Modelling the field personnel resources to control foot-and-mouth disease outbreaks in New Zealand.

The objective of the study was to simulate New Zealand's foot-and-mouth disease (FMD) operational plan to determine personnel requirements for an FMD response and understand how the numbers of front-line staff available could affect the size and duration of FMD outbreaks, when using stamping-out (SO) measures with or without vaccination. The model utilized a national dataset of all known livestock farms. Each simulation randomly seeded infection into a single farm. Transmission mechanisms included direct and indirect contacts, local and airborne spread. Prior to each simulation, the numbers of personnel available for front-line tasks (including contact tracing, surveillance of at-risk farms, depopulation and vaccination) were set randomly. In a random subset of simulations, vaccination was allowed to be deployed as an adjunct to SO. The effects of personnel numbers on the size and duration of epidemics were explored using machine learning methods. In the second stage of the study, using a subset of iterations where numbers of personnel were unconstrained, the number of personnel used each day were quantified. When personnel resources were unconstrained, the 95th percentile and maximum number of infected places (IPs) were 78 and 462, respectively, and the 95th percentile and maximum duration were 69 and 217 days, respectively. However, severe constraints on personnel resources allowed some outbreaks to exceed the size of the UK 2001 FMD epidemic which had 2026 IPs. The number of veterinarians available had a major influence on the size and duration of outbreaks, whereas the availability of other personnel types did not. A shortage of veterinarians was associated with an increase in time to detect and depopulate IPs, allowing for continued transmission. Emergency vaccination placed a short-term demand for additional staff at the start of the vaccination programme, but the overall number of person days used was similar to SO-only strategies. This study determined the optimal numbers of front-line personnel required to implement the current operational plans to support an FMD response in New Zealand. A shortage of veterinarians was identified as the most influential factor to impact disease control outcomes. Emergency vaccination led to earlier control of FMD outbreaks but at the cost of a short-term spike in demand for personnel. In conclusion, a successful response needs to have access to sufficient personnel, particularly veterinarians, trained in response roles and available at short notice.

Use of scenario tree modelling to plan freedom from infection surveillance: Mycoplasma bovis in New Zealand.

Since mid-2018, the New Zealand (NZ) Ministry for Primary Industries (MPI) has been operating an eradication program for an incursion of Mycoplasma bovis. Although NZ is still delimiting the outbreak, consideration is being given to how freedom from M. bovis will be demonstrated. Rapid demonstration of freedom will minimise the length of the program, significantly reducing its financial burden. This collaborative research was undertaken to help inform planning of surveillance to demonstrate freedom after M. bovis is believed eradicated. Scenario tree modelling (STM) involves assimilating multiple surveillance system components to determine whether disease is absent. STM has infrequently been used to plan appropriate surveillance but this was the approach used here. A stochastic simulation model was implemented in R. The model represented the NZ commercial dairy and non-dairy cattle industries and the current surveillance components that are also planned to be used to gather evidence of absence of M. bovis once it is eradicated. Different surveillance intensities and risk based versus random surveillance were simulated and compared for probability of freedom, financial cost of sampling and testing and the time to demonstrate freedom. The results indicate that the current surveillance components will enable demonstration of freedom. Surveillance components included bulk tank milk testing, herd testing and testing at meat processing plants, predominantly using an imperfect ELISA. Several combinations of surveillance components appeared most efficient achieving >95 % confidence of freedom over 2-4 years, whilst sampling 4-7 % of the non-dairy herds and less than 25 % of dairy herds annually. The results indicate that surveillance intensity can be lower than is currently occurring to support the delimiting phase, thereby saving significant resources in the post eradication phase (proof of freedom phases). Further consideration is required to enable the assumption of 100 % herd specificity made in the model to be achieved. The ELISA used is very specific, but will yield some false positives that must be resolved to their true status. This may occur for example through modified diagnostic test interpretation (e.g. cut point optimisation at individual and herd level) or resolution of putative false positive herds with epidemiological investigation. In conclusion this research demonstrates the utility of STM for planning surveillance programs, and in this instance has highlighted efficient and effective surveillance components for demonstrating freedom from M. bovis in NZ. It also highlights the need to achieve 100 % specificity for M. bovis in herds tested during the proof of freedom phases.

Detection of Foot-and-Mouth Disease Virus in the Absence of Clinical Disease in Cattle and Buffalo in South East Asia.

Foot-and-mouth disease virus (FMDV) is widespread throughout much of the world, including parts of South East Asia. Surveillance is often limited in endemic areas, relying predominantly on passive outbreak reporting. As part of the World Organisation for Animal Health (OIE)'s South East Asia and China Foot-and-Mouth Disease Project (SEACFMD), field sampling was performed to help understand evidence of widespread virus exposure observed in previous studies. Serum and dry mucosal swabs were collected to evaluate the presence of FMDV RNA on the nasal, oral, and dorsal nasopharyngeal mucosal surfaces of 262 healthy cattle (n = 84 in Laos; n = 125 in Myanmar) and buffalo (n = 48 in Laos; n = 5 in Myanmar) immediately following slaughter in three slaughterhouses. Swabs and serum were tested by the OIE/FAO World Reference Laboratory for foot-and-mouth disease (WRLFMD) using pan-serotypic real-time reverse transcription-PCR (rRT-PCR) and serum was evaluated using the FMD PrioCHECK non-structural protein (NSP) ELISA. In total, 7.3% of animals had detectable FMDV RNA in one or more of the three sites including 5.3% of nasopharyngeal swabs, 2.3% of oral swabs, and 1.5% of nasal swabs. No FMDV RNA was detected in serum. Overall, 37.8% of animals were positive for NSP antibodies, indicating likely past natural exposure to FMDV. Results were comparable for Laos and Myanmar, and for both cattle and buffalo, and were not significantly different between age groups. Detectable FMDV RNA present on the oral and nasal mucosa of clinically-healthy large ruminants in Laos and Myanmar demonstrates the importance of sampling asymptomatic animals as part of surveillance, and may indicate that subclinical infection plays a role in the epidemiology of FMD in these countries.

Challenges and opportunities for using national animal datasets to support foot-and-mouth disease control.

National level databases of animal numbers, locations and movements provide the essential foundations for disease preparedness, outbreak investigations and control activities. These activities are particularly important for managing and mitigating the risks of high-impact transboundary animal disease outbreaks such as foot-and-mouth disease (FMD), which can significantly affect international trade access and domestic food security. In countries where livestock production systems are heavily subsidized by the government, producers are often required to provide detailed animal movement and demographic data as a condition of business. In the remaining countries, it can be difficult to maintain these types of databases and impossible to estimate the extent of missing or inaccurate information due to the absence of gold standard datasets for comparison. Consequently, competent authorities are often required to make decisions about disease preparedness and control based on available data, which may result in suboptimal outcomes for their livestock industries. It is important to understand the limitations of poor data quality as well as the range of methods that have been developed to compensate in both disease-free and endemic situations. Using FMD as a case example, this review first discusses the different activities that competent authorities use farm-level animal population data for to support (1) preparedness activities in disease-free countries, (2) response activities during an acute outbreak in a disease-free country, and (3) eradication and control activities in an endemic country. We then discuss (4) data requirements needed to support epidemiological investigations, surveillance, and disease spread modelling both in disease-free and endemic countries.

Informing adaptive management strategies: Evaluating a mechanism to predict the likely qualitative size of foot-and-mouth disease outbreaks in New Zealand using data available in the early response phase of simulated outbreaks.

The objective of the study was to define and then evaluate an early decision indicator (EDI) trigger that operated within the first 5 weeks of a response that would indicate a large and/or long outbreak of FMD was developing, to be able to inform control options within an adaptive management framework. To define the EDI trigger, a previous dataset of 10,000 simulated FMD outbreaks in New Zealand, controlled by the standard stamping-out approach, was re-analysed at various time points between Days 11 and 35 of each response to find threshold values of cumulative detected infected premises (IPs) that indicated upper quartile sized outbreaks and estimated dissemination rate (EDR) values that indicated sustained spread. Both sets of thresholds were then parameterized within the InterSpread Plus modelling framework, such that if either the cumulative IPs or the EDR exceeded the defined thresholds, the EDI trigger would fire. A new series of simulations were then generated. The EDI trigger was like two diagnostic tests interpreted in parallel, with the diagnostic outcome positive if either test was positive at any time point between Days 11 and 35 inclusive. The diagnostic result was then compared to the final size of each outbreak, to see if the outbreak was an upper quartile outbreak in terms of cumulative IPs and/or final duration. The performance of the EDI trigger was then evaluated across the population of outbreaks, and the sensitivity (Se), specificity (Sp), positive predictive value (PPV) and negative predictive value (NPV) were calculated. The Se, Sp, PPV and NPV for predicting large outbreaks were 0.997, 0.513, 0.404 and 0.998, respectively. The study showed that the EDI trigger was very sensitive to detecting large outbreaks, although not all outbreaks predicted to be large were so, whereas outbreaks predicted to be small invariably were small. Therefore, it shows promise as a mechanism that could support an adaptive management approach to FMD control.

Mycoplasma bovis outbreak in New Zealand cattle: An assessment of transmission trends using surveillance data.

Mycoplasma bovis most likely infected New Zealand cattle in the latter half of 2015. Infection was detected in mid-2017 after which control activities were implemented. An official eradication programme commenced in mid-2018, which is ongoing. We examined farm-level tracing and surveillance data to describe the outbreak, analyse transmission trends and make inference on progress towards eradication. Results indicate that cattle movements were the primary means of spread. Although case farms were distributed throughout both islands of New Zealand, most animal movements off infected farms did not result in newly infected farms, indicating Mycoplasma bovis is not highly transmissible between farms. To describe and analyse outbreak trends, we undertook a standard descriptive outbreak investigation, including construction of an epidemic curve and calculation of estimated dissemination ratios. We then employed three empirical models-a non-linear growth model, time series model and branching process model based on time-varying effective reproduction numbers-to further analyse transmission trends and provide short-term forecasts of farm-level incidence. Our analyses suggest that Mycoplasma bovis transmission in New Zealand has declined and progress towards eradication has been made. Few incident cases were forecast for the period between 8 September and 17 December 2019. To date, no case farms with an estimated infection date assigned to this period have been detected; however, case detection is ongoing, and these results need to be interpreted cautiously considering model validation and other important contextual information on performance of the eradication programme, such as the time between infection, detection and implementation of movement controls on case farms.

Real-Time Standard Analysis of Disease Investigation (SADI)-A Toolbox Approach to Inform Disease Outbreak Response.

An incursion of an important exotic transboundary animal disease requires a prompt and intensive response. The routine analysis of up-to-date data, as near to real time as possible, is essential for the objective assessment of the patterns of disease spread or effectiveness of control measures and the formulation of alternative control strategies. In this paper, we describe the Standard Analysis of Disease Investigation (SADI), a toolbox for informing disease outbreak response, which was developed as part of New Zealand's biosecurity preparedness. SADI was generically designed on a web-based software platform, Integrated Real-time Information System (IRIS). We demonstrated the use of SADI for a hypothetical foot-and-mouth disease (FMD) outbreak scenario in New Zealand. The data standards were set within SADI, accommodating a single relational database that integrated the national livestock population data, outbreak data, and tracing data. We collected a well-researched, standardised set of 16 epidemiologically relevant analyses for informing the FMD outbreak response, including farm response timelines, interactive outbreak/network maps, stratified epidemic curves, estimated dissemination rates, estimated reproduction numbers, and areal attack rates. The analyses were programmed within SADI to automate the process to generate the reports at a regular interval (daily) using the most up-to-date data. Having SADI prepared in advance and the process streamlined for data collection, analysis and reporting would free a wider group of epidemiologists during an actual disease outbreak from solving data inconsistency among response teams, daily "number crunching," or providing largely retrospective analyses. Instead, the focus could be directed into enhancing data collection strategies, improving data quality, understanding the limitations of the data available, interpreting the set of analyses, and communicating their meaning with response teams, decision makers and public in the context of the epidemic.

Efficacy of citric acid and sodium hypochlorite as disinfectants against Mycoplasma bovis.

Mycoplasma bovis, a cattle pathogen of major economic importance across the globe, causes a range of diseases, including pneumonia and mastitis. Because of the limited options for effective treatment of these diseases, prevention and control are preferred to diagnosis and treatment. In this study, the efficacies of citric acid and sodium hypochlorite as disinfectants against M. bovis were tested using a modification of a standardised method for assessing the efficacy of disinfectants against bacteria. A citric acid concentration of 0.5 % was found to be an effective disinfectant, reducing infectivity by close to 106 fold, while sodium hypochlorite at 1% was found to have similar efficacy to 0.5 % citric acid. A 0.04 % concentration of sodium hypochlorite was effective against M. bovis only in the absence of any organic material. Under these conditions, 0.25 % citric acid found to have similar efficacy. These findings indicate that 0.5 % citric acid or 1 % sodium hypochlorite are likely to be effective disinfectants for M. bovis under field conditions and 0.04 % sodium hypochlorite or 0.25 % citric acid are likely to be effective following removal of organic material.

Evaluating the utility of national-scale data to estimate the local risk of foot-and-mouth disease in endemic regions.

Knowledge of the distribution of foot-and-mouth disease (FMD) is required if control programmes are to be successful. However, data on the seroprevalence and incidence of affected villages in developing countries with endemic disease are scarce. This is partly due to resource constraints as well as the logistical challenges of conducting intensive surveys and diagnostic testing in remote locations. In this study, we evaluated the performance of low resolution national-scale data against high resolution local survey data to predict the FMD serological status of 168 villages in the Mandalay and Sagaing Regions of central Myanmar using both logistic regression and random forest modelling approaches. Blood samples for ELISA testing were collected from approximately 30 cattle per village in both the 6 to 18 month age range and in the over 18 month age range to distinguish between recent and historical exposure, respectively. The results of the animal level tests were aggregated to the village level to provide the outcome of interest (village positive or not positive for FMD), and three explanatory data sets were constructed: using only nationally available data, using only data collected by survey and using the combined survey and nationally available data. The true seroprevalence of FMD at the village level was 61% when only young animals were included, but increased to 87% when all animals were included. The best performing model was a logistic regression model using the combined national and survey data to predict recent infection in villages. However, this still incorrectly classified 40% of villages, which suggests that using national-level data were not reliable enough for extrapolating seroprevalence in regions where conducting detailed surveys is impractical. Other methods for collected data on FMD such as the use of local reporting should be explored.

Estimating foot-and-mouth disease (FMD) prevalence in central Myanmar: Comparison of village headman and farmer disease reports with serological findings.

The impacts of foot-and-mouth disease (FMD) on food security in developing countries are difficult to quantify due to the scarcity of accurate data on the prevalence and incidence of affected villages. This is partly due to resource constraints as well as the logistical challenges of conducting regular diagnostic testing in remote locations. In this study, we used descriptive analysis and latent class analysis (LCA) models to analyse data collected during a field survey of 160 villages in central Myanmar in the Mandalay and Sagaing Regions over the 2012-2016 time period. We evaluated the performance of verbal reports made by village householders and headmen against serological data to retrospectively determine the FMD-infection status of our study area and to identify factors contributing to under-reporting. Blood samples were collected from approximately 30 cattle per village in both the 6- to 18-month age range and over 18-month age range to distinguish between recent and historic exposure. Village householders were asked to identify pictures of FMD-affected cattle amongst pictures of cattle affected with other common endemic diseases to assess the accuracy of their verbal reporting. The serological results confirmed that FMD is endemic in central Myanmar with village-level seroprevalence estimated at 56% for animals 6-18 months of age and 80% when all age groups were considered together. Most village householders were familiar with the clinical signs of FMD-affected cattle (72%). Based on the results from the LCA models, the village headman had a sensitivity of 77% and specificity of 75% for identifying FMD outbreaks in their village, whereas individual householders had a higher sensitivity and lower specificity of 80% and 56%, respectively. The level of disagreement between the different sources was correlated with the total number of cattle in the village and may potentially be worse in villages where endemic FMD may have led to a high level of natural immunity in cattle and subsequent masking of clinical signs. However, other regional effects such as the intensity of FMD extension efforts cannot be ruled out. Overall, the results suggest that verbal reports of FMD outbreaks from village headmen may be a useful tool to integrate into active FMD surveillance programmes in developing countries.

Does Size Matter to Models? Exploring the Effect of Herd Size on Outputs of a Herd-Level Disease Spread Simulator.

Disease spread modeling is widely used by veterinary authorities to predict the impact of emergency animal disease outbreaks in livestock and to evaluate the cost-effectiveness of different management interventions. Such models require knowledge of basic disease epidemiology as well as information about the population of animals at risk. Essential demographic information includes the production system, animal numbers, and their spatial locations yet many countries with significant livestock industries do not have publically available and accurate animal population information at the farm level that can be used in these models. The impact of inaccuracies in data on model outputs and the decisions based on these outputs is seldom discussed. In this analysis, we used the Australian Animal Disease model to simulate the spread of foot-and-mouth disease seeded into high-risk herds in six different farming regions in New Zealand. We used three different susceptible animal population datasets: (1) a gold standard dataset comprising known herd sizes, (2) a dataset where herd size was simulated from a beta-pert distribution for each herd production type, and (3) a dataset where herd size was simplified to the median herd size for each herd production type. We analyzed the model outputs to compare (i) the extent of disease spread, (ii) the length of the outbreaks, and (iii) the possible impacts on decisions made for simulated outbreaks in different regions. Model outputs using the different datasets showed statistically significant differences, which could have serious implications for decision making by a competent authority. Outbreak duration, number of infected properties, and vaccine doses used during the outbreak were all significantly smaller for the gold standard dataset when compared with the median herd size dataset. Initial outbreak location and disease control strategy also significantly influenced the duration of the outbreak and number of infected premises. The study findings demonstrate the importance of having accurate national-level population datasets to ensure effective decisions are made before and during disease outbreaks, reducing the damage and cost.

Predicting farm-level animal populations using environmental and socioeconomic variables.

Accurate information on the geographic distribution of domestic animal populations helps biosecurity authorities to efficiently prepare for and rapidly eradicate exotic diseases, such as Foot and Mouth Disease (FMD). Developing and maintaining sufficiently high-quality data resources is expensive and time consuming. Statistical modelling of population density and distribution has only begun to be applied to farm animal populations, although it is commonly used in wildlife ecology. We developed zero-inflated Poisson regression models in a Bayesian framework using environmental and socioeconomic variables to predict the counts of livestock units (LSUs) and of cattle on spatially referenced farm polygons in a commercially available New Zealand farm database, Agribase. Farm-level counts of cattle and of LSUs varied considerably by region, because of the heterogeneous farming landscape in New Zealand. The amount of high quality pasture per farm was significantly associated with the presence of both cattle and LSUs. Internal model validation (predictive performance) showed that the models were able to predict the count of the animal population on groups of farms that were located in randomly selected 3km zones with a high level of accuracy. Predicting cattle or LSU counts on individual farms was less accurate. Predicted counts were statistically significantly more variable for farms that were contract grazing dry stock, such as replacement dairy heifers and dairy cattle not currently producing milk, compared with other farm types. This analysis presents a way to predict numbers of LSUs and cattle for farms using environmental and socio-economic data. The technique has the potential to be extrapolated to predicting other pastoral based livestock species.

Species distribution models: A comparison of statistical approaches for livestock and disease epidemics.

In livestock industries, reliable up-to-date spatial distribution and abundance records for animals and farms are critical for governments to manage and respond to risks. Yet few, if any, countries can afford to maintain comprehensive, up-to-date agricultural census data. Statistical modelling can be used as a proxy for such data but comparative modelling studies have rarely been undertaken for livestock populations. Widespread species, including livestock, can be difficult to model effectively due to complex spatial distributions that do not respond predictably to environmental gradients. We assessed three machine learning species distribution models (SDM) for their capacity to estimate national-level farm animal population numbers within property boundaries: boosted regression trees (BRT), random forests (RF) and K-nearest neighbour (K-NN). The models were built from a commercial livestock database and environmental and socio-economic predictor data for New Zealand. We used two spatial data stratifications to test (i) support for decision making in an emergency response situation, and (ii) the ability for the models to predict to new geographic regions. The performance of the three model types varied substantially, but the best performing models showed very high accuracy. BRTs had the best performance overall, but RF performed equally well or better in many simulations; RFs were superior at predicting livestock numbers for all but very large commercial farms. K-NN performed poorly relative to both RF and BRT in all simulations. The predictions of both multi species and single species models for farms and within hypothetical quarantine zones were very close to observed data. These models are generally applicable for livestock estimation with broad applications in disease risk modelling, biosecurity, policy and planning.

Network analysis of translocated Takahe populations to identify disease surveillance targets.

Social network analysis is being increasingly used in epidemiology and disease modeling in humans, domestic animals, and wildlife. We investigated this tool in describing a translocation network (area that allows movement of animals between geographically isolated locations) used for the conservation of an endangered flightless rail, the Takahe (Porphyrio hochstetteri). We collated records of Takahe translocations within New Zealand and used social network principles to describe the connectivity of the translocation network. That is, networks were constructed and analyzed using adjacency matrices with values based on the tie weights between nodes. Five annual network matrices were created using the Takahe data set, each incremental year included records of previous years. Weights of movements between connected locations were assigned by the number of Takahe moved. We calculated the number of nodes (i(total)) and the number of ties (t(total)) between the nodes. To quantify the small-world character of the networks, we compared the real networks to random graphs of the equivalent size, weighting, and node strength. Descriptive analysis of cumulative annual Takahe movement networks involved determination of node-level characteristics, including centrality descriptors of relevance to disease modeling such as weighted measures of in degree (k(i)(in)), out degree (k(i)(out)), and betweenness (B(i)). Key players were assigned according to the highest node measure of k(i)(in), k(i)(out), and B(i) per network. Networks increased in size throughout the time frame considered. The network had some degree small-world characteristics. Nodes with the highest cumulative tie weights connecting them were the captive breeding center, the Murchison Mountains and 2 offshore islands. The key player fluctuated between the captive breeding center and the Murchison Mountains. The cumulative networks identified the captive breeding center every year as the hub of the network until the final network in 2011. Likewise, the wild Murchison Mountains population was consistently the sink of the network. Other nodes, such as the offshore islands and the wildlife hospital, varied in importance over time. Common network descriptors and measures of centrality identified key locations for targeting disease surveillance. The visual representation of movements of animals in a population that this technique provides can aid decision makers when they evaluate translocation proposals or attempt to control a disease outbreak.