact, despite criticism, preventive stamping out of uninfected at-risk farms has been considered for several high-consequence diseases, including HPAI (1616. Lewis N, Dorjee S, Dubé C, VanLeeuwen J, Sanchez J. Assessment of Effectiveness of Control Strategies Against Simulated Outbreaks of Highly Pathogenic Avian Influenza in Ontario, Canada. Transbound Emerg Dis. junio de 2015;64(3):938-50. , 4343. Bergevoet RHM, van Asseldonk M. Análisis económico de la erradicación de brotes de fiebre aftosa usando estrategias alternativas de control. Archivos de medicina veterinaria. 2014;46(3):381-8. -4646. Bradhurst R, Garner G, Roche S, Iglesias R, Kung N, Robinson B, et al. 2021. Modelling the spread and control of African swine fever in domestic and feral pigs. Technical Report. Australian Government, Department of Agriculture, Water and the Environment. Disponible en: https://era.daf.qld.gov.au/id/eprint/8363/ ).

Such decisions may depend on several factors including the magnitude and timing of the epidemic, as well as its implications for poultry and poultry product exports. Modeling is a useful resource for early decision making, particularly when facing of the potential incursion of pathogens causing high-consequence diseases. Poultry production in Cuba is predominantly based on laying, and constitutes one of the main sources of native animal proteins supporting an average per capita consumption of about 210 eggs/person/year (4747. Oficina Nacional de Estadística e Información, Sitio en Actualización [Internet]. 2020 [citado 13 de septiembre de 2022]. ONEI. Disponible en: http://www.onei.gob.cu/ ). Hence it is an important component of food security (4848. Bradhurst R, Garner G, Hóvári M, de la Puente M, Mintiens K, Yadav S, et al. Development of a transboundary model of livestock disease in Europe. Transboundary and Emerging Diseases. 2022;69(4):1963-82. ).

The main limitation of this stochastic model was the need to establish, by different hypotheses, variables that could not be measured. Although the expert´s survey was useful to collect primary information, it could incur in the omission of significant data. In addition, in Cuba there have never been outbreaks of HPAIV infection in poultry. Therefore, there is no case available to compare the results and validate this study, considering that this is the first time that the presence of HPAIV has been simulated in Cuba.

Previous Cuban studies on this subject have strengthened surveillance system by stratifying the areas of greatest risk of disease occurrence (1010. de las Nieves Montano D, Irian Percedo M, Vioel Rodríguez S, Fonseca O, Centelles Y, Ley O, Abreu Y, Delgado B, Capdevila Y, Regis Santoro K, Quesada T. Influenza aviar. Oportunidades de mejora del sistema de vigilancia activa basado en riesgo en Cuba. Revista de Salud Animal. 2020 Dec;42(3):1-10.,2121. Mongkolsawat C, Kamchai T. GIS Modeling for Avian Influenza Risk Areas | IJG:International Journal of Geoinformatics. International Journal of Geoinformatics [Internet]. 2009 [citado 7 de enero de 2024];5(1); ISSN. 1686-6576. Disponible en: http://creativecity.gscc.osaka-cu.ac.jp/IJG/article/view/483 ,2424. Clemmons EA, Alfson KJ, Dutton III JW. Transboundary animal diseases, an overview of 17 diseases with potential for global spread and serious consequences. Animals. 2021;11(7):2039. https://doi.org/10.3390/ani11072039 ) . These areas of interest, together with the results obtained in the present study, can provide early warning and resilience in case of outbreaks caused by HPAIVs.

The quick dissemination of infection in farm contiguity is consistent with the potential for aerogenic transmission attributed to the virus associated with the surprising magnitude of the HPAI epidemic in the United States during 2015 (5050. Zhao Y, Richardson B, Takle E, Chai L, Schmitt D, Xin H. Airborne transmission may have played a role in the spread of 2015 highly pathogenic avian influenza outbreaks in the United States. Scientific reports. 2019;9(1):11755. ), which were taken into account. Therefore, areas with high farm contiguity should be prioritized for investment of resources in biosecurity and early warning of virus introduction. Decisions of this nature would reduce the risk of introducing the agent into poultry populations and even, in the event of its introduction, detect it quickly and initiate a timely response to contain the spread.

According to the results obtained in this study, special attention should be paid to improve biosecurity and to promptly impose movement restrictions after the detection of suspected or confirmed cases. Preventive stamping out was not taken into account in this model, although it has been used by several countries (4545. Salvador R, Tanquilut N, Macmac R, Na Lampang K, Chaisowwong W, Pfeiffer D, et al. Evaluation of strategies using simulation model to control a potential outbreak of highly pathogenic avian influenza among poultry farms in Central Luzon, Philippines. PloS one. 2020;15(9):e0238815. ,5151. Andronico A, Courcoul A, Bronner A, Scoizec A, Lebouquin-Leneveu S, Guinat C, et al. Highly pathogenic avian influenza H5N8 in south-west France 2016-2017: A modeling study of control strategies. Epidemics. 2019;28:100340. ,5252. Seymour RG, Kypraios T, O’Neill PD, Hagenaars TJ. A Bayesian nonparametric analysis of the 2003 outbreak of highly pathogenic avian influenza in the Netherlands. Journal of the Royal Statistical Society Series C: Applied Statistics. 2021;70(5):1323-43. ). However, there are large differences among countries in terms of the intensity of production systems and market aspects suggesting not to directly transfer this measure. In fact, a similar model in Canada showed no improvement in HPAI containment by this measure (1616. Lewis N, Dorjee S, Dubé C, VanLeeuwen J, Sanchez J. Assessment of Effectiveness of Control Strategies Against Simulated Outbreaks of Highly Pathogenic Avian Influenza in Ontario, Canada. Transbound Emerg Dis. junio de 2015;64(3):938-50. ).

Strategic and systematic vaccination against HPAI was not taken into account in this model, even though it is a topic that is currently receiving much attention as an additional protective measure (5353. Alqazlan N, Astill J, Raj S, Sharif S. Strategies for enhancing immunity against avian influenza virus in chickens: A review. Avian Pathology. 2022;51(3):211-35. ). However, this would require further studies, considering that vaccination is not a substitute for lack of biosecurity.

CONCLUSIONS

The study highlights the importance of strengthening preventive measures and effective control policies to limit or at least mitigate the spread of HPAIV under commercial poultry farming conditions in Cuba. Biosecurity proved to be the most influential factor requiring special attention in the prevention phase of emergency management and disaster risk reduction. This model suggests that it is possible to contain the spread of HPAI virus if detection is reached within three days of infection and depopulation of animals is completed within six days. This requires a high level of producer awareness promoting early warning and effective timely response.