Informatics progress of the Global Burden of Animal Diseases programme towards data for One Health

The Global Burden of Animal Diseases (GBADs) programme will provide data-driven evidence that policy-makers can use to evaluate options, inform decisions, and measure the success of animal health and welfare interventions. The GBADs' Informatics team is developing a transparent process for identifying, analysing, visualising and sharing data to calculate livestock disease burdens and drive models and dashboards. These data can be combined with data on other global burdens (human health, crop loss, foodborne diseases) to provide a comprehensive range of information on One Health, required to address such issues as antimicrobial resistance and climate change. The programme began by gathering open data from international organisations (which are undergoing their own digital transformations). Efforts to achieve an accurate estimate of livestock numbers revealed problems in finding, accessing and reconciling data from different sources over time. Ontologies and graph databases are being developed to bridge data silos and improve the findability and interoperability of data. Dashboards, data stories, a documentation website and a Data Governance Handbook explain GBADs data, now available through an application programming interface. Sharing data quality assessments builds trust in such data, encouraging their application to livestock and One Health issues. Animal welfare data present a particular challenge, as much of this information is held privately and discussions continue regarding which data are the most relevant. Accurate livestock numbers are an essential input for calculating biomass, which subsequently feeds into calculations of antimicrobial use and climate change. The GBADs data are also essential to at least eight of the United Nations Sustainable Development Goals.

Development and assessment of an epidemiologic dashboard for surveillance of Varroa destructor in Ontario apiaries

Varroosis (caused by the Varroa destructor mite) is a key health issue for honey bees in North America. Because these mites can exist in reservoirs of feral honey bee colonies, eradication is impossible, and instead efforts are made to maintain mites below a critical threshold. Monitoring for Varroa mites within a population is key for allocating resources and targeting interventions but surveillance can be difficult and/or expensive. This project aims to reflect on the success of data dashboards developed throughout the 2019-coronavirus pandemic and showcase how these methods can improve surveillance of Varroa mite infestations in Ontario, Canada. Dashboards provide a consistent source of information and epidemiologic metrics through data visualizations, and mobilize data otherwise bound to tables and intermittent reports. In the present work, an interactive dashboard for the surveillance of Varroa mite infestations across the province is proposed. This dashboard was developed using routine ministry inspection data to depict the spatio-temporal distribution of mites across a five-year data collection period. Through interactive figures and plots, able to be disaggregated to a specific region and time frame, this dashboard will allow for members of the beekeeping community to monitor provincial mite levels throughout the season. Seven criteria found to be common across highly actionable COVID-19 dashboards were used in a beta testing stage of development to assess the quality of the dashboard, and critically reflect on its strengths and weaknesses. Furthermore, future directions for surveillance dashboards are explored, including integration with citizen science data collection to develop a comprehensive province-wide surveillance system. The outcome of this project is a functional dashboard proof-of-concept for population-level monitoring of Varroa mites and a model for future tools designed for other species and diseases.

Global Burden of Animal Diseases: a novel approach to understanding and managing disease in livestock and aquaculture

Investments in animal health and Veterinary Services can have a measurable impact on the health of people and the environment. These investments require a baseline metric that describes the burden of animal health and welfare in order to justify and prioritise resource allocation and from which to measure the impact of interventions. This paper is part of a process of scientific enquiry in which problems are identified and solutions sought in an inclusive way. It poses the broad question: what should a system to measure the animal disease burden on society look like and what value would it add? Moreover, it aims to do this in such a way as to be accessible by a wide audience, who are encouraged to engage in this debate. Given that farmed animals, including those raised by poor smallholders, are an economic entity, this system should be based on economic principles. These poor farmers are negatively impacted by disparities in animal health technology, which can be addressed through a mixture of supply-led and demand-driven interventions, reinforcing the relevance of targeted financial support from government and non-governmental organisations. The Global Burden of Animal Diseases (GBADs) Programme will glean existing data to measure animal health losses within carefully characterised production systems. Consistent and transparent attribution of animal health losses will enable meaningful comparisons of the animal disease burden to be made between diseases, production systems and countries, and will show how it is apportioned by people's socio-economic status and gender. The GBADs Programme will produce a cloud-based knowledge engine and data portal, through which users will access burden metrics and associated visualisations, support for decisionmaking in the form of future animal health scenarios, and the outputs of wider economic modelling. The vision of GBADs, strengthening the food system for the benefit of society and the environment, is an example of One Health thinking in action.

Innovating at the human-technology interface in disasters and disease outbreaks

Disasters and disease outbreaks have long been a catalyst for innovative applications of emerging technologies. The urgent need to respond to an emergency leads to resourceful uses of the technologies at hand. However, the best and most cost-effective use of new technologies is to prevent disease and improve resilience. In this paper, the authors present a range of approaches through which both opportunities can be grasped. Global connectedness enables more data to be collected and processed in emergencies, especially with the rise of open-source data, including social media. In general, the poorest and most remote populations are most vulnerable to disaster. However, with smaller, faster, smarter, cheaper and more connected technology, reliable, efficient, and targeted response and recovery can be provided. Initially, crowdsourcing was used to find people, map affected areas, and determine resource allocation. This led to the generation of an overwhelming amount of data, and the need to extract valuable information from that data in a timely manner. As technology evolved, organisations started outsourcing many tasks, first to other people, then to machines. Since the volume of data generated outpaces human capacity, data analysis is being automated using artificial intelligence and machine learning, which furthers our abilities in predictive analytics. As we move towards prevention rather than remediation, information collection and processing must become faster and more efficient while maintaining accuracy. Moreover, these new strategies and technologies can help us to move forwards, by integrating layers of human, veterinary, public, and environmental health data for a One Health approach.

The Caribbean Animal and Plant Information Network

The Inter-American Institute for Cooperation on Agriculture (IICA) is presently implementing a four-year project in the Caribbean region to establish a regional information network on animal and plant health--the Caribbean Animal and Plant Health Information Network (CARAPHIN). CARAPHIN's immediate goal is to give the participating countries the technical capability to process and analyze the phyto- and zoo-sanitary information that they generate, share this information within the region, and use it as an effective instrument for decision making for agricultural development and trade. Sharing of information is further facilitated through the periodic regional and hemispheric meetings organized by IICA, by the Inter-American Animal Health Commission, and by the Technical Advisory Committee on Plant Protection. The Veterinary and Plant Protection Services of the region have expressed interest in the continuation of CARAPHIN as a permanent mechanism for agricultural health information management in the Caribbean. This motivated IICA to seek financing for a second phase of CARAPHIN, which will be executed from 1992 through 1996. Phase II of the project will seek to ensure the continuing application of the techniques and methodologies acquired by the agricultural health services of the different countries, and the functioning of the disease/pest information systems at the regional level. This will be achieved through technical assistance, continuing education, publications, provision of useful databases, and transfer of the project to a regional institution.

Ascariasis, respiratory diseases and production indices in selected Prince Edward Island swine herds

The levels of production, ascarid burden and respiratory disease were measured on 15 purposively selected swine herds, and the relationships between the various measures of ascarid burden were examined. On each farm 30 randomly selected pigs were weighed and rectal fecal samples were collected at approximately 11, 15, 19 and 22 weeks of age, and at slaughter. Fecal ascarid-egg counts and duration of infection were combined to calculate a composite measure of ascarid burden called "lifetime burden". At the abattoir the carcass weight and levels of anteroventral pneumonia, atrophic rhinitis, and liver lesions were recorded for each hog. The number of ascarids in the small intestines were counted. Study hogs were marketed at an average of 189 +/- 22 days. The average dressed carcass weight was 77.0 +/- 5.9 kg and the mean average daily gain was 0.519 +/- 0.071 kg/day. The percent of hogs with ascariasis varied widely among farms, no matter what measure of ascariasis was used; the percent with intestinal ascarids at slaughter ranged from 0% to 96%, the percent that shed ascarid eggs during their lifetime ranged from 0% to 100%, and the range for hogs with liver lesions ranged from 27% to 100%. Of the hogs slaughtered, 82% had milk spot lesions, 32% shed ascarid eggs during their lifetime and 35% had intestinal ascarids. The latter had an average of 12 intestinal ascarids. Anteroventral pneumonia occurred in 55% of the slaughtered hogs and 9% had atrophic rhinitis scores of five. The percent of hogs per farm with pneumonia ranged from 17% to 96%. The percent of hogs per farm with atrophic rhinitis scores of five ranged from 0% to 57%.

A critical assessment of abattoir surveillance as a screening test for swine ascariasis

The following data on ascarid burden were collected on an individual basis for 380 hogs marketed in the fall of 1987: a series of fecal ascarid-egg counts during the growing period; the level of milk spot lesions on the liver at slaughter; and the number of ascarids in the small intestines at slaughter. The presence of milk spots had a high sensitivity, very low specificity, and a high negative predictive value as a screening test for ascariasis in individual hogs. Results were consistent whether ascariasis was measured as the presence of intestinal ascarids at slaughter (sensitivity 91%, specificity 22%, negative predictive value 82%), or by a positive fecal egg count during the hog's lifetime (sensitivity 96%, specificity 24%, negative predictive value 93%). The presence of milk spots does not necessarily indicate that an ascarid infection has been established in the small intestine. The absence of milk spots, however, is a reliable indicator of the absence of an established ascarid infection, provided that the prevalence of ascariasis is equal to or less than that observed in this study. The severity of the ascarid infection in an individual hog could not be ascertained by the number of milk spot lesions on the liver.

Effect of ascariasis and respiratory diseases on growth rates in swine

Growth rates, measures of ascarid burden, and the levels of anteroventral pneumonia and atrophic rhinitis at slaughter were determined for 352 hogs born between March 8 and March 28, 1987 on 15 farms located in Prince Edward Island. Regression analyses were used to determine associations between average daily gain (ADG) and independent variables controlling for sex, farm, and litters nested within farm. The regression model accounted for 75.4% of the variation in ADG. The number of intestinal ascarids at slaughter did not affect ADG. However, the "life-time burden" (a composite measure based on fecal egg counts and duration of infection) was associated with ADG (p less than 0.05) in a quadratic manner. Although heavy ascarid burdens decreased the growth rate of swine, the magnitude of the effect was minimal. The maximum improvement one could expect from reducing the ascarid burden on heavily infected farms would be less than 1%. Severe atrophic rhinitis and the presence of anteroventral pneumonia each had a detrimental effect on ADG (p less than 0.001). The corresponding reductions in mean ADG were 7.7% and 2.8% respectively. There was significant interaction between the effects of atrophic rhinitis and anteroventral pneumonia on ADG (p less than 0.05). Hogs with both anteroventral pneumonia and severe atrophic rhinitis had a 17.6% lower ADG than hogs with neither disease. There is much greater potential for improvement in ADG through control of respiratory diseases than through control of ascariasis.