Adding the spatial dimension to the social network analysis of an epidemic: investigation of the 2007 outbreak of equine influenza in Australia

Social Network Analysis of Ebola Virus Disease During the 2014 Outbreak in Sukudu, Sierra Leone

Background

Transmission by unreported cases has been proposed as a reason for the 2013-2016 Ebola virus (EBOV) epidemic decline in West Africa, but studies that test this hypothesis are lacking. We examined a transmission chain within social networks in Sukudu village to assess spread and transmission burnout.

Methods

Network data were collected in 2 phases: (1) serological and contact information from Ebola cases (n = 48, including unreported); and (2) interviews (n = 148), including Ebola survivors (n = 13), to identify key social interactions. Social links to the transmission chain were used to calculate cumulative incidence proportion as the number of EBOV-infected people in the network divided by total network size.

Results

The sample included 148 participants and 1522 contacts, comprising 10 social networks: 3 had strong links (>50% of cases) to the transmission chain: household sharing (largely kinship), leisure time, and talking about important things (both largely non-kin). Overall cumulative incidence for these networks was 37 of 311 (12%). Unreported cases did not have higher network centrality than reported cases.

Conclusions

Although this study did not find evidence that explained epidemic decline in Sukudu, it excluded potential reasons (eg, unreported cases, herd immunity) and identified 3 social interactions in EBOV transmission.

Coupling spatial statistics with social network analysis to estimate distinct risk areas of disease circulation to improve risk-based surveillance

Most animal disease surveillance systems concentrate efforts in blocking transmission pathways and tracing back infected contacts while not considering the risk of transporting animals into areas with elevated disease risk. Here, we use a suite of spatial statistics and social network analysis to characterize animal movement among areas with an estimated distinct risk of disease circulation to ultimately enhance surveillance activities. Our model utilized equine infectious anemia virus (EIAV) outbreaks, between-farm horse movements, and spatial landscape data from 2015 through 2017. We related EIAV occurrence and the movement of horses between farms with climate variables that foster conditions for local disease propagation. We then constructed a spatially explicit model that allows the effect of the climate variables on EIAV occurrence to vary through space (i.e., non-stationary). Our results identified important areas in which in-going movements were more likely to result in EIAV infections and disease propagation. Municipalities were then classified as having high 56 (11.3%), medium 48 (9.66%), and low 393 (79.1%) spatial risk. The majority of the movements were between low-risk areas, altogether representing 68.68% of all animal movements. Meanwhile, 9.48% were within high-risk areas, and 6.20% were within medium-risk areas. Only 5.37% of the animals entering low-risk areas came from high-risk areas. On the other hand, 4.91% of the animals in the high-risk areas came from low- and medium-risk areas. Our results demonstrate that animal movements and spatial risk mapping could be used to make informed decisions before issuing animal movement permits, thus potentially reducing the chances of reintroducing infection into areas of low risk.

Analyzing Spatial Dependency of the 2016-2017 Korean HPAI Outbreak to Determine the Effective Culling Radius

Highly pathogenic avian influenza (HPAI) outbreaks are a threat to human health and cause extremely large financial losses to the poultry industry due to containment measures. Determining the most effective control measures, especially the culling radius, to minimize economic impacts yet contain the spread of HPAI is of great importance. This study examines the factors influencing the probability of a farm being infected with HPAI during the 2016-2017 HPAI outbreak in Korea. Using a spatial random effects logistic model, only a few factors commonly associated with a higher risk of HPAI infection were significant. Interestingly, most density-related factors, poultry and farm, were not significantly associated with a higher risk of HPAI infection. The effective culling radius was determined to be two ranges: 0.5-2.2 km and 2.7-3.0 km. This suggests that the spatial heterogeneity, due to local characteristics and/or the characteristics of the HPAI virus(es) involved, should be considered to determine the most effective culling radius in each region. These findings will help strengthen biosecurity control measures at the farm level and enable authorities to quickly respond to HPAI outbreaks with effective countermeasures to suppress the spread of HPAI.

Effects of Spatial Characteristics on the Spread of the Highly Pathogenic Avian Influenza (HPAI) in Korea

This paper examines the effects of regional characteristics on the spread of the highly pathogenic avian influenza (HPAI) during Korea’s 2016–2017 outbreak. A spatial econometric model is used to determine the effects of regional characteristics on HPAI dispersion using data from 162 counties in Korea. Results indicate the existence of spatial dependence, suggesting that the occurrence of HPAI in a county is significantly influenced by neighboring counties. We found that larger size poultry, including laying hens, breeders, and ducks are significantly associated with a greater incidence of HPAI. Among poultry, we found ducks as the greatest source of the spread of HPAI. Our findings suggest that those regions that are spatially dependent with respect to the spread of HPAI, such as counties that intensively breed ducks, should be the focus of surveillance to prevent future epidemics of HPAI.

Social structure defines spatial transmission of African swine fever in wild boar

The spatial spread of infectious disease is determined by spatial and social processes such as animal space use and family group structure. Yet, the impacts of social processes on spatial spread remain poorly understood and estimates of spatial transmission kernels (STKs) often exclude social structure. Understanding the impacts of social structure on STKs is important for obtaining robust inferences for policy decisions and optimizing response plans. We fit spatially explicit transmission models with different assumptions about contact structure to African swine fever virus surveillance data from eastern Poland from 2014 to 2015 and evaluated how social structure affected inference of STKs and spatial spread. The model with social structure provided better inference of spatial spread, predicted that approximately 80% of transmission events occurred within family groups, and that transmission was weakly female-biased (other models predicted weakly male-biased transmission). In all models, most transmission events were within 1.5 km, with some rare events at longer distances. Effective reproductive numbers were between 1.1 and 2.5 (maximum values between 4 and 8). Social structure can modify spatial transmission dynamics. Accounting for this additional contact heterogeneity in spatial transmission models could provide more robust inferences of STKs for policy decisions, identify best control targets and improve transparency in model uncertainty.

Spatiotemporal analysis and hotspots detection of COVID-19 using geographic information system (March and April, 2020)

Understanding the spatial distribution of coronavirus disease 2019 (COVID-19) cases can provide valuable information to anticipate the world outbreaks and in turn improve public health policies. In this study, the cumulative incidence rate (CIR) and cumulative mortality rate (CMR) of all countries affected by the new corona outbreak were calculated at the end of March and April, 2020. Prior to the implementation of hot spot analysis, the spatial autocorrelation results of CIR were obtained. Hot spot analysis and Anselin Local Moran's I indices were then applied to accurately locate high and low-risk clusters of COVID-19 globally. San Marino and Italy revealed the highest CMR by the end of March, though Belgium took the place of Italy as of 30th April. At the end of the research period (by 30th April), the CIR showed obvious spatial clustering. Accordingly, southern, northern and western Europe were detected in the high-high clusters demonstrating an increased risk of COVID-19 in these regions and also the surrounding areas. Countries of northern Africa exhibited a clustering of hot spots, with a confidence level above 95%, even though these areas assigned low CIR values. The hot spots accounted for nearly 70% of CIR. Furthermore, analysis of clusters and outliers demonstrated that these countries are situated in the low-high outlier pattern. Most of the surveyed countries that exhibited clustering of high values (hot spot) with a confidence level of 99% (by 31st March) and 95% (by 30th April) were dedicated higher CIR values. In conclusion, hot spot analysis coupled with Anselin local Moran's I provides a scrupulous and objective approach to determine the locations of statistically significant clusters of COVID-19 cases shedding light on the high-risk districts.

Quantitative Risk Assessment for the Introduction of Bovine Leukemia Virus-Infected Cattle Using a Cattle Movement Network Analysis

The cattle industry is suffering economic losses caused by bovine leukemia virus (BLV) and enzootic bovine leukosis (EBL), the clinical condition associated with BLV infection. This pathogen spreads easily without detection by farmers and veterinarians due to the lack of obvious clinical signs. Cattle movement strongly contributes to the inter-farm transmission of BLV. This study quantified the farm-level risk of BLV introduction using a cattle movement analysis. A generalized linear mixed model predicting the proportion of BLV-infected cattle was constructed based on weighted in-degree centrality. Our results suggest a positive association between weighted in-degree centrality and the estimated number of introduced BLV-infected cattle. Remarkably, the introduction of approximately six cattle allowed at least one BLV-infected animal to be added to the farm in the worst-case scenario. These data suggest a high risk of BLV infection on farms with a high number of cattle being introduced. Our findings indicate the need to strengthen BLV control strategies, especially along the chain of cattle movement.

Epidemic investigations within an arm's reach - role of google maps during an epidemic outbreak

Epidemics such as novel Coronavirus 2019 (COVID-19) can be contained and the rate of infection reduced by public health measures such as epidemiologic inquiries and social distancing. Epidemiologic inquiry requires resources and time which may not be available or reduced when the outbreak is excessive. We evaluated the use of Google Maps Timeline (GMTL) for creating spatial epidemiologic timelines. The study compares locations, routes, and means of transport between GMTL and user recall for 17 suitable users who were recruited during March 2020. They were interviewed about their timeline using the Timeline Follow-Back (TLFB) method which was then compared to their GMTL and discrepancies between both methods were addressed. Interviewer conclusions were divided into categories: (1) participant recalled, (2) no recall (until shown). Categories were subdivided by GMTL accuracy: [a] GMTL accurate, [b] GMTL inaccurate, [c] GMTL data missing. A total of 362 locations were compared. Participants recalled 322 (88.95% SD = 8.55) locations compared with 40 (11.05%, SD = 2.05) locations not recalled. There were 304 locations found accurate on GMTL (83.98%, SD = 9.49), 29 (8.01%, SD = 1.11) inaccurate locations, and 29 (8.01%, SD = 0.54) missing locations. The total discrepancy between GMTL and TLFB recall was 95 cases (26.24%, SD = 3.25). Despite variations between users, Google Maps with GMTL technology may be useful in identifying potentially exposed individuals in a pandemic. It is especially useful when resources are limited. Further research is required with a larger number of users who are undergoing a real epidemiologic investigation to corroborate findings and establish further recommendations.

The Relationship between the Migrant Population's Migration Network and the Risk of COVID-19 Transmission in China-Empirical Analysis and Prediction in Prefecture-Level Cities

The outbreak of COVID-19 in China has attracted wide attention from all over the world. The impact of COVID-19 has been significant, raising concerns regarding public health risks in China and worldwide. Migration may be the primary reason for the long-distance transmission of the disease. In this study, the following analyses were performed. (1) Using the data from the China migrant population survey in 2017 (Sample size = 432,907), a matrix of the residence-birthplace (R-B matrix) of migrant populations is constructed. The matrix was used to analyze the confirmed cases of COVID-19 at Prefecture-level Cities from February 1-15, 2020 after the outbreak in Wuhan, by calculating the probability of influx or outflow migration. We obtain a satisfactory regression analysis result (R2 = 0.826-0.887, N = 330). (2) We use this R-B matrix to simulate an outbreak scenario in 22 immigrant cities in China, and propose risk prevention measures after the outbreak. If similar scenarios occur in the cities of Wenzhou, Guangzhou, Dongguan, or Shenzhen, the disease transmission will be wider. (3) We also use a matrix to determine that cities in Henan province, Anhui province, and Municipalities (such as Beijing, Shanghai, Guangzhou, Shenzhen, Chongqing) in China will have a high risk level of disease carriers after a similar emerging epidemic outbreak scenario due to a high influx or outflow of migrant populations.

Preparedness Through Urban Resilience

This book explores the topic of resilience at the city level. The focus is more on outbreak events at the city level, or how cities should prepare and react in facing the larger events of epidemic and pandemic.

Descriptive analysis of horse movement networks during the 2015 equestrian season in Ontario, Canada

Horses are a highly mobile population, with many travelling locally, nationally, and internationally to participate in shows and sporting events. However, the nature and extent of these movements, as well as the potential impact they may have on disease introduction and spread, is not well documented. The objective of this study was to characterise the movement network of a sample of horses in Ontario, Canada, over a 7-month equestrian season. Horse owners (n = 141) documented their travel patterns with their horse(s) (n = 330) by completing monthly online questionnaires between May and November 2015. Directed networks were constructed to represent horse movements in 1-month time periods. A total of 1754 horse movements met the inclusion criteria for analysis. A variety of location types were included in each monthly network, with many including non-facilities such as parks, trails, and private farms. Only 34.3% of competitions attended by participants during the study period were regulated by an official equestrian organisation. Comparisons of the similarity between monthly networks indicated that participants did not travel to the same locations each month, and the most connected locations varied between consecutive months. While the findings should not be generalized to the wider horse population, they have provided greater insight into the nature and extent of observed horse movement patterns. The results support the need to better understand the variety of locations to which horses can travel in Ontario, as different types of locations may have different associated risks of disease introduction and spread.

Burkholderia mallei: The dynamics of networks and disease transmission

Glanders is a highly infectious zoonotic disease caused by Burkholderia mallei. The transmission of B. mallei occurs mainly by direct contact, and horses are the natural reservoir. Therefore, the identification of infection sources within horse populations and animal movements is critical to enhance disease control. Here, we analysed the dynamics of horse movements from 2014 to 2016 using network analysis in order to understand the flow of animals in two hierarchical levels, municipalities and farms. The municipality-level network was used to investigate both community clustering and the balance between the municipality's trades and the farm-level network associations between B. mallei outbreaks and the network centrality measurements, analysed by spatio-temporal generalized additive model (GAM). Causal paths were established for the dispersion of B. mallei outbreaks through the network. Our approach captured and established a direct relationship between movement of infected equines and predicted B. mallei outbreaks. The GAM model revealed that the parameters in degree and closeness centrality out were positively associated with B. mallei. In addition, we also detected 10 communities with high commerce among municipalities. The role of each municipality within the network was detailed, and significant changes in the structures of the network were detected over the course of 3 years. The results suggested the necessity to focus on structural changes of the networks over time to better control glanders disease. The identification of farms with a putative risk of B. mallei infection using the horse movement network provided a direct opportunity for disease control through active surveillance, thus minimizing economic losses and risks for human cases of B. mallei.

VetCompass Australia: A National Big Data Collection System for Veterinary Science

VetCompass Australia is veterinary medical records-based research coordinated with the global VetCompass endeavor to maximize its quality and effectiveness for Australian companion animals (cats, dogs, and horses). Bringing together all seven Australian veterinary schools, it is the first nationwide surveillance system collating clinical records on companion-animal diseases and treatments. VetCompass data service collects and aggregates real-time, clinical records for researchers to interrogate, delivering sustainable and cost-effective access to data from hundreds of veterinary practitioners nationwide. Analysis of these clinical records will reveal geographical and temporal trends in the prevalence of inherited and acquired diseases, identify frequently prescribed treatments, revolutionize clinical auditing, help the veterinary profession to rank research priorities, and assure evidence-based companion-animal curricula in veterinary schools. VetCompass Australia will progress in three phases: (1) roll-out of the VetCompass platform to harvest Australian veterinary clinical record data; (2) development and enrichment of the coding (data-presentation) platform; and (3) creation of a world-first, real-time surveillance interface with natural language processing (NLP) technology. The first of these three phases is described in the current article. Advances in the collection and sharing of records from numerous practices will enable veterinary professionals to deliver a vastly improved level of care for companion animals that will improve their quality of life.

Attitudes Toward the Ethics of Research Using Social Media: A Systematic Review

BACKGROUND

Although primarily used for social networking and often used for social support and dissemination, data on social media platforms are increasingly being used to facilitate research. However, the ethical challenges in conducting social media research remain of great concern. Although much debated in the literature, it is the views of the public that are most pertinent to inform future practice.

OBJECTIVE

The aim of our study was to ascertain attitudes on the ethical considerations of using social media as a data source for research as expressed by social media users and researchers.

METHODS

A systematic review was conducted, wherein 16 databases and 2 Internet search engines were searched in addition to handsearching, reference checking, citation searching, and contacting authors and experts. Studies that conducted any qualitative methods to collect data on attitudes on the ethical implications of research using social media were included. Quality assessment was conducted using the quality of reporting tool (QuaRT) and findings analyzed using inductive thematic synthesis.

RESULTS

In total, 17 studies met the inclusion criteria. Attitudes varied from overly positive with people expressing the views about the essential nature of such research for the public good, to very concerned with views that social media research should not happen. Underlying reasons for this variation related to issues such as the purpose and quality of the research, the researcher affiliation, and the potential harms. The methods used to conduct the research were also important. Many respondents were positive about social media research while adding caveats such as the need for informed consent or use restricted to public platforms only.

CONCLUSIONS

Many conflicting issues contribute to the complexity of good ethical practice in social media research. However, this should not deter researchers from conducting social media research. Each Internet research project requires an individual assessment of its own ethical issues. Guidelines on ethical conduct should be based on current evidence and standardized to avoid discrepancies between, and duplication across, different institutions, taking into consideration different jurisdictions.

Emergence of the giant weak component in directed random graphs with arbitrary degree distributions

The weak component generalizes the idea of connected components to directed graphs. In this paper, an exact criterion for the existence of the giant weak component is derived for directed graphs with arbitrary bivariate degree distributions. In addition, we consider a random process for evolving directed graphs with bounded degrees. The bounds are not the same for different vertices but satisfy a predefined distribution. The analytic expression obtained for the evolving degree distribution is then combined with the weak-component criterion to obtain the exact time of the phase transition. The phase-transition time is obtained as a function of the distribution that bounds the degrees. Remarkably, when viewed from the step-polymerization formalism, the new results yield Flory-Stockmayer gelation theory and generalize it to a broader scope.

Using heterogeneity in the population structure of U.S. swine farms to compare transmission models for porcine epidemic diarrhoea

In 2013, U.S. swine producers were confronted with the disruptive emergence of porcine epidemic diarrhoea (PED). Movement of animals among farms is hypothesised to have played a role in the spread of PED among farms. Via this or other mechanisms, the rate of spread may also depend on the geographic density of farms and climate. To evaluate such effects on a large scale, we analyse state-level counts of outbreaks with variables describing the distribution of farm sizes and types, aggregate flows of animals among farms, and an index of climate. Our first main finding is that it is possible for a correlation analysis to be sensitive to transmission model parameters. This finding is based on a global sensitivity analysis of correlations on simulated data that included a biased and noisy observation model based on the available PED data. Our second main finding is that flows are significantly associated with the reports of PED outbreaks. This finding is based on correlations of pairwise relationships and regression modeling of total and weekly outbreak counts. These findings illustrate how variation in population structure may be employed along with observational data to improve understanding of disease spread.

Sample size considerations for livestock movement network data

The movement of animals between farms contributes to infectious disease spread in production animal populations, and is increasingly investigated with social network analysis methods. Tangible outcomes of this work include the identification of high-risk premises for targeting surveillance or control programs. However, knowledge of the effect of sampling or incomplete network enumeration on these studies is limited. In this study, a simulation algorithm is presented that provides an estimate of required sampling proportions based on predicted network size, density and degree value distribution. The algorithm may be applied a priori to ensure network analyses based on sampled or incomplete data provide population estimates of known precision. Results demonstrate that, for network degree measures, sample size requirements vary with sampling method. The repeatability of the algorithm output under constant network and sampling criteria was found to be consistent for networks with at least 1000 nodes (in this case, farms). Where simulated networks can be constructed to closely mimic the true network in a target population, this algorithm provides a straightforward approach to determining sample size under a given sampling procedure for a network measure of interest. It can be used to tailor study designs of known precision, for investigating specific livestock movement networks and their impact on disease dissemination within populations.

Spatial and temporal epidemiological analysis in the Big Data era

Concurrent with global economic development in the last 50 years, the opportunities for the spread of existing diseases and emergence of new infectious pathogens, have increased substantially. The activities associated with the enormously intensified global connectivity have resulted in large amounts of data being generated, which in turn provides opportunities for generating knowledge that will allow more effective management of animal and human health risks. This so-called Big Data has, more recently, been accompanied by the Internet of Things which highlights the increasing presence of a wide range of sensors, interconnected via the Internet. Analysis of this data needs to exploit its complexity, accommodate variation in data quality and should take advantage of its spatial and temporal dimensions, where available. Apart from the development of hardware technologies and networking/communication infrastructure, it is necessary to develop appropriate data management tools that make this data accessible for analysis. This includes relational databases, geographical information systems and most recently, cloud-based data storage such as Hadoop distributed file systems. While the development in analytical methodologies has not quite caught up with the data deluge, important advances have been made in a number of areas, including spatial and temporal data analysis where the spectrum of analytical methods ranges from visualisation and exploratory analysis, to modelling. While there used to be a primary focus on statistical science in terms of methodological development for data analysis, the newly emerged discipline of data science is a reflection of the challenges presented by the need to integrate diverse data sources and exploit them using novel data- and knowledge-driven modelling methods while simultaneously recognising the value of quantitative as well as qualitative analytical approaches. Machine learning regression methods, which are more robust and can handle large datasets faster than classical regression approaches, are now also used to analyse spatial and spatio-temporal data. Multi-criteria decision analysis methods have gained greater acceptance, due in part, to the need to increasingly combine data from diverse sources including published scientific information and expert opinion in an attempt to fill important knowledge gaps. The opportunities for more effective prevention, detection and control of animal health threats arising from these developments are immense, but not without risks given the different types, and much higher frequency, of biases associated with these data.

Modelling influenza A(H1N1) 2009 epidemics using a random network in a distributed computing environment

In this paper we propose the use of a random network model for simulating and understanding the epidemics of influenza A(H1N1). The proposed model is used to simulate the transmission process of influenza A(H1N1) in a community region of Venezuela using distributed computing in order to accomplish many realizations of the underlying random process. These large scale epidemic simulations have recently become an important application of high-performance computing. The network model proposed performs better than the traditional epidemic model based on ordinary differential equations since it adjusts better to the irregularity of the real world data. In addition, the network model allows the consideration of many possibilities regarding the spread of influenza at the population level. The results presented here show how well the SEIR model fits the data for the AH1N1 time series despite the irregularity of the data and returns parameter values that are in good agreement with the medical data regarding AH1N1 influenza virus. This versatile network model approach may be applied to the simulation of the transmission dynamics of several epidemics in human networks. In addition, the simulation can provide useful information for the understanding, prediction and control of the transmission of influenza A(H1N1) epidemics.

Flying-fox species density--a spatial risk factor for Hendra virus infection in horses in eastern Australia

Hendra virus causes sporadic but typically fatal infection in horses and humans in eastern Australia. Fruit-bats of the genus Pteropus (commonly known as flying-foxes) are the natural host of the virus, and the putative source of infection in horses; infected horses are the source of human infection. Effective treatment is lacking in both horses and humans, and notwithstanding the recent availability of a vaccine for horses, exposure risk mitigation remains an important infection control strategy. This study sought to inform risk mitigation by identifying spatial and environmental risk factors for equine infection using multiple analytical approaches to investigate the relationship between plausible variables and reported Hendra virus infection in horses. Spatial autocorrelation (Global Moran's I) showed significant clustering of equine cases at a distance of 40 km, a distance consistent with the foraging 'footprint' of a flying-fox roost, suggesting the latter as a biologically plausible basis for the clustering. Getis-Ord Gi* analysis identified multiple equine infection hot spots along the eastern Australia coast from far north Queensland to central New South Wales, with the largest extending for nearly 300 km from southern Queensland to northern New South Wales. Geographically weighted regression (GWR) showed the density of P. alecto and P. conspicillatus to have the strongest positive correlation with equine case locations, suggesting these species are more likely a source of infection of Hendra virus for horses than P. poliocephalus or P. scapulatus. The density of horses, climate variables and vegetation variables were not found to be a significant risk factors, but the residuals from the GWR suggest that additional unidentified risk factors exist at the property level. Further investigations and comparisons between case and control properties are needed to identify these local risk factors.

Applying Bayesian network modelling to understand the links between on-farm biosecurity practice during the 2007 equine influenza outbreak and horse managers' perceptions of a subsequent outbreak

Australia experienced its first ever outbreak of equine influenza in August 2007. Horses on 9359 premises were infected over a period of 5 months before the disease was successfully eradicated through the combination of horse movement controls, on-farm biosecurity and vaccination. In a previous premises-level case-control study of the 2007 equine influenza outbreak in Australia, the protective effect of several variables representing on-farm biosecurity practices were identified. Separately, factors associated with horse managers' perceptions of the effectiveness of biosecurity measures have been identified. In this analysis we applied additive Bayesian network modelling to describe the complex web of associations linking variables representing on-farm human behaviours during the 2007 equine influenza outbreak (compliance or lack thereof with advised personal biosecurity measures) and horse managers' perceptions of the effectiveness of such measures in the event of a subsequent outbreak. Heuristic structure discovery enabled identification of a robust statistical model for 31 variables representing biosecurity practices and perceptions of the owners and managers of 148 premises. The Bayesian graphical network model we present statistically describes the associations linking horse managers' on-farm biosecurity practices during an at-risk period in the 2007 outbreak and their perceptions of whether such measures will be effective in a future outbreak. Practice of barrier infection control measures were associated with a heightened perception of preparedness, whereas horse managers that considered their on-farm biosecurity to be more stringent during the outbreak period than normal practices had a heightened perception of the effectiveness of other measures such as controlling access to the premises. Past performance in an outbreak setting may indeed be a reliable predictor of future perceptions, and should be considered when targeting infection control guidance to horse owners and managers.

Understanding the associations between on-farm biosecurity practice and equine influenza infection during the 2007 outbreak in Australia

In a previous premises-level case-control study of the 2007 equine influenza outbreak in Australia, the protective effect of several variables representing on-farm biosecurity practices was identified. However, using logistic regression it was not possible to definitively identify individual effects and associations between each of the personal biosecurity measures implemented by horse premises owners and managers in the face of the outbreak. In this study we apply Bayesian network modelling to identify the complex web of associations between these variables, horse premises infection status and other premises-level covariates. We focussed this analysis primarily on the inter-relationship between the nine variables representing on-farm personal biosecurity measures (of people residing on the premises and those visiting), and all other variables from the final logistic regression model of our previous analysis. Exact structure discovery was used to identify the globally optimal model from across the landscape of all directed acyclic graphs possible for our dataset. Bootstrapping was used to adjust the model for over-fitting. Our final Bayesian graphic network model included 18 variables linked by 23 arcs, each arc analogous to a single multivariable generalised linear model, combined in a probabilistically coherent way. Amongst the personal biosecurity measures, having a footbath in place, certain practices of visitors (hand-washing, changing clothes and shoes) in contact with the horses, and the regularity of horse handling were statistically associated with premises infection status. The results of this in-depth analysis provide new insight into the complex web of direct and indirect associations between risk factors and horse premises infection status during the first 7 weeks of the 2007 equine influenza outbreak in Australia. In future outbreaks, unnecessary contact and handling of horses should be avoided, especially by those coming from off the premises. Prior to any such contact, persons handling horses should use a footbath (if present), change their clothes and shoes, and wash their hands.

Risk map and spatial determinants of pancreas disease in the marine phase of Norwegian Atlantic salmon farming sites

Background

Outbreaks of pancreas disease (PD) greatly contribute to economic losses due to high mortality, control measures, interrupted production cycles, reduced feed conversion and flesh quality in the aquaculture industries in European salmon-producing countries. The overall objective of this study was to evaluate an effect of potential factors contributing to PD occurrence accounting for spatial congruity of neighboring infected sites, and then create quantitative risk maps for predicting PD occurrence. The study population included active Atlantic salmon farming sites located in the coastal area of 6 southern counties of Norway (where most of PD outbreaks have been reported so far) from 1 January 2009 to 31 December 2010.

Results

Using a Bayesian modeling approach, with and without spatial component, the final model included site latitude, site density, PD history, and local biomass density. Clearly, the PD infected sites were spatially clustered; however, the cluster was well explained by the covariates of the final model. Based on the final model, we produced a map presenting the predicted probability of the PD occurrence in the southern part of Norway. Subsequently, the predictive capacity of the final model was validated by comparing the predicted probabilities with the observed PD outbreaks in 2011.

Conclusions

The framework of the study could be applied for spatial studies of other infectious aquatic animal diseases.

The influence of meteorology on the spread of influenza: survival analysis of an equine influenza (A/H3N8) outbreak

The influences of relative humidity and ambient temperature on the transmission of influenza A viruses have recently been established under controlled laboratory conditions. The interplay of meteorological factors during an actual influenza epidemic is less clear, and research into the contribution of wind to epidemic spread is scarce. By applying geostatistics and survival analysis to data from a large outbreak of equine influenza (A/H3N8), we quantified the association between hazard of infection and air temperature, relative humidity, rainfall, and wind velocity, whilst controlling for premises-level covariates. The pattern of disease spread in space and time was described using extraction mapping and instantaneous hazard curves. Meteorological conditions at each premises location were estimated by kriging daily meteorological data and analysed as time-lagged time-varying predictors using generalised Cox regression. Meteorological covariates time-lagged by three days were strongly associated with hazard of influenza infection, corresponding closely with the incubation period of equine influenza. Hazard of equine influenza infection was higher when relative humidity was <60% and lowest on days when daily maximum air temperature was 20–25°C. Wind speeds >30 km hour⁻¹ from the direction of nearby infected premises were associated with increased hazard of infection. Through combining detailed influenza outbreak and meteorological data, we provide empirical evidence for the underlying environmental mechanisms that influenced the local spread of an outbreak of influenza A. Our analysis supports, and extends, the findings of studies into influenza A transmission conducted under laboratory conditions. The relationships described are of direct importance for managing disease risk during influenza outbreaks in horses, and more generally, advance our understanding of the transmission of influenza A viruses under field conditions.