Use of epidemiologic models in the control of highly pathogenic avian influenza

A systematic review of mechanistic models used to study avian influenza virus transmission and control

The global spread of avian influenza A viruses in domestic birds is causing increasing socioeconomic devastation. Various mechanistic models have been developed to better understand avian influenza transmission and evaluate the effectiveness of control measures in mitigating the socioeconomic losses caused by these viruses. However, the results of models of avian influenza transmission and control have not yet been subject to a comprehensive review. Such a review could help inform policy makers and guide future modeling work. To help fill this gap, we conducted a systematic review of the mechanistic models that have been applied to field outbreaks. Our three objectives were to: (1) describe the type of models and their epidemiological context, (2) list estimates of commonly used parameters of low pathogenicity and highly pathogenic avian influenza transmission, and (3) review the characteristics of avian influenza transmission and the efficacy of control strategies according to the mechanistic models. We reviewed a total of 46 articles. Of these, 26 articles estimated parameters by fitting the model to data, one evaluated the effectiveness of control strategies, and 19 did both. Values of the between-individual reproduction number ranged widely: from 2.18 to 86 for highly pathogenic avian influenza viruses, and from 4.7 to 45.9 for low pathogenicity avian influenza viruses, depending on epidemiological settings, virus subtypes and host species. Other parameters, such as the durations of the latent and infectious periods, were often taken from the literature, limiting the models' potential insights. Concerning control strategies, many models evaluated culling (n = 15), while vaccination received less attention (n = 6). According to the articles reviewed, optimal control strategies varied between virus subtypes and local conditions, and depended on the overall objective of the intervention. For instance, vaccination was optimal when the objective was to limit the overall number of culled flocks. In contrast, pre-emptive culling was preferred for reducing the size and duration of an epidemic. Early implementation consistently improved the overall efficacy of interventions, highlighting the need for effective surveillance and epidemic preparedness.

A review of estimated transmission parameters for the spread of avian influenza viruses

Avian influenza poses an increasing problem in Europe and around the world. Simulation models are a useful tool to predict the spatiotemporal risk of avian influenza spread and evaluate appropriate control actions. To develop realistic simulation models, valid transmission parameters are critical. Here, we reviewed published estimates of the basic reproduction number (R₀ ), the latent period and the infectious period by virus type, pathogenicity, species, study type and poultry flock unit. We found a large variation in the parameter estimates, with highest R₀ estimates for H5N1 and H7N3 compared with other types; for low pathogenic avian influenza compared with high pathogenic avian influenza types; for ducks compared with other species; for estimates from field studies compared with experimental studies; and for within-flock estimates compared with between-flock estimates. Simulation models should reflect this observed variation so as to produce more reliable outputs and support decision-making. How to incorporate this information into simulation models remains a challenge.

The African swine fever modelling challenge: Model comparison and lessons learnt

Robust epidemiological knowledge and predictive modelling tools are needed to address challenging objectives, such as: understanding epidemic drivers; forecasting epidemics; and prioritising control measures. Often, multiple modelling approaches can be used during an epidemic to support effective decision making in a timely manner. Modelling challenges contribute to understanding the pros and cons of different approaches and to fostering technical dialogue between modellers. In this paper, we present the results of the first modelling challenge in animal health - the ASF Challenge - which focused on a synthetic epidemic of African swine fever (ASF) on an island. The modelling approaches proposed by five independent international teams were compared. We assessed their ability to predict temporal and spatial epidemic expansion at the interface between domestic pigs and wild boar, and to prioritise a limited number of alternative interventions. We also compared their qualitative and quantitative spatio-temporal predictions over the first two one-month projection phases of the challenge. Top-performing models in predicting the ASF epidemic differed according to the challenge phase, host species, and in predicting spatial or temporal dynamics. Ensemble models built using all team-predictions outperformed any individual model in at least one phase. The ASF Challenge demonstrated that accounting for the interface between livestock and wildlife is key to increasing our effectiveness in controlling emerging animal diseases, and contributed to improving the readiness of the scientific community to face future ASF epidemics. Finally, we discuss the lessons learnt from model comparison to guide decision making.

Movement Restriction and Increased Surveillance as Efficient Measures to Control the Spread of Highly Pathogenic Avian Influenza in Backyard Productive Systems in Central Chile

During the last 5 years there has been an alarming number of reports of highly pathogenic avian influenza worldwide. However, little is known about the status of this disease in South America. Chile has been the only country in South America where an HPAI outbreak was reported. This outbreak occurred in 2002 and was due to an H7N3 HPAI, where the most plausible hypothesis that explained the entrance of the disease to the country, had relation to migratory wild birds. Commercial poultry farms in Chile are highly integrated and have high biosecurity standards. Nevertheless, poultry backyard production systems lack biosecurity measures and are widely distributed. Since 2002 outbreak, avian influenza viruses have been identified in wild birds and different animal species kept in backyard productive systems (BPS) in Chile. The aim of this study was to simulate the possible natural history of HPAI after its introduction to BPS in central Chile and to simulate different intervention strategies. To do so, the North American Animal Disease Spread Model version 3.3 was used. The results showed that a median of 15,930 BPS would be affected if HPAI spread among BPS in central Chile, representing 97.8% of the current amount of BPS existing in study zone. Movement restrictions, pre-emptive destruction, passive surveillance, tracing of infected premises and combinations of the three, where the intervention strategies tested in the simulation model. From all the interventions simulated, movement restrictions together with increasing surveillance (through increasing passive surveillance and good tracing of infected premises) had the biggest effect, reducing the median number of infected BPS in 90.8%. However, more studies are needed to more accurately estimate local contact rates. These results can guide the official veterinary services to consider potential mechanisms to control or prevent an HPAI emergency situation.

Exploring contacts facilitating transmission of influenza A(H5N1) virus between poultry farms in West Java, Indonesia: A major role for backyard farms?

Highly pathogenic avian influenza virus (HPAIV) H5N1 has been reported in Asia, including Indonesia since 2003. Although several risk factors related to the HPAIV outbreaks in poultry in Indonesia have been identified, little is known of the contact structure of farms of different poultry production types (backyard chickens, broilers, layers, and ducks). This study aims to quantify the contact rates associated with the movement of people, and movements of live birds and products and equipment that affect the risk of HPAIV H5N1 transmission between poultry farms in Indonesia. On 124 poultry farms in 6 districts in West Java, logbooks were distributed to record the movements of farmers/staff and visitors and their poultry contacts. Most movements in backyard chicken, commercial native chicken, broiler and duck farms were visits to and from other poultry farms, whilst in layer farms visits to and from poultry companies, visits to egg collection houses and visit from other poultry farms were most frequent. Over 75% of persons visiting backyard chicken and duck farms had previously visited other poultry farms on the same day. Visitors of backyard chicken farms had the highest average contact rate, either direct contact with poultry on other farms before the visits (1.35 contact/day) or contact during their visits in the farms (10.03 contact/day). These results suggest that backyard chicken farms are most at risk for transmission of HPAIV compared to farms of the other poultry production types. Since visits of farm-to-farm were high, backyard farms could also a potential source for HPAIV transmission to commercial poultry farms.

Estimating the between-farm transmission rates for highly pathogenic avian influenza subtype H5N1 epidemics in Bangladesh between 2007 and 2013

Highly Pathogenic Avian Influenza (HPAI) is classified by the World Organization for Animal Health as one of the notifiable diseases. Its occurrence is associated with severe socio-economic impacts and is also zoonotic. Bangladesh HPAI epidemic data for the period between 2007 and 2013 were obtained and split into epidemic waves based on the time lag between outbreaks. By assuming the number of newly infected farms to be binomially distributed, we fit a Generalized Linear Model to the data to estimate between-farm transmission rates (β). These parameters are then used together with the calculated infectious periods to estimate the respective basic reproduction numbers (R₀ ). The change in β and R₀ with time during the course of each epidemic wave was explored. Finally, sensitivity analyses of the effects of reducing the delay in detecting infection on a farm as well as extended infectiousness of a farm beyond the day of culling were assessed. The point estimates obtained for β ranged from 0.08 (95% CI: 0.06-0.10) to 0.11 (95% CI: 0.08-0.20) per infectious farm per day while R₀ ranged from 0.85 (95% CI: 0.77-1.02) to 0.96 (95% CI: 0.72-1.20). Sensitivity analyses reveal that the estimates are quite robust to changes in the assumptions about the day in reporting infection and extended infectiousness. In the analysis allowing for time-varying transmission parameters, the rising and declining phases observed in the epidemic data were synchronized with the moments when R₀ was greater and less than one, respectively. From an epidemiological perspective, the consistency of these estimates and their magnitude (R₀  ≈ 1) indicate that the effectiveness of the deployed control measures was largely invariant between epidemic waves and the trend of the time-varying R₀ supports the hypothesis of sustained farm-to-farm transmission that is possibly initiated by a few unique introductions.

Assessment of China's H5N1 routine vaccination strategy

In this study, a simulation model was used to assess the epidemiological and economic impacts of the routine vaccination strategy for H5N1 in China at the national level. The results of the simulation show that a routine vaccination strategy at the national level could have a substantial impact on decreasing H5N1 outbreaks; it could therefore markedly reduce the severity and duration of an H5N1 epidemic at the national level in China. Under a routine vaccination programme, when a flock is infected, the number of depopulated birds could be reduced by approximately 91%, the outbreak duration could be reduced by one-fourth, and the probability of an H5N1 outbreak could be reduced to 51.5%. Although the use of vaccination has obvious benefits, if indirect costs were not considered, the mean direct cost of simulated disease control without vaccination was only 1.36% of the cost of the routine vaccination strategy, and this former approach would have saved 57 billion yuan for all poultry in China from 2004 to 2012. Traditional H5N1 control strategies with culling programmes at the national level represented a better choice for China.

Controlling highly pathogenic avian influenza outbreaks: An epidemiological and economic model analysis

Outbreaks of highly pathogenic avian influenza (HPAI) can cause large losses for the poultry sector and for animal disease controlling authorities, as well as risks for animal and human welfare. In the current simulation approach epidemiological and economic models are combined to compare different strategies to control highly pathogenic avian influenza in Dutch poultry flocks. Evaluated control strategies are the minimum EU strategy (i.e., culling of infected flocks, transport regulations, tracing and screening of contact flocks, establishment of protection and surveillance zones), and additional control strategies comprising pre-emptive culling of all susceptible poultry flocks in an area around infected flocks (1 km, 3 km and 10 km) and emergency vaccination of all flocks except broilers around infected flocks (3 km). Simulation results indicate that the EU strategy is not sufficient to eradicate an epidemic in high density poultry areas. From an epidemiological point of view, this strategy is the least effective, while pre-emptive culling in 10 km radius is the most effective of the studied strategies. But these two strategies incur the highest costs due to long duration (EU strategy) and large-scale culling (pre-emptive culling in 10 km radius). Other analysed pre-emptive culling strategies (i.e., in 1 km and 3 km radius) are more effective than the analysed emergency vaccination strategy (in 3 km radius) in terms of duration and size of the epidemics, despite the assumed optimistic vaccination capacity of 20 farms per day. However, the total costs of these strategies differ only marginally. Extending the capacity for culling substantially reduces the duration, size and costs of the epidemic. This study demonstrates the strength of combining epidemiological and economic model analysis to gain insight in a range of consequences and thus to serve as a decision support tool in the control of HPAI epidemics.

Towards universal influenza vaccines?

Vaccination is the most cost-effective way to reduce the considerable disease burden of seasonal influenza. Although seasonal influenza vaccines are effective, their performance in the elderly and immunocompromised individuals would benefit from improvement. Major problems related to the development and production of pandemic influenza vaccines are response time and production capacity as well as vaccine efficacy and safety. Several improvements can be envisaged. Vaccine production technologies based on embryonated chicken eggs may be replaced by cell culture techniques. Reverse genetics techniques can speed up the generation of seed viruses and new mathematical modelling methods improve vaccine strain selection. Better understanding of the correlates of immune-mediated protection may lead to new vaccine targets besides the viral haemagglutinin, like the neuraminidase and M2 proteins. In addition, the role of cell-mediated immunity could be better exploited. New adjuvants have recently been shown to increase the breadth and the duration of influenza vaccine-induced protection. Other studies have shown that influenza vaccines based on different viral vector systems may also induce broad protection. It is to be expected that these developments may lead to more universal influenza vaccines that elicit broader and longer protection, and can be produced more efficiently.

A single vaccination of commercial broilers does not reduce transmission of H5N1 highly pathogenic avian influenza

Vaccination of chickens has become routine practice in Asian countries in which H5N1 highly pathogenic avian influenza (HPAI) is endemically present. This mainly applies to layer and breeder flocks, but broilers are usually left unvaccinated. Here we investigate whether vaccination is able to reduce HPAI H5N1 virus transmission among broiler chickens. Four sets of experiments were carried out, each consisting of 22 replicate trials containing a pair of birds. Experiments 1-3 were carried out with four-week-old birds that were unvaccinated, and vaccinated at day 1 or at day 10 of age. Experiment 4 was carried out with unvaccinated day-old broiler chicks. One chicken in each trial was inoculated with H5N1 HPAI virus. One chicken in each trial was inoculated with virus. The course of the infection chain was monitored by serological analysis, and by virus isolation performed on tracheal and cloacal swabs. The analyses were based on a stochastic SEIR model using a Bayesian inferential framework. When inoculation was carried out at the 28^th day of life, transmission was efficient in unvaccinated birds, and in birds vaccinated at first or tenth day of life. In these experiments estimates of the latent period (~1.0 day), infectious period (~3.3 days), and transmission rate parameter (~1.4 per day) were similar, as were estimates of the reproduction number (~4) and generation interval (~1.4 day). Transmission was significantly less efficient in unvaccinated chickens when inoculation was carried out on the first day of life. These results show that vaccination of broiler chickens does not reduce transmission, and suggest that this may be due to the interference of maternal immunity.

Ongoing estimation of the epidemic parameters of a stochastic, spatial, discrete-time model for a 1983-84 avian influenza epidemic

We formulate a stochastic, spatial, discrete-time model of viral "Susceptible, Exposed, Infectious, Recovered" animal epidemics and apply it to an avian influenza epidemic in Pennsylvania in 1983-84. Using weekly data for the number of newly infectious cases collected during the epidemic, we find estimates for the latent period of the virus and the values of two parameters within the transmission kernel of the model. These data are then jackknifed on a progressive weekly basis to show how our estimates can be applied to an ongoing epidemic to generate continually improving values of certain epidemic parameters.