An integrated geospatial analysis framework for neglected zoonotic diseases

Infectious diseases which can spread from animals to humans (and vice versa) are called zoonotic diseases. Many of these diseases are common in developing countries, particularly in poor, rural areas where humans live in close proximity to their livestock. The control of zoonotic diseases requires input from both the medical and veterinary services, but in the deprived communities most affected, access to these services is poor and the responsibility for control tends to fall between the two, resulting in underestimation of the true burden of disease, lack of understanding of the disease and minimal control efforts. For this reason, a number of zoonotic diseases occurring in poor and isolated communities have been termed "neglected zoonotic diseases" (NZDs). The NZDs cause an adverse impact on both humans and livestock (due to illness and, in some cases, death), continuing the cycle of poverty due to reduced productivity in humans and reduced outputs (e.g. meat and milk) from livestock. Therefore, improvements in our understanding of these diseases and how they can best be controlled are urgently needed.

Zoonotic disease control programmes are beginning to understand the need for integration of human and animal targeted measures, but to ensure their success a better understanding of the linkages between animal and human disease is required. This research will focus on several NZDs, including trypanosomiasis (which causes nagana in livestock and sleeping sickness in humans), cysticercosis and Rift Valley fever. Many infectious diseases (including NZDs) depend on specific environmental conditions (e.g. the amount of rainfall or the presence of particular land cover types, such as forest or swamp) for transmission and spread. The study will investigate these dependencies to provide an understanding of (a) the relationships between the amount of disease in animals and the amount of disease in humans, (b) how this relationship varies in different types of landscape (e.g. with different land cover) and (c) the reasons why these diseases occur in particular locations and not in others. These findings will provide valuable information to support NZD control, benefiting both human and livestock health and increasing livestock productivity. Due to the dependence of poor, rural communities on their livestock as a main source of income, improved control can, in the long run, help to break the cycle of poverty

The research will be conducted by Dr Nicola Wardrop, who works within the academic unit Geography and Environment at the University of Southampton, in collaboration with Professor Peter Atkinson (Geography and Environment, University of Southampton), Professor Sue Welburn (Biomedical Sciences, University of Edinburgh), Dr Eric Fèvre (Biological Science, University of Edinburgh) and Dr Archie Clements (School of Population Health, University of Queensland). The study will use state-of-the-art methods including geographical information systems (GIS), mapping, statistics and spatial modelling to assess the relationships between the transmission of NZDs and environmental factors (e.g. land cover or temperature). Disease data, with spatial references to allow mapping, will be provided by collaborating projects, which are testing human and animal populations for the presence of several NZDs. The planned research is an original and innovative extension of the geographical methods normally used to understand disease distributions. The project will integrate information on both human and animal disease into a single analysis, whereas usually they would be analysed separately. This will provide an additional layer of information, describing the relationship between disease in animals and disease in humans. These methods also provide a means to assess which other (currently unaffected) areas may be at risk of disease transmission, providing vital information for future disease control programmes.

This project aims to develop a "One Health analytical framework" to provide advances in our understanding of the spatially heterogeneous risk of neglected zoonotic disease (NZD) in humans and animals and the relationships between animal and human NZD prevalence within different landscapes (e.g. varying land cover).
Geo-referenced human and animal prevalence of several NZDs will be provided by collaborators. Remote sensor images and field visits will provide environmental covariate data (e.g. land cover, precipitation). Exploratory analysis will include: spatial mapping; descriptive analysis; cluster detection and spatial correlation analysis, to allow the selection of up to three datasets for further analysis. Preliminary, non-spatial regression will examine the influence of covariates on human and livestock NZD prevalence. This will be extended with a spatially explicit model-based geostatistics approach (to account for spatial autocorrelation) and the integrated analysis of both human and animal prevalence using joint-distribution modelling, to (a) quantify the influence of external variables on the spatial distributions of human and animal infections, (b) quantify the relationship between livestock and human prevalence of disease and (c) identify variations in relationship b in different environmental conditions. Model extrapolation will provide spatial predictions of disease prevalence in humans and livestock across the study areas. Further spatial analysis will provide estimates of the population at risk, identify priority control areas and assess the utility of sentinel screening in livestock to detect risk to humans.
The results will provide significant evidence to support the control of NZDs, and the One Health analytical framework represents a new paradigm for the analysis of zoonotic disease data, allowing integrated analysis of human and animal data in a single model and generating a substantially greater depth of understanding than individual analyses.

The proposed research is seeking an important step forward in our scientific understanding of the spatially heterogeneous risk of neglected zoonotic diseases (NZDs) in human populations. The novel integration of human and livestock disease in a holistic, spatially explicit analysis framework, while considering the influence of environmental factors, will deliver significant advances in the understanding of zoonotic disease transmission, providing valuable evidence to support control. Potential beneficiaries include academia, the public, private and third sectors within the study areas, and the general public. The end-users of this research will be communities in areas affected by (or at risk of) the NZDs which the research will focus on.

Improved knowledge of the relationships which exist between livestock disease prevalence and human disease prevalence within different landscapes, and an understanding of environmental risk factors for the occurrence of zoonotic diseases in humans, will allow the tailoring of control programmes. The delivery of a robust evidence base will support the development and implementation of appropriate and effective One Health interventions, contributing to improvements in the cost-efficiency of zoonotic disease control in the medium- to long-term. The provision of easily interpreted predicted disease distributions, along with estimates of uncertainty, will provide a valuable tool for advocacy and enable improvements in the future planning and delivery of interventions. The spatial visualisation of predicted prevalence of disease in humans and livestock will provide information vital for the allocation of resources and spatial targeting of control measures. An assessment of the potential utility of sentinel surveillance in livestock populations to assess the risk of zoonotic disease in human populations may also provide a starting point for the improvement of zoonotic disease surveillance. An understanding of the relationships that exist between livestock prevalence and human prevalence of disease is vital to judge whether sentinel surveillance may be possible. Where sentinel surveillance may be of use, the research outputs will enable an understanding of what disease detection in livestock means in terms of risk to human health. In the mid- to long-term, these findings may contribute to the implementation of more cost-effective surveillance, if disease detection is easier or cheaper in animals than humans.

In the long term, improvements to the design, targeting and implementation of NZD surveillance and control can result in several important indirect impacts. The public sector and others involved in disease control will benefit from improved cost-efficiency of control programmes, resulting in financial savings. Improved control will reduce disease prevalence, thus reducing morbidity and mortality in both humans and animals. The reduction in disease prevalence will also result in fiscal savings by the public sector, individuals and communities due to reduced demand for medical and veterinary care. An overall increase in productivity (human and livestock), may also be anticipated following improved control of NZDs, ultimately, providing a means for improved rural livelihoods.

The research will also deliver a novel, holistic framework for the integrated analysis of human and animal disease data in a spatially explicit manner. Future application of the integrated One Health analytical framework to other zoonotic diseases will allow the continued delivery of impacts, by providing significant evidence to support zoonotic disease control programmes.