84956 rev A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 0 7 REDUCING CLIMATE-SENSITIVE DISEASE RISKS WORLD BANK REPORT NUMBER 84956-GLB APRIL 2014 A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 0 7 REDUCING CLIMATE-SENSITIVE DISEASE RISKS WORLD BANK REPORT NUMBER 84956-GLB APRIL 2014 © 2014 The International Bank for Reconstruction and Development/The World Bank 1818 H Street NW Washington, DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org Email: feedback@worldbank.org All rights reserved This volume is a product of the staff of the International Bank for Reconstruction and Development/The World Bank. The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development/The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone: 978-750-8400, fax: 978-750-4470, http://www.copyright.com/. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax: 202-522-2422, e-mail: pubrights@worldbank.org. Cover Photos: Shutterstock, LLC. CONTENTS iii CONTENTS Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii Abbreviations and Acronyms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix Foreword . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi Executive Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xiii Scope of this Report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xix Chapter 1: Knowledge for Action . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1 1.1.1 Livestock Diseases and Collaborative Health Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .2 1.1.2 Relationship between Climate Change, Climate Variability, and Livestock Disease . . . . . . . . . . .3 1.1.3 Climate, Diseases, and the Emergence of “One Health” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5 1.2 Key Climate-Sensitive Disease Threats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6 1.3 Economic and Livelihood Impacts of Climate-Sensitive Livestock Diseases . . . . . . . . . . . . . . . . . . . 10 1.3.1 Economic/Livelihood Impacts of Three Diseases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 1.3.2 Quality/Robustness of Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 1.4 Vulnerable Populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13 Chapter 2: Actionable Tools to Reduce Climate-Sensitive Disease Risks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 2.1 Surveillance Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 2.1.1 Surveillance Systems: General Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 2.1.2 Surveillance Systems: Knowledge and Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 2.2. Climate-Sensitive Disease Risk Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 2.2.1 Risk Maps: General Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 2.2.2 Utility of Risk Maps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 2.2.3 Risk Maps: Knowledge and Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 2.3 Climate-Sensitive Disease Outlooks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 2.3.1 Disease Outlooks: General Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 2.3.2 Disease Outlook: Knowledge and Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R iv CONTENTS 2.4 Early Warning Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 2.4.1 Early Warning Systems: General Principles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 2.4.2 Early Warning Systems: Knowledge and Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 2.5 C  omplementary Nature of Surveillance Systems, Risk Maps, and Disease Outlooks Within Early Warning Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Chapter 3: I  nvestments and Approaches for Establishing Early Action Climate-Sensitive Disease Risk Reduction Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 3.1 Requirements for Early Action Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 3.2 Requirement 1: Baseline Knowledge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 3.3 Requirements 2 and 3: Policy and Human Resources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 3.4 Requirement 4: Information and Communication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 3.5 Requirement 5: Physical Infrastructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39 3.6 Co-Benefits of Improving Infrastructure for Climate-Sensitive Disease Risk Reduction Tools . . . . . . . 41 Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 References���������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������������45 FIGURES Figure 1.1: G  eneral Context of Emerging Infectious Diseases [Changes in Demography (Demographic Transition, Urbanization), Livestock (Increasing Densities, Off-land Production), Wildlife (Land Pressure), and the Environment (Climate Change, Land Use Change) are the Main Components Affecting the Conditions of Emerging Infectious Diseases] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 Figure 1.2: Mechanistic Pathways by Which Climate Change May Influence Climate-Sensitive Vector-Borne Diseases . . . . . . 4 Figure 1.3: Distribution Map and Number of Outbreaks for Rift Valley Fever During an Eight-Year Period . . . . . . . . . . . . . . 7 Figure 1.4: Distribution Map and Number of Outbreaks for Bluetongue During an Eight-Year Period . . . . . . . . . . . . . . . . 8 Figure 1.5: D  istribution Map and Number of Outbreaks for Theileriosis During an Eight-Year Period—That Is, Caused by Theileria Parva (ECF) and T. annulata . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 Figure 1.6: D  istribution of BT Restriction Zones in the EU in January 2008, February 2009, and March 2012, with Each Color Indicating Regions Within Which Movements Were Permitted . . . . . . . . . . . . . . . . . . . . . . 13 Figure 2.1: Inputs, Analytical Process, and Outputs Typically Involved in Disease Risk Mapping . . . . . . . . . . . . . . . . . . . 19 Figure 2.2: Framework for Developing Early Warning Systems for Climate-Sensitive Diseases . . . . . . . . . . . . . . . . . . . . 27 Figure 2.3: Different Components of Early Warning Systems and Their Relationship . . . . . . . . . . . . . . . . . . . . . . . . . . 28 R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S CONTENTS v TABLES Table 1.1: Key Disease Characteristics for Rift Valley Fever, Bluetongue, and East Coast Fever . . . . . . . . . . . . . . . . . . . . . 10 Table 2.1: Early Warning System Data and Risk Mapping Technical Resources. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 Table 3.1: Climate-Sensitive Disease Investments. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 Note: All dollars are U.S. dollars unless otherwise noted. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R ACKNOWLEDGMENTS v ii ACKNOWLEDGMENTS This report was written by Timothy Bouley (World Bank) and Caroline Planté (World Bank). The team was under the leadership of Pai-Yei Whung and Francois Le Gall. Important contributions were made by Marius Gilbert (Université Libre de Bruxelles), Mimako Kobayashi, Anne Maryse Pierre-Louis, Norman Piccioni, Dipti Thapa, Cory Belden, Jim Cantrell, Angie Wahi (World Bank), and William Wint (University of Oxford). Communications were handled by Elisabeth Mealey. Editing was done by Linda Starke. Peer-review was performed by Ademola Braimoh, Shiyong Wang, Laurent Msellati (World Bank), Stephane de la Rocque (WHO-OIE), Judy Omumbo (independent consultant), and Joy Guillemot (WHO consultant). Special efforts were made by a team of international experts assembled to advise the development of this work, including Matthew Baylis (University of Liverpool), Diarmid Campbell-Lendrum (WHO), Cyril Caminade (University of Liverpool), Pietro Ceccato (Columbia University), Nicoline DeHaan (FAO), Stephane de la Rocque (WHO-OIE), Stephane Forman (World Bank), Marius Gilbert (Université Libre de Bruxelles), Peter Ithondeka (Department of Veterinary Services, Kenya), Bhoj Joshi (Nepal Agricultural Research Council), Steve Kemp (ILRI), Mark Nanyingi (Department of Veterinary Services, Kenya), Judy Omumbo (Independent Consultant), Madeleine Thomson (Columbia University), Joep Van Mierlo (Vétérinaires Sans Frontiéres, Belgium), Bill Wilson (US Department of Agriculture). The final product remains the responsibility of the authors. Much gratitude is owed to Juergen Voegele, Mark Cackler, and Valerie Hickey of the World Bank’s Department of Agriculture and Environmental Services and Rachel Kyte of the Climate Change Group for their leadership and guidance. Special thanks is due to the Rockefeller Foundation and the staff of the Rockefeller Bellagio Center for their support. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R A B B R E V I AT I O N S A N D A C R O N Y M S ix ABBREVIATIONS AND ACRONYMS BT Bluetongue DPCF Disaster-preparedness and contingency fund DPSIP Disaster-preparedness strategy and investment program ECF East Coast fever EID emerging infectious disease ENSO El Niño–Southern Oscillation EWS early warning system FAO Food and Agriculture Organization GCM global climate model GDP gross domestic product GIS geographic information system GLEWS Global Early Warning System for Major Animal Diseases HPAI highly pathogenic avian influenza ICT information and communication technology IFAD International Fund for Agricultural Development ILRI International Livestock Research Institute IPCC Intergovernmental Panel on Climate Change NDVI Normalized Difference Vegetation Index OIE World Organisation for Animal Health PVS Performance of Veterinary Services RVF Rift Valley fever SST sea surface temperature TTL Task Team Leader WHO World Health Organization WMO World Meteorological Organization A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R FOREWORD xi FOREWORD It is clear that climate change is increasingly affecting life on the planet. Average temperatures are rising, rainfall patterns are changing, and weather is becoming highly volatile. At the World Bank Group, we consider climate change a fundamental threat to sustainable economic development and the fight against poverty. We are concerned that without bold action now, the warming planet threatens to put prosperity out of reach of millions and to roll back decades of development. It is our hope that shedding light on the various links between climate change and development will help practitioners and governments to better respond to the challenges posed by global warming. This study focuses on livestock diseases that are “sensitive” to climate change, with a view to help practitioners reduce the risks of key climate-sensitive infectious diseases by strengthening risk management systems for disease outbreaks. The three diseases chosen for the study—Rift Valley fever, Bluetongue, and East Coast fever—spread through “vectors” such as insects and parasites, the prevalence of which fluctuates depending on key weather and climate variables such as temperature and humidity. As the symptoms of climate change continue, the frequency and extent of these diseases are expected to escalate. This research highlights the need for better understanding of the evolving interactions between the environment and emerging and re- emerging disease pathogens. It also points to the inseparable interactions between animal health and human health, which climate change appears to be reinforcing and even diversifying. In this context, the burgeoning concept and approach of “one health”—defined as “the collaborative effort of multiple disciplines—working locally, nationally, and globally—to attain optimal health for people, animals and the environment”—becomes increasingly relevant. Going forward, it is clear that partnerships are essential to implementing the interventions recommended in this report (and those beyond) and to facilitating the mobilization of information and knowledge, technical capacity, and financial resources. The Global Livestock Agenda under development at the World Bank and the Livestock Global Alliance, with health, environment, and livelihoods at its core, envisages accelerating and scaling up systematic and coordinated interventions, including those for climate-sensitive diseases. Dr. Juergen Voegele Director Agriculture and Environmental Services Department World Bank A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R EXECUTIVE SUMMARY x iii EXECUTIVE SUMMARY Disease risks to humans, animals, and plants are determined by interconnected environmental variables that affect incidence, transmission, and outbreak. Climate change affects many of the environmental variables that lead to disease. Regardless of the species involved, the impacts can ultimately affect the health, livelihood, and economic security of humans. The objective of this World Bank Economic and Sector Work is to build on scientific and operational knowledge of early action tools to help practitioners reduce the risks of key climate-sensitive infectious diseases by strengthening risk management systems for disease outbreaks. The report includes an assessment of known interventions such as the establishment of surveillance systems, the development of region- and nation-specific disease outlooks, the creation of climate-sensitive disease risk maps, and the construction and implementation of early warning advisory systems. The assessment then looks at proposed investments that can lead to the development of these tools, working toward reducing global climate-sensitive disease risk. Because of the breadth of species affected by climate-sensitive disease, it has been helpful to select a model through which the specific impact of climate change and disease can be traced. In this instance, livestock has been chosen, given its significant global presence, economic importance, and susceptibility to disease outbreak. The livestock sector plays a vital role in the economies of many develop- ing countries. Globally it accounts for 40 percent of agricultural gross domestic product (GDP). It employs 1.3 billion people and creates livelihoods for 1 billion of the world’s poor. Livestock products provide one-third of human protein intake and are a potential remedy for undernourishment. Climate-sensitive diseases pose a permanent threat to this important sector, and disease outbreaks have major economic implications—both through private and public costs of the outbreak and through the costs of the measures taken at individual, collective, and international levels to prevent or control infection and disease outbreaks. Yet despite increasing evidence, including the Intergovernmental Panel on Climate Change’s Fourth Assessment Report that linked climate variability and change to the emergence and re-emergence of infectious disease, concrete actions to address the climate impacts on disease outbreaks and livelihoods remain lagging. The World Bank’s World Development Report 2010 estimated the costs associated with climate-sensitive health impacts to be as high as 9 percent of GDP in some countries. Investments to reduce climate-associated diseases and health risks are only about 1–2 percent of overall climate sector investments. There are many climate-sensitive livestock diseases; virtually any that are dependent on vectors or are waterborne could be included on this list. To narrow the scope, three of particular economic and health importance were chosen: Rift Valley fever (RVF), Bluetongue (BT), and East Coast fever (ECF). Rift Valley fever is a vector-borne viral disease transmitted by several species of mosquitoes that have facilitated epidemics in Africa and in the Arabian Peninsula, with dramatic impact on animal and human health due to its zoonotic dimension. Bluetongue is a vector-borne viral disease transmitted by several species of Culicoides midges. The disease is endemic to many tropical climates, though it has invaded Europe in the last decades with a massive economic impact, mainly through disruption of trade. East Coast fever is a vector- borne parasitic disease transmitted by ticks that is endemic in many southeast African countries, where it has a continuous and significant economic impact. The vector-borne nature of each implies climate sensitivity. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R x iv EXECUTIVE SUMMARY Livestock diseases can be classified and ranked according to various criteria, such as overall economic impact, impact on livelihood, poten- tial to adapt and invade a new host and become zoonotic, potential to spread into new geographical areas, or a combination of these. The relative impact of each of these diseases is highly specific to the region and the capacity of agricultural and health systems to mitigate the effects. Furthermore, global data on each of these three, which can be taken as a sample of other climate-sensitive diseases, are incomplete and variable in quality and by type of disease, making it difficult to assess comprehensive economic impact. The data for health impacts on humans are even less robust, highlighting the need for individual case studies to clarify total costs. Bearing in mind these limitations, rough estimates of costs can be assessed, providing a reference point for decision makers. For example, RVF epidemics in Somalia have prevented 8.2 million small ruminants, 110,000 camels, and 57,000 cattle from being exported, correspond- ing to economic losses for the livestock industry estimated at $109 million in 1998–99 and $326 million in 2000–02. For Bluetongue, in the Netherlands alone the 2006 and 2007 epidemics had net costs of 32.4 million and 164–175 million euros, respectively. The annual cost of ECF is estimated as $88.6 million in Kenya, $2.6 million in Malawi, $133.9 million in Tanzania, and $8.8 million in Zambia. The impact of climate change on diseases is not unique to livestock. Human, plant, and other animal diseases are all affected by changing climatic conditions. Further, each affects the other and can lead to serious harm to economic and human well-being. Plant and animal diseases can lead to malnutrition and famine in humans, and many animal and human diseases can be exchanged via zoonotic (animal to human) or anthroponotic (human to animal) transmission. One Health is a recognized framework that acknowledges the systemic connectedness of human, animal, and environmental health. These considerations have long been important to health care practitioners, as humans have historically lived intimately in the environment. As cities have emerged, as technology progressed, and as allopathic medicine become the predominant medical paradigm, this inherent understanding about disease and health has been displaced by disciplinary silos. In recent decades, however, renewed interest in jointly considering these different spheres of health has occurred. Global trends in environmental change, travel, population growth, and the live- stock industry have resulted in a booming era of emerging infectious disease (EIDs). A total of 335 EIDs have been identified in humans since 1940, of which three-quarters are zoonotic, including HIV, Ebola, SARS, and avian influenza. Climate is thought to have a role in some of these emergent events; for example, recent work has suggested that variations in climate may have established environmental conditions ripe for avian influenza—a disease with catastrophic financial impacts that span sectors as diverse as livestock, tourism, trade, and health care. The additional effect of climate change on health is difficult to calculate for one species, let alone for collaborative health systems that include humans, animals, and the environment. Nevertheless, in pairing what is known about the effect of climate change on the health of one species with what is known about how the health of one species affects another, logic can help us see how climate change is undeni- ably linked to health in many spheres of life. It is not necessary to establish clear causal links between climate change and environmental change before adaptation strategies can be developed and implemented. In order to identify approaches to reducing these disease risks (and costs), it is important to understand the pressure points where climate affects disease. Climate may influence virtually all components of disease systems: the pathogen (for example, influencing the development rate or the survival outside the host or vector), the host (through the immune response or changes in host distribution), and the vectors (arthropod vector development is tightly linked to climatic parameters such as temperature and humidity). In addition, climate change and climate variability may strongly influence disease by indirect effects such as the movements of hosts resulting from floods or heat waves or climate-induced changes in land use or land cover. People in developing regions are particularly vulnerable to negative economic, social, and health impacts resulting from climate change. Human vulnerability (inclusive of health, economics, livelihoods) is affected by the vulnerability of animals’ health to climate change, but this has been the focus of few studies. Reducing climate-sensitive livestock disease risks overall can be aided by understanding how both R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S EXECUTIVE SUMMARY xv animals and humans are vulnerable to climate change so that collaborative and comprehensive systems can be developed to increase resilience. Through upstream disease prevention efforts, the health, livelihood, and economic security of downstream human populations can be protected. The tools identified and assessed in this report for upstream prevention include surveillance systems, disease outlooks, disease risk maps, and early warning systems (EWS). � Surveillance systems are key to knowing where and when a disease will occur, providing baseline data for risk models. Both active and passive surveillance are important tools that can be used to generate most-accurate disease profiles. Geospatial and informa- tion technology is increasingly important for developing accurate surveillance methodologies. � Risk maps rely heavily on the appropriate collection of disease data. They enable better prioritization of surveillance, prevention, and mitigation efforts. In many studies, risk maps have consisted of mapping the distribution of vectors; recent works enable modeling diseases themselves. � Disease outlooks aim to provide long-term projection of disease trends so that disease control and mitigation efforts can be inte- grated into long-term planning. Unfortunately, few disease outlooks are yet available for any diseases. � Early warning systems aim to provide short or midterm disease forecasting so that appropriate interventions and mitigation efforts can reduce the impact of an epidemic. Climate-based EWS have been developed for RVF in East Africa and have proved useful in predicting recent outbreaks. Preparing climate-sensitive disease risk-reduction tools requires basic levels of underlying infrastructure in a number of areas: knowledge, policy, human resources, information and communication technology, and physical building. Investment in individual project compo- nents in each of these infrastructure areas will help build the capacities of countries so that they can effectively implement and use risk management tools. Many of the actions and project components leading to the strengthening of this infrastructure are interrelated and co-dependent, necessitating investment packages that address a portfolio of needs. Further, the actions required to bolster underlying infrastructure are not necessarily specific to any one disease, and investment in project components that lead to improved infrastructure can have co-benefits for a variety of non-disease-related development needs. (See following page for detail.) Building an investment package requires a chronological deployment of activities. A three-phased approach has been recommended. Phase 1: � Knowledge: Needs assessments and baseline surveys of basic capacities of institutions, individuals, and technical and physical infrastructures Phase 2: � Information and communication technology (ICT): Climate-sensitive disease web-portals inclusive of integrated EWS information, risk maps, disease outlooks ­ � ICT: Mapping, geographic information system (GIS), and modeling software � ICT: New and/or integrated with current hydro-met information systems � Human Resources: Workforce trainings (policy makers, veterinarians, physicians, environmental scientists, communication experts, others) through short courses, workshops, and sponsored advanced degree programs on general climate-sensitive disease information as well as specialized technical aspects of the work (for example, disease diagnostics, GIS, computer programming) � Policy: Coordinated animal health–human health collaboration mechanisms through, for example, committees and cross- sectoral working groups at national/regional levels Phase 3: � ICT: EWS messages disseminated through new media of websites, mobile phones, social media A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R xvi EXECUTIVE SUMMARY INFRASTRUCTURE REQUIREMENT INVESTMENT FAMILY PROJECT COMPONENTS REQUIRING INVESTMENT Baseline Knowledge Information Product and Knowledge • Needs assessments and baseline surveys of basic capacities of institutions, individuals, Generation and technical/physical infrastructures • Climate-sensitive disease risk catalogues and impact assessments at national and regional level • Feasibility studies for risk management tools, such as EWS messaging Policy and Human Resources Institutional Strengthening and Professional • Workforce trainings (policy makers, veterinarians, physicians, environmental scientists, Capacity Building communication experts, others) through short courses, workshops, and sponsored advanced degree programs on general climate-sensitive disease research as well as specialized technical aspects of the work (such as disease diagnostics, disease risk mapping (GIS and spatiotemporal modeling), computer programming) • Environment, disease, and ICT workforce recruitment • Coordinated animal–human health collaboration committees and cross-sectoral working groups at national/regional levels • Early warning protocols for specific climate-sensitive diseases Community Capacity Building • Climate-sensitive disease and ICT user trainings at local and subnational levels • Community support groups and knowledge exchanges Information and Communication Information Dissemination • Climate-sensitive disease publications disseminated to professional and lay audiences • Climate-sensitive disease and EWS messages to be disseminated through traditional media resources: print, television, radio, community theatre • EWS messages disseminated through new media: websites, mobile phones, social media ICT Capacity Building • Digital climate-sensitive disease libraries at regional/national level • Climate-sensitive disease web-portals inclusive of integrated EWS information, risk maps, disease outlooks • Mapping, GIS, and modeling software • New and/or integrated with current hydro-met information systems • Innovative data collection approaches Physical Building and Construction • New or retrofitting of current facilities to create coordinated animal-human health–environmental data collection and collaboration centers at national/regional levels; to include meeting facilities, high speed Internet, resource libraries, and computers equipped with mapping, modeling, climate, and disease monitoring software • Rapid diagnostic laboratories equipped to process climate-sensitive diseases • ICT networks R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S The real problems are setting up the delivery systems that can protect people not only from the diseases of today, but from the diseases of tomorrow, and there’s enough money out there in the world that we can begin moving in that direction. That’s how I would like to see the World Bank engage. —Jim Yong Kim at Brookings Institution, July 19, 2012 President The World Bank SCOPE OF THIS REPORT x ix SCOPE OF THIS REPORT The risk of disease to humans, animals, and plants is determined by interconnected factors of political, structural, organizational and technical nature. Environmental variables affect incidence, transmission, and spread. In this Economic and Sector Work, we focus only on the additional risk posed by climate change, which affects many of the environmental variables that lead to disease. We further narrow the scope by focusing on livestock. Many of the lessons learned from this exercise are transferrable to other diseases, including those that also affect humans. Through upstream disease prevention efforts, the health, livelihood, and economic security of downstream human populations can be protected. Audience This report was written as guidance for investment and project implementation to reduce risks from climate-sensitive disease. Internal World Bank audiences that stand to benefit the most are Task Team Leaders (TTLs) in countries and regions that are interested in developing coordinated programs incorporating agricultural, health, and environmental activities. Given the cross-cutting themes identified, TTLs with interests in any of these disciplines singularly may also benefit from learning of the co-benefits of inter-sectoral investment and action. Audiences external to the World Bank who may be interested in this work include governments, agencies, and nongovernmental organiza- tions working at any of the intersections of agriculture, health, and environment. Guide for Readers This report does not need to be read cover to cover to be useful. Sections have been highlighted below in attempt to draw attention to certain audiences and corresponding “most relevant” sections. Potentially Most Relevant to Researchers General information on climate and disease or One Health: pp. 1–5 Climate-sensitive livestock diseases: pp. 5–10 Economic impacts of climate-sensitive diseases: pp. 10–13 Descriptions of tools for reducing climate-sensitive disease risks Surveillance systems: pp. 15–18 Risk maps: pp. 18–23 Disease outlooks: pp. 23–24 Early warning systems: pp. 25–29 Potentially Most Relevant to Operational Teams Approaches to using tools and making investments for reducing climate-sensitive disease risks: pp. 31–32 Building knowledge infrastructure: pp. 32–33 Building policy and human resource infrastructure: pp. 33–37 Building ICT infrastructure: pp. 37–39 Building physical infrastructure: pp. 39–40 A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R xx SCOPE OF THIS REPORT Zoonotic Pathogens from Wildlife (Jones et al. 2008) Adapted from and reprinted by permission from Macmillan Publishers Ltd: Nature. Jones, K. E., N. G. Patel, M. A. Levy, A. Storeygard, D. Balk, J. L. Gittleman, and P. Daszak. 2008. “Global Trends in Emerging Infectious Diseases.” 451 (7181): 990–93. Zoonotic Pathogens from Non-Wildlife (Jones et al. 2008) Adapted from and reprinted by permission from Macmillan Publishers Ltd: Nature. Jones, K. E., N. G. Patel, M. A. Levy, A. Storeygard, D. Balk, J. L. Gittleman, and P. Daszak. 2008. “Global Trends in Emerging Infectious Diseases.” 451 (7181): 990–93. R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S SCOPE OF THIS REPORT xxi Vector-Borne Pathogens (Jones et al. 2008) Adapted from and reprinted by permission from Macmillan Publishers Ltd: Nature. Jones, K. E., N. G. Patel, M. A. Levy, A. Storeygard, D. Balk, J. L. Gittleman, and P. Daszak. 2008. “Global Trends in Emerging Infectious Diseases.” 451 (7181): 990–93. MAPS: Global distribution of relative risk of emerging infectious disease events. Derived for EID events caused by zoonotic pathogens from wildlife, zoonotic pathogens from non-wildlife, and vector-borne pathogens. The relative risk is mapped on a linear scale from green (lower values) to red (higher values). A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R CHAPTER 1 — KNOWLEDGE FOR AC TION 1 Chapter 1 KNOWLEDGE FOR ACTION power, and bestows status, collectively making it an important lever This chapter includes state-of-the-art knowledge of key of economic and social development (Randolph et al. 2007). livestock diseases, their relationship to climate change, and the economic and social impacts that these diseases have on Infectious livestock diseases pose a serious threat, and disease out- human health, well-being, and livelihood. breaks have major socioeconomic impacts through losses incurred by outbreaks and through costly measures taken at individual, national, and international levels for prevention and control (Otte, Key messages: Nugent, and McLeod 2004). For zoonotic diseases (those that can • Livestock are fundamental to the health and livelihood be transmitted from animals to humans), the costs are even higher of many in the developing world, accounting for 40 per- cent of agricultural gross domestic product, employing because they include the additional impact on human health 1.3 billion people, and providing one-third of human (Rushton, Heffernan, and Pilling 2002). A recent study carried out by protein intake. the International Livestock Research Institute (ILRI) noted that the • Climate-sensitive diseases have significant impact on the greatest burden of zoonotic disease falls on the poorest livestock livestock sector in poor countries, having already led to keepers, with 2.3 billion human illness and 1.7 million human deaths hundreds of millions of dollars in losses in recent decades. • Infectious disease stands to be particularly affected per year. Unsurprisingly, these burdens are felt in countries with by climate change; many viral, bacterial, and parasitic large pastoralist populations, with Ethiopia, Nigeria, Tanzania, and infections depend on climate variables like temperature, India having some of the highest burdens (Grace et al. 2012). humidity, and precipitation. • There are many climate-sensitive livestock diseases; The World Development Report 2010 estimated the costs associated this report highlights three of particular economic and with climate-sensitive health impacts (in humans and animals) to health importance: Rift Valley fever, Bluetongue, and East Coast fever. be as high as 9 percent of gross domestic product (GDP) in certain • All regions are subject to climate change/variability and countries (World Bank 2010b). Despite this, investments to mitigate are at risk of emerging new climate-sensitive diseases; the climate-induced diseases and health risks are only about 1–2 per- three disease examples provided in this document pro- cent of overall climate sector investment. vide important examples that are relevant to all regions. Several examples illustrate these aggregated numbers. Rift Valley 1.1 BACKGROUND fever, a mosquito-borne virus that is responsible for significant mor- Livestock is a key economic component of the agriculture sector, bidity and mortality in humans and animals, had estimated trade- accounting for 40 percent of global agricultural gross domestic related economic losses as high as $60 million between 2006 and product, employing 1.3 billion people, and providing one-third of 2007 in East Africa alone (Little 2009). Tick-borne diseases, such as human protein intake (Steinfeld et al. 2006). Livestock plays an espe- theileriosis (in animals), have been estimated to cost $384.3 million cially crucial role in developing countries. In addition to protein and annually in India and $54.4 million in Kenya (Jongejan and Uilenberg income, it provides draught power, transport, fertilizer, holds cultural 2004). The cost of inaction against livestock trypanosomiasis in A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 2 CHAPTER 1 — KNOWLEDGE FOR AC TION Nigeria is estimated to be 10 billion Nigerian Naira/year (~$60 mil- FIGURE 1.1: General Context of Emerging Infectious Diseases lion) (Fadiga, Jost, and Ihedioha 2011). In the developed world, the [Changes in Demography (Demographic Transition, Urbanization), risks are also real, as evinced in the 2006 Bluetongue epidemic in the Livestock (Increasing Densities, Off-land Production), Wildlife (Land Pressure), and the Environment (Climate Change, Land Use Netherlands, which accounted for €32.4 million in country losses Change) are the Main Components Affecting the Conditions of and €164–175 million in losses upon its European spread (Velthuis Emerging Infectious Diseases] et al. 2010). People Global demand for animal-based protein is predicted to grow and to lead to increasing livestock populations, with the cattle population increasing from 1.5 billion to 2.6 billion and sheep and goats increasing from 1.7 billion to 2.7 billion between Pathogens 2000 and 2050 (FAO 2009). The human population is expected & vectors to reach 9 billion by 2050, and the average global temperature is on track to increase by several degrees Celsius by the end of Livestock Wildlife century (IPCC 2007). Together, these projections portend prob- lems for the incidence and transmission of diseases that thrive in overpopulated, warm environments. Environment 1.1.1 Livestock Diseases and Collaborative Health Systems Along with the most dramatic increase in human population in The genetic pool and origin of most emerging infectious diseases history, the twentieth century saw a profoundly accelerated rate of can be identified in wildlife (Jones et al. 2008; Cleaveland, Laurenson, urbanization (from 13 percent living in cities in 1900 to 49 percent in and Taylor 2001). The domestication of animals, however, has led to 2005), providing new habitats for vectors, concentrated populations new evolutionary opportunities (Diamond 2002; Wolfe, Dunavan, of human hosts, and a substantial increase in demand for livestock and Diamond 2007). During the “first epidemiological transition,” products. In parallel, and following the green revolution in crop humans began congregating in more sedentary agricultural soci- production, a livestock revolution fundamentally changed the way eties and establishing larger communities in cities, providing op- livestock were raised and traded (Delgado 1999; Steinfeld et al. 2006). portunities for pathogens such as malaria, smallpox, measles, and The shift can be characterized by a change in practices from local tuberculosis to emerge and spread (Harper and Armelagos 2010). multi-purpose activity to market-oriented production and integrat- The “second epidemiological transition” coincided with the indus- ed processes, a decreasing importance of ruminants compared to trial revolution, when improved nutrition and more-effective public monogastric species (pigs, poultry), more large-scale industrial pro- health measures resulted in a decline in early mortality from infec- duction closer to urban consumption centers, an increase in the use tious disease. Yet as populations aged, people began to experience of cereal-based feed, and an increase in the volume of trade of live a concomitant rise in chronic disease such as heart failure, cancer, animals and animal products. Such changes in human and livestock and diabetes (Barrett et al. 1998), effectively substituting one cause numbers and distribution have been so significant that people and of death for another. their livestock represent today more than 95 percent of the terrestrial vertebrate biomass (estimate based on Smil 2002 and on FAO 2009). We are now in the midst of the “third epidemiological transition,” which is characterized by the emergence or re-emergence of patho- The epidemiological connectivity of the human and livestock popu- gens in a context of fast demographic changes, globalization of pro- lations has also been considerably expanded through the globaliza- duction and trade, and changes in land use and climate (figure 1.1). tion of trade and travel. While epidemiological theory predicts that R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S CHAPTER 1 — KNOWLEDGE FOR AC TION 3 an increase in the number and connectivity of hosts should result in indirect effects such as movements of hosts resulting from floods higher disease persistence and spread, the effects of those changes or heat waves or climate-induced changes in land use or land cover. in human and animal populations have been largely tempered by Our current capacity to predict the actual impact of climate change the rapid parallel developments in human and animal health—that on livestock diseases is somewhat limited, but it can be improved in is, with the advent of vaccination, medication, and increased in- the future, particularly with some of the methods detailed later in vestments in disease prevention and control. Yet as the history of this report (Heffernan, Salman, and York 2012). emerging diseases has demonstrated, pathogens adapt to these Climate change occurs at a global scale, yet it impacts regional and new patterns of human and animal populations (Antia et al. 2003) local environmental systems—and subsequently affects regional and challenge our ability to control them (Daszak, Cunningham, disease profiles. The focus of this report is the additionality of this and Hyatt 2000). The direct costs of four emerging disease out- climate-change-related burden, independent of other environ- breaks over the last decade are estimated at over $20 billion, and mental pressures. Determining the exact degree to which climate indirect costs to economies are thought to be greater than $200 change affects diseases is challenging in light of the multitude of billion (World Bank 2010b). If indirect losses to other parts of the factors that determine disease transmission, such as species inter- animal production chain, trade, and tourism are included, these actions, vegetation, land degradation, food resources, population, costs are considerably higher. and the baseline health of species. And yet it is possible to make Alongside the changes in host abundance and connectiv- some inferences about how climate change will affect diseases in ity, emerging and re-emerging diseases affecting people and any given region with an understanding of how certain diseases are livestock are strongly influenced by environmental changes in sensitive to the environment. land use (Patz et al. 2004) and climate (Epstein 2001). The World Health Organization (WHO) estimates that one-quarter of the It is essential that the global community consider climate global burden of diseases in humans, disproportionately felt in the change’s impact on health—both animal and human—given developing world, is due to environmental change (Prüss-Ustün its potential to undermine global economic systems and livelihoods, particularly of the least-resilient populations. and Corvalán 2007). Similar estimates have not been made for livestock, although given the shared environmental pressures on In most disease systems, climate change is occurring concurrently them, it can be inferred that a corresponding statement could be with anthropogenic and natural drivers of change, making it diffi- made for animal diseases. cult to disentangle respective impacts. Rapid transformation in land use, increase in trade and movements of live animals, increase in 1.1.2 Relationship between Climate Change, Climate trade of goods that may harbor breeding vectors, changes in the Variability, and Livestock Disease distribution and abundance of livestock, or changes in the genetic A number of studies have explored the potential effect of climate composition of hosts, for example, may all affect disease systems. change on infectious diseases in animals (de la Rocque 2008; Baylis and Githeko 2006; Heffernan, Salman, and York 2012). Collectively, The relationships between climate change, climate variability, and dis- these reviews highlight that climate change and climate vari- ease are disease-specific. In 2008, the World Organisation for Animal ability may influence virtually all components of disease systems Health (OIE) Scientific and Technical Review published a review of the (figure  1.2): the pathogen (for instance, influencing the develop- impact of climate change on both the epidemiology and the control ment rate or survival outside the host or vector), the host (through of animal diseases (de la Rocque 2008), and the state of knowledge the immune response or changes in host distribution), and the on the effect of climate change was reviewed for several diseases vectors (arthropod vector development is tightly linked to climatic and disease groups, including Rift Valley fever, Bluetongue, avian parameters such as temperature and humidity). In addition, climate influenza, tick-borne diseases, mosquito-borne diseases, leishmani- change and climate variability may strongly influence diseases by asis, and helminthiasis. The review demonstrated a contrasting range A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 4 CHAPTER 1 — KNOWLEDGE FOR AC TION FIGURE 1.2: Mechanistic Pathways by Which Climate Change May Influence Climate- Sensitive Vector-Borne Diseases Climate change Affects Disease mitigation Regional climate Meteorological variables For example, population density, sanitation infrastructure, land use surveillance Humidity, temperature, precipitation Affects Non-climate change pressures Vector-borne disease Pathogen Vector Disease surveillance Replication, Transmission Distribution, virulence exposure reproduction, Vector and pathogen control maturation, feeding behavior, Disease outlooks longevity Vaccination Host Animal and human Impacts Vulnerable populations Risk mapping Economic Health Livelihood of potential effects in different diseases systems. Strong evidence sup- on Climate Change (IPCC) predictions for future temperature and ported the impact of climate change on some diseases, for example rainfall are notoriously heterogeneously distributed throughout the on the northward expansion of Bluetongue (Purse et al. 2008). globe (IPCC 2007), and highlight not only changes in the averages but also the extremes. The impact of such changes on different vec- The potential impact of climate on diseases transmitted by tors is therefore likely to vary substantially from place to place, with arthropod vectors is strongly scale-dependent in space and time. higher vector populations in some parts of the world (regions that Temperature affects arthropod vector development at embryonic, have recently become part of the niches) and in lower populations larval, and pupal stages, it influences adult feeding behavior, and in others (regions that are no longer suitable). In addition, a general it affects adult life spans. Similarly, aquatic or moist environments rise in annual mean temperature may have very different effects on are often needed for breeding stages so that high precipitation can vector populations according to the season: increased winter tem- create more reservoirs and thus amplify the number of breeding perature may have a positive effect through a lower winter mortal- sites. All vector species distributions can therefore be defined by ity, whereas increased summer temperature may have the opposite their temperature and moisture ecological niche, which are defined effect through an increased mortality in adults. Both spatial and by both a lower and an upper bound. The Intergovernmental Panel R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S CHAPTER 1 — KNOWLEDGE FOR AC TION 5 temporal heterogeneity of the predicted changes, combined with for avian influenza—a disease with catastrophic financial impacts the inherent uncertainties in climate projections, make predictions that span sectors as diverse as livestock, tourism, and health care difficult, even for a single disease system. (Shaman and Lipsitch 2012). While there are clear difficulties in assessing the overall impact of cli- The additional effect of climate change on health is difficult to mate change and variability on livestock diseases, careful assessment calculate for one species, let alone for collaborative health systems and prediction in some regions and disease systems remain possible, that include humans, animals, and the environment. Nevertheless, as demonstrated by the numerous studies that have used statistical in pairing what is known about the effect of climate change on the modeling to forecast the future distribution of species or disease health of one species with what is known about how the health (Rogers, Hay, and Packer 1996; McDermott et al. 2002; Purse et al. of one species affects another, logic can help us see how climate 2008). Even if other mechanistic causes are implicated, addressing change is undeniably linked to health in many spheres of life, and mitigating the potential effects of climate change and climate regardless of our incomplete understanding. It is not necessary to variability on livestock disease has much benefit, particularly in the establish clear causal links between climate change and environ- developing world, where humans and livestock live so close together. mental change before adaptation strategies can be developed and implemented (Black and Nunn 2009). 1.1.3 Climate, Diseases, and the Emergence of “One Health” The impact of climate change on diseases is not unique to livestock. Recent One Health Actions Human, plant, and other animal diseases are all affected by chang- In recent years the international community has taken increas- ing climatic conditions. Further, each affects the other and can lead ing notice of both the threat that climate change poses to to serious harm to economic and human well-being. Plant and disease and the importance of collaborative health among hu- animal diseases can lead to malnutrition and famine in humans, mans, animals, and the environment. and animals and humans can be affected by common pathogenic In April 2011, the African Union Commissioner of Rural Economy agents (zoonotic diseases). and Agriculture, jointly with the UN Economic Commission for Africa, WHO’s Regional Office for Africa, Columbia University One Health is a recognized framework that acknowledges the sys- International Research Institute, and Ethiopia Climate and temic connectedness among human, animal, and environmental Health Working Group, committed to take actions to build a climate-resilient healthy community through integrating health. These considerations have long been important to health climate-health risk management. In October 2011, a confer- care practitioners, as humans have historically lived intimately in the ence was held among the International Institute for Strategic environment. As cities have emerged, technology progressed, and Studies and members of the medical community, exploring allopathic medicine become the predominant medical paradigm, the linkages between climate change, security, and health this inherent understanding about disease and health has been (inclusive of disease). Recent reports by the Natural Resource Defense Council, the Union of Concerned Scientists, and displaced by disciplinary silos. In recent decades, however, renewed Accenture have tallied the health costs due to climate change interest in jointly considering these different spheres of health has and found billions of dollars worth of impact. In 2009, the occurred. Global trends in environmental change, travel, population World Organisation for Animal Health released a list of diseases growth, and the livestock industry have resulted in a booming era of that are at risk of being affected by climate change. In 2008, the American Veterinary Medical Association issued a  statement emerging infectious disease. A total of 335 EIDs have been identified drawing attention to the impact of climate change on animal in humans since 1940, of which three-quarters are zoonotic, includ- health. Over the past decade, the World Health Organization ing HIV, Ebola, SARS, and avian influenza (Jones et al. 2008; Taylor has published a number of articles and reports on how climate et al. 2001). Climate is thought to have a role in some of these emer- change will affect health and, in particular, disease—and WHO is also currently partnering with the UN Development gent events; for example, recent work has suggested that variations (Continued) in climate may have established environmental conditions ripe A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 6 CHAPTER 1 — KNOWLEDGE FOR AC TION Rift Valley fever, bluetongue, and East Coast fever were chosen Recent One Health Actions (Continued) for this report given the breadth of successes and challenges Programme and the World Meteorological Organization (WMO) each embodies and the opportunity to derive a spectrum of les- to build paired climate and disease surveillance systems. sons learned. RVF, for example, is present in both humans and Global One Health conferences were held in Switzerland and animals, has affected large parts of Africa, and has been piloted Thailand in 2012 and 2013 to discuss the threats and oppor- in several early warning system models. Bluetongue is found only tunities of One Health actions. In 2011, the Global Initiative for Food System Leadership at the University of Minnesota in livestock and is currently a significant problem for European convened a conference to assess the global implementation countries, yet it has also been the subject of early warning sys- of One Health. In 2010 and 2012, the World Bank published tems and monitoring and surveillance programs. ECF is relatively two volumes of One Health work: People, Pathogens, and Our less tracked in early warning systems, but it has had significant Planet (World Bank 2010a). In 2009, a One Health Commission was endorsed by a number of UN organizations (the Food and economic impact in East Africa and is transmitted by ticks, as Agriculture Organization [FAO] and WHO), the OIE, and the opposed to either RVF or BT, which are transmitted by mosquitoes Centers for Disease Control and Prevention. And in the past and midges, respectively. decade, dozens of nonprofits, professional organizations, and universities have initiated One Health programs. Rift Valley fever is a viral zoonosis transmitted by mosquitoes that primarily affects animals, though sometimes it infects humans. The disease has had significant impact on both animal and human 1.2 KEY CLIMATE-SENSITIVE DISEASE THREATS health in East Africa and, recently, the Middle East. Outbreaks have occurred in Kenya (1968, 1978–79, 1997–98), Sudan (1973, 1976), Key Messages: Somalia (1997–98), Tanzania (1977, 1987, 1997), Zambia (1973–74, • Climate-sensitive livestock diseases can be ranked 1978, 1985), Zimbabwe (1955, 1957, 1969–70, 1978), Mozambique according to various criteria, such as economic impact, (1969), South Africa (1974–76, 1981, 1996), Namibia (1955), and for epidemic potential, and zoonotic or public health the first time off the African continent in 1998–2000 in Saudi Arabia dimension. • Rift Valley fever is a vector-borne viral disease transmit- and Yemen (figure 1.3). These outbreaks have caused widespread ted by several species of mosquitoes that have facili- morbidity and mortality and resulted in hundreds of millions of tated epidemics in Africa and in the Arabian Peninsula, dollars in agricultural, trade, health care, and tourism losses (Rich with dramatic impact on animal and human health and Wanyoike 2010). The 1997–98 occurrence was the largest docu- due to its zoonotic dimension. The disease has been shown to be sensitive to climate. mented outbreak ever in the Horn of Africa, involving five countries, • Bluetongue is a vector-borne viral disease transmitted the loss of ~100,000 domestic animals, and ~90,000 human infec- by several species of Culicoides midges. The disease tions (Woods et al. 2002). is endemic to many tropical climates, though it has invaded Europe in the last decades with a massive The disease is transmitted by a broad range of mosquitoes, though economic impact, mainly through disruption of trade. certain Aedes species can act as reservoirs during inter-epidemic Recent evidence suggests that the northward shift of years. Increased precipitation in dry areas leads to explosive hatch- the disease could have been caused by climate change. • East Coast fever is a vector-borne parasitic disease ing of RVF-harboring mosquito eggs, which when combined with transmitted by ticks that is endemic in many southeast immune-naive animal populations can lead to outbreaks. Juveniles African countries, where it has a continuous and are most at risk, with mortality rates ranging from 20 percent to 100 significant economic impact. The disease has received percent, depending on animal (OIE 2009). comparatively less attention than RVF and BT, though its transmission via vector species suggests it may be The disease affects animals and humans. In animals, it primarily sensitive to climate. affects sheep, cattle, goats, camels, and wild ruminants, resulting in R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S CHAPTER 1 — KNOWLEDGE FOR AC TION 7 FIGURE 1.3: Distribution Map and Number of Outbreaks for Rift Valley fever During an Eight-Year Period (OIE 2014). high rates of abortion and neonatal mortality (LaBeaud, Kazura, and persist for four to seven days. A small minority can experience eye King 2010). The majority of human infections result from direct or lesions, meningo-encephalitis, or hemorrhagic fever (Davies 2010). indirect contact with the blood and organs of infected animals. Transmission often occurs during slaughtering or butchering, There are no specific treatments once infection has occurred in assisting in animal births, providing veterinary care, or disposing of animals or humans; prevention and control are therefore the only carcasses and fetuses. As a result, certain occupational groups such measures for avoiding disease transmission and outbreak. There as herders, farmers, butchers, and veterinarians are at higher risk. are multiple vaccine options for animals (attenuated/inactivated), The virus infects humans either through direct body fluid contact although none are currently licensed for humans. Sanitary pro- or via aerosols produced during slaughter. In some cases, humans phylaxis is also recommended, including wearing protection at can also be infected by mosquitoes and blood-feeding flies. So far slaughterhouses and during veterinary procedures, draining stand- no human-to-human transmission has been observed (WHO 2010). ing water and providing vector control in mosquito-prone areas, The total case fatality rate varies widely by epidemic, though it is less and running community awareness campaigns that highlight the than 1 percent overall. Most sufferers typically experience a mild unsafe consumption of raw animal tissues and protection against form of the disease that is characterized by flu-like symptoms that mosquitoes (OIE 2009; WHO 2010). A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 8 CHAPTER 1 — KNOWLEDGE FOR AC TION FIGURE 1.4: Distribution Map and Number of Outbreaks for Bluetongue During an Eight-Year Period (OIE 2014). A Closer Look: The Epidemiologic Cycling Aedes has been suggested by field data (Linthicum et al. 1985) of Rift Valley Fever and provides a mechanistic pathway for the reservoir of the virus during inter-epidemic periods. Culex females, in contrast, Mosquitoes of the Aedes and Culex genera are the main lay their eggs on the surface of water and need permanent vectors of RVF, but they have been shown to have a different water to develop, as the eggs cannot survive desiccation ecology and potential role in the persistence and spread of the and are abundant in irrigated areas. No evidence of vertical disease. Aedes females typically lay their eggs in the mud of transmission of RVF in Culex has been found. Some particular small water bodies. Even when they become desiccated, the sequence of floods inundating small ponds and water bodies eggs can survive several years and will hatch when they are can thus affect the epidemiologic cycle of RVF in Aedes, and exposed again to a short period of flood. If no further cycles result in the subsequent infection of nearby livestock (Wilson of desiccation and flooding occur, the population remains low 1994; Chevalier et al. 2004). If the natural RVF immunity in sus- because the eggs need a period of desiccation for embryogen- ceptible livestock is low, the disease spreads rapidly and can esis. As a result, regions characterized by a succession of dry become amplified through horizontal transmission by Culex and wet periods provide the most suitable environment for mosquitoes. egg survival and development. Vertical transmission of RVF in R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S CHAPTER 1 — KNOWLEDGE FOR AC TION 9 Bluetongue is a vector-borne virus that affects ruminants, primar- The disease is characterized by fever, oral and nasal hemorrhage, ily sheep, occasionally goats and deer, and cattle (Sperlova and excessive salivation, and nasal discharge. In some cases, the tongue Zendulkova 2011). It is transmitted by various Culicoides species will appear cyanotic and swollen, hence the name. Among domes- of biting midge (Purse et al. 2005) and can result in severe clinical tic animals, disease is most severe in sheep, with mortality rates up symptoms, sometimes leading to death (Wilson and Mellor 2009). to 70 percent in most susceptible breeds, and in some wild spe- The OIE has had it historically listed as a “notifiable disease” since cies such as whitetail deer and pronghorn antelope, with mortality the 1960s due to the high associated economic costs. The virus is rates up to 90 percent. Cattle usually do not express clinical signs found on all continents (figure 1.4), although different serotypes except with the BT virus strain 8 recently found in Europe. BT does result in markedly different impacts. In recent years it has achieved not establish persistent infections in animals, though the disease is significant visibility after emerging in Europe in 2006 (Saegerman, maintained through infected cattle and wild ruminant reservoirs. Berkvens, and Mellor 2008). In the Netherlands alone, the costs as- The disease is then transmitted by the midge vector, which is itself sociated with this outbreak amounted to hundreds of millions of positively sensitive to certain climatic factors, such as high rainfall, euros (Velthuis et al. 2010). warm temperatures, and high humidity. FIGURE 1.5: Distribution Map and Number of Outbreaks for Theileriosis During an Eight-Year Period—That Is, Caused by Theileria parva (ECF) and T. annulata (OIE 2014). A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 10 CHAPTER 1 — KNOWLEDGE FOR AC TION There is no efficient treatment for BT other than to engage in pro- TABLE 1.1: Key Disease Characteristics for Rift Valley Fever, phylactic measures. In disease-free areas, animal movement control Bluetongue, and East Coast fever and quarantine must be enforced. In infected areas, vector control is RVF BT ECF Current Distribution Africa Latitudinal Southeast Africa recommended. Vaccines are also currently available, although they Primary Regional African, Arabian Europe Southeast Africa sometimes require serotype specificity in order to be fully effective. Impact peninsula The disease is not zoonotic and cannot infect humans (OIE 2009). Zoonotic Yes No No Vector Mosquito—various Biting midges, Ixodid ticks East Coast fever is a cattle disease endemic to regions from south- species, for example, especially various Aedes and Culex Culicoides species ern Sudan to South Africa and west, to eastern Democratic Republic Species Affected Primarily sheep, Primarily sheep, African buffalo, of Congo (figure 1.5). It is caused by the parasite Theileria parva, one cattle, goats, and occasionally goats cattle wild ruminants; and deer, and cattle of six species of Theileria that infects cattle. Human theileriosis is also humans caused by genus Theileria, though it is of a different species: microti Treatment Prophylaxis only Prophylaxis only Prophylaxis, acaricide (OIE 2009). A related disease, tropical (or Mediterranean) theileriosis is caused by T. annulata and is endemic in North Africa, southern 1.3. ECONOMIC AND LIVELIHOOD IMPACTS OF Europe, parts of eastern Europe, the Indian subcontinent, China, and CLIMATE-SENSITIVE LIVESTOCK DISEASES the Middle East. Annual costs associated with the livestock disease are in the hundreds of millions of dollars. Key Messages: Parasites are transmitted by several species of Ixodid ticks that have • The individual and collective economic impact of historically been hosted by African buffalo; only the relatively recent RVF, BT, and ECF has not been estimated globally, but examples are available for some countries that illustrate introduction of cattle to the region has resulted in the new parasitic the magnitude of the impact. target. Infected ticks can remain in grazing lands for up to two years, • For RVF, in Somalia the epidemic prevented 8.2 million depending on climate. Warmer temperatures speed up parasite small ruminants, 110,000 camels, and 57,000 cattle from maturation and subsequent diminished tick-attachment to infection being exported, corresponding to economic losses for the livestock industry estimated at $109 million in time. The ticks can be found from sea level to over 2,500 m in any area 1998–99 and at $326 million in 2000–02. where the annual rainfall exceeds 500 mm. Without tick presence, the • For BT, in Netherlands alone the 2006 and 2007 epidem- parasite is unable to complete its life cycle and disappears (OIE 2009). ics had a net cost of 32.4 million and 164–175 million euros, respectively. ECF affects cattle differentially, with exotic species mortality ap- • The annual cost of ECF was estimated at $88.6 million in proaching 100 percent in some areas. Indigenous Zebu cattle tend Kenya, $2.6 million in Malawi, $133.9 million in Tanzania, be less severely affected, although they nearly always show some and $8.8 million in Zambia. morbidity (OIE 2009). • Robust assessment of economic and livelihood impact in many countries is impaired by difficulties in assessing indirect impacts and by the lack of epidemiological data. Treatment can be both preventative and therapeutic (see table 1.1.) Acaricide pour-ons are frequently used to kill the ticks, although they are expensive, can be environmentally detrimental, and can lead to 1.3.1 Economic/Livelihood Impacts of Three Diseases resistance in the targeted species. A number of vaccines are also RVF is enzootic in most sub-Saharan African countries (Davies 2010) available in various forms to prevent parasitism. Chemotherapeutic and has been recorded (Clements et al. 2007a), either through agents, such as buparvaquone, are used to treat cattle once in- diagnosis during epidemic or through sero-surveillance surveys fected, but they do not always completely eradicate the infections. (Gonzalez et al. 1992; Mariner, Morrill, and Ksiazek 1995), even in Best-practice methods tend to use a combination of tick control, countries where there have never been any significant outbreaks vaccination, and chemotherapy (OIE 2009). (such as Niger, Burkina Faso, and Gabon). The virus has also spread R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S CHAPTER 1 — KNOWLEDGE FOR AC TION 11 to Egypt (Meegan, Hoogstraal, and Moussa 1979; Arthur et al. 1993), 2002). However, the disease has had a much more serious impact Madagascar (Andriamandiby et al. 2010), and the Arabian Peninsula when occurring beyond its historical range, and it caused epidemics (Ahmad 2000; Davies 2006), predominantly through the (legal or il- with significant mortality in the Mediterranean countries between legal) commercial transport of live animals. In these naive areas, RVF 1998 and 2004 (Mellor et al. 2008) and in northwestern Europe in develops as epidemics, usually resulting in significant human as well 2006–08 (Saegerman, Berkvens, and Mellor 2008). The overall eco- as animal fatalities. In areas where RVF is known to regularly circulate, nomic impact of these epidemics has not been estimated at the extensive epidemics are commonly reported when there is a concor- European scale. However, studies have attempted to integrate all dance of reduced population immunity of livestock (Thiongane et al. direct and indirect costs in some countries. In the Netherlands, for 1994) and an increase in vector activity triggered by climatic or other example, Velthuis et al. (2010) estimated that the BT epidemics had a environmental events (Chevalier et al. 2004). In particular, floods may net cost of 32.4 million and 164–175 million euros in 2006 and 2007, trigger the “en masse” hatching of Aedes vector eggs that can harbor respectively. Comparable figures were obtained in the analysis by the virus during inter-epidemic periods (Linthicum et al. 1985). Häsler et al. (2012) in Switzerland. Similarly, the 2007 BTV-8 epidemic in France was estimated to have cost $1.4 billion (Tabachnick, Smartt, The impact of RVF has only been formally quantified in a limited and Connelly 2008), and it is important to note that a large share of number of studies carried out in Kenya (Rich and Wanyoike 2010), these costs arose from restrictions on movement and trade. Given Tanzania (Sindato, Karimuribo, and Mboera 2012), and Somalia that the Netherlands has a standing stock of approximately 4.5 mil- (Cagnolati, Tempia, and Abdi 2006). The direct economic impact of lion cattle and small ruminants, that the EU has 271 million cattle the disease is due to the loss of livestock; the indirect impact on the and small ruminants according to FAOSTAT (FAO 2009), and that the value of surviving stock and levels of trade can also be considerable. extent of the Bluetongue invasion included many different European In Tanzania, for example, the losses caused by the 2006/07 epidemic countries, one can safely assume that the financial impact of the epi- were estimated as 16,973 cattle, 20,913 goats and 12,124 sheep, demics must have been in the range of hundreds of millions of euros. corresponding to a value of $6.44 million (Sindato, Karimuribo, and Even in those countries where the disease did not fully take hold, Mboera 2012). In Somalia, Cagnolati et al. (2006) estimated that the such as the United Kingdom, costs of vaccination were considerable. epidemic prevented 8.2 million small ruminants, 110,000 camels, and 57,000 cattle from being exported, corresponding to economic A comprehensive economic impact assessment of East Coast fever losses for the livestock industry estimated at $109 million for the was made by Minjauw and McLeod (2003) in their study of the first ban (February 1998–May 1999) and at $326 million for the impact of tick-borne diseases in Asia and Africa. They calculated an second ban (September 2000–December 2002). In Kenya, Rich and annual cost of $88.6 million in Kenya, $2.6 million in Malawi, $133.9 Wanyoike (2010) estimated the overall cost of an RVF outbreak to the million in Tanzania, and $8.8 million in Zambia. In a different study, economy, including all potential indirect impacts, to be $32 million. Mukhebi, Perry, and Kruska (1992) estimated the cost of ECF to be Simulation studies have also been looking at the potential outbreaks $168 million in eastern, central, and southern Africa. In Tanzania, the in countries where the disease is currently absent. They concluded, total economic loss caused by ECF was estimated at $247.7 million for example, that an RVF epidemic spreading through Southeast by Kivari (2006)—somewhat higher than that estimated by Minjauw Texas could lead to total costs ranging between $121  million and and McLeod (2003). $2.3 billion (Hughes-Fraire et al. 2011). The economic and livelihood impacts of RVF and ECF in Africa are Bluetongue has been historically broadly distributed and endemic therefore extremely high and diverse. The impact of these diseases between 35°S and 40°N parallels, where it has a relatively limited on poor livestock owners, in addition to the direct loss of animals, clinical impact except on exotic breeds that have been recently includes reduced production and household meat consumption, introduced. For this reason, BT was not listed as a disease with a high high cost of animal health care, reduction of inputs to crop systems, impact on the poor by the ILRI study (Perry, Randolph, and Thornton falls in stock value, inhibited access to communal grazing areas, and A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 12 CHAPTER 1 — KNOWLEDGE FOR AC TION a decrease in social capital such as perceived wealth and status. All of the costs that need be integrated into any economic impact these factors influence the cash flow and income of poor livestock assessment (Häsler et al. 2012). With economic assessment having to owners, their nutritional status, and ultimately their entire livelihood include both direct and indirect impacts influenced by the epidemic (Minjauw and McLeod 2003). and national disease control strategies themselves, the overall assess- ment across multiple countries for a disease like BT is challenging. Primary regions of concern for these diseases are Africa, the Middle East, and North Africa for RVF; Africa, the Middle East, North Africa, With RVF, systematic cross-sectional or longitudinal surveillance Europe, Central Asia, South Asia, East Asia and the Pacific, and Latin data are scarce. The impact of the disease in countries where ani- America and the Caribbean for Bluetongue; and Africa for ECF. mals are found to be sero-positive but do not have outbreaks, or in periods between epidemics, can be assumed to be relatively low. 1.3.2 Quality/Robustness of Data In epidemic conditions, detailed data have been obtained through The figures provided in the preceding section demonstrate that case reports and targeted surveys. The available economic impact while RVF, BT, and ECF have a substantial economic impact, the assessments relate primarily to the number of outbreaks and to estimates vary considerably in terms of the extent and level at the morbidity and mortality in animals and people. In these cir- which they have been estimated, and the quality of the data on the cumstances, assessing the indirect impact of the epidemics can be economic impact of these diseases can be improved. The obstacles carried out through market surveys quantifying the economic loss to effective assessment are different for each disease. along the value chain as well as the consequences of trade disrup- tion (Rich and Wanyoike 2010). For BT, the developed countries affected by outbreaks generally have good veterinary services and infrastructure, as well as a disease With respect to ECF, the lack of available estimates of the prevalence registry that allows the estimation of direct impact in terms of mor- and incidence of different tick-borne diseases makes it difficult to tality and/or abortion caused by the disease (Perrin et al. 2010). The determine individual impact (recalling that ECF is merely one type assessment can become somewhat more difficult when the direct of theileriosis). Nevertheless, the regions (see figure 1.5) are poten- impact of the disease includes reductions in productivity of milk tially at risk from the introduction (or re-introduction) of tick-borne or meat, which may vary according to management practice and pathogens. This situation is illustrated by T. parva, the cause of breed. In addition, as highlighted by Wilson and Mellor (2009), the classical East Coast fever, which was eradicated from South Africa, direct costs of BT represent only a fraction of the cost incurred by Swaziland, and southern Mozambique some 40 years ago (Norval the disease, and a large share of the cost is due to trade restrictions. et al. 1991), although its tick vector, Rhipicephalus appendiculatus, is As a consequence, the costs are strongly dependent on the control still abundant. By using available data on incidence, and by combin- strategy that is being implemented and that may change over time ing these data with livestock and vector distribution and the costs in response to the epidemiological situation. of tick control, Minjauw and McLeod (2003) were able to estimate the annual financial impact of ECF at a continental level. Estimates Movement restrictions, for example, that were implemented in of costs by different authors are not always comparable, however, Europe at the start of the BT epidemics had a very high economic because they distribute tick control costs between each of the dis- impact. As the epidemic progressed, the restriction zones were eases being controlled (for example, anaplasmosis, babesiosis). modified. Figure 1.6 shows the distribution of BT restriction zones in January 2008, February 2009, and March 2012, with each color In sum, the different diseases (and composite ecology) present a indicating regions from which movements were restricted. This range of challenges in estimating total economic costs and illustrate illustrates how the spatial structure of these restrictions evolved over the difficulty in applying any one methodology to a portfolio of time, effectively complicating economic assessments. Other preven- diseases that are related only in that they are vector-borne and are tion and control strategies such as serological surveillance in sentinel sensitive to climate. Accurate estimations of costs require uniquely animals, entomological surveillance, and vaccination are also part considered methodology on a disease-by-disease basis. R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S 14 CHAPTER 1 — KNOWLEDGE FOR AC TION of climate change” (IPCC 2007). Understanding a population’s capacity assess human vulnerability to climate-sensitive animal disease. to adapt to new climate conditions is critical to realistically assessing The unique characteristics of each climate-sensitive livestock the impacts of climate change (Kovats, Ebi, and Menne 2003). disease and specific region within which it acts ensures that there will be a spectrum of vulnerable populations that require This report explores human vulnerability (inclusive of health, econom- country-level and local investigation. Climate change will simply ics, and livelihoods) to the impacts of climate change on animal dis- magnify the risk to these populations and compound existing eases. Little work has yet been done on this facet of climate change issues of poverty and disease (Woodward 2011). Identification impact, although some insights can be derived from the work done of vulnerability should be performed on regional, national, on human health, which states that health vulnerability to climate subnational, local, and individual levels. Multiple vulnerabilities change can be defined as a function of sensitivity—the extent to will increase the relative risk to certain populations and can be which health, or systems upon which it depends, are influenced by considered as either additive or multiplicative, depending on the changes in weather and climate (the exposure–response relation- specifics. Identifying and comparing relative vulnerabilities pro- ship), of the levels of exposure to weather or climate-related hazards vides some insight as to which adaptations can be most useful (including the magnitude, rate, and character of climate variation), and and where they are most effectively implemented. And this will of adaptation—the measures and actions that can reduce the burden help establish a project scope. In particular, minimizing climate- of specific adverse health outcome (Kovats, Ebi, and Menne 2003). sensitive livestock disease risk requires identification of vulnerable Applying this framework to assess the health vulnerability of ani- populations before the threats can be targeted and reduced or mals provides the first component of a tool that can be used to eliminated (Ebi et al. 2011). R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S C H A P T E R 2 — A C T I O N A B L E TO O L S TO R E D U C E C L I M AT E - S E N S I T I V E D I S E A S E R I S K S 15 Chapter 2 ACTIONABLE TOOLS TO REDUCE CLIMATE- SENSITIVE DISEASE RISKS time period, and location; the integration of data at a higher level This chapter describes the tools and components of early and their analysis; and dissemination of results and recommended warning systems for the risk management of climate-sensitive actions to stakeholders. In many countries, animal disease surveil- disease: surveillance systems, risk mapping, and disease out- lance is typically organized through an active network of veteri- looks. Underlying knowledge, applications, and best-practice nary officers at different administrative and field-based levels. In examples are provided. recent years, a number of surveillance systems have developed to include a centralized authority with inputs from stockholders 2.1 SURVEILLANCE SYSTEMS themselves. Key Messages: Good diagnostic capacity underlies accurate data collection. Clinical diagnosis and sample collection are carried out in the field, although • Surveillance systems are key to knowing where and when a disease occurs; they also provide baseline data laboratory diagnosis can be used to confirm clinical suspicions. for risk models. Laboratories equipped to run these diagnostics are often centralized • Both active and passive surveillance are important tools at subnational, national, or international levels (OIE reference labora- that can be used to generate most accurate disease tories, for instance). Increasingly, laboratory diagnosis can be decen- profiles. tralized through the use of rapid field test kits. The benefit of these is • Geospatial and information technology is increas- ingly important for developing accurate surveillance that they provide a first screening of field samples so that only posi- methodologies. tive samples are sent for final confirmation at centralized (national or reference) labs. Effective data management and information systems 2.1.1 Surveillance Systems: General Principles are essential to ensure a smooth and rapid flow of information back Knowing where and when a disease is circulating is key to informed and forth between the central veterinary services and field veterinar- disease prevention and control strategy. Good surveillance systems ians. Together, field kits and centralized facilities can provide a rapid have high detection sensitivities (the capacity to detect disease and comprehensive perspective of the disease situation. events), specificity (avoiding false positive detection), simplicity, For the purposes of this report, surveillance can be thought of in two adaptability (ability to scale up in the case of an unexpected event ways: passive and active. Passive surveillance is the routine collec- or epidemic and to scale down when disease impact is low), and tion of disease reports from field practitioners who are themselves cost-efficiency (Dufour, Hendrikx, and Toma 2006). These attributes informed of potential cases by livestock owners reporting clinical are true regardless of whether a disease is human- or animal-specific. manifestations. Active surveillance is the active search for new cases Surveillance is a continuous and systematic process that can be in the field and is usually carried out during high-risk periods or in characterized by the collection of relevant data for a population, high-risk regions (that is, at the beginning of an epidemic). A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 16 C H A P T E R 2 — A C T I O N A B L E TO O L S TO R E D U C E C L I M AT E - S E N S I T I V E D I S E A S E R I S K S Other web-based initiatives developed for human health, such as Active Surveillance in Action Healthmap (http://healthmap.org), which is based on automated The Thailand Department of Livestock Development launched data mining of digital news reports, also offer useful templates a country-wide survey involving several hundred thousand for animal health surveillance. A number of initiatives have been trained volunteers to search door-to-door for evidence of highly pathogenic avian influenza (HPAI) H5N1 (Tiensin et al. undertaken in recent years to improve disease surveillance (IOM 2005). The surveys enabled the creation of an unprecedentedly 2007), with support from top philanthropic organizations such as detailed data set of HPAI cases and poultry census data at the the Rockefeller foundation, the Bill & Melinda Gates Foundation, and village level and effectively helped the country to efficiently Google.org, tackling many of the challenges common to human target control and surveillance. and animal health. In some developing countries, epidemiological networks enabling 2.1.2 Surveillance Systems: Knowledge and Applications disease surveillance have been gradually developed (despite Rift Valley Fever sometimes unfavorable contexts of declining resources and infra- Due to limited veterinary resources in countries affected by RVF, structure) and are becoming technically and institutionally well surveillance during inter-epidemic periods is neither continuous established (Bendali 2006; Ouagal et al. 2008). Yet they are often nor systematic. Rather, a number of project-based or local studies limited by poor access to field information and insufficient labora- have tried to establish the presence of RVF, predominantly through tory diagnostic capacities. Further, limited external financing that is sero-prevalence studies to confirm the disease in livestock or hu- disease-crisis-specific can prevent long-term sustainable solutions mans. This methodology has been employed in countries where the (Ouagal et al. 2008). disease was not previously known to occur, such as Burkina Faso Participatory disease surveillance is a particular form of active sur- (Gonzalez et al. 1992), Cameroon (LeBreton et al. 2006), Chad (Ringot veillance that occurs at the village or household level and that can et al. 2004), Gabon (Pourrut et al. 2010), Niger (Mariner, Morrill, and complement centralized surveillance programs. Livestock owners Ksiazek 1995), and Nigeria (Olaleye, Tomori, and Schmitz 1996). In are most often able to recognize major disease problems in their countries where RVF is known, such as Egypt (Abd el-Rahim, Abd area. Questionnaires and active community engagement can aid in el-Hakim, and Hussein 1999), Kenya (Munyua et al. 2010; Murithi the epidemiological risk assessment of any given area and can be et al. 2011), Madagascar (Andriamandimby et al. 2010), Mauritania carried out by an investigation team going from village to village to (El Mamy et al. 2011), Senegal (Chevalier, Thiongane, and Lancelot administer the questionnaires. The technique has been instrumen- 2009), Somalia (Soumare et al. 2007), Sudan (Hassan et al. 2011), tal in global Rinderpest eradication programs (Jost et al. 2007) and South Africa (Archer et al. 2011), Tanzania (Mohamed et al. 2010), has been used in both rural and urban settings in Africa (Malak et al. Yemen (Abdo-Salem et al. 2006), and Zambia (Samui et al. 1997), 2012) and Asia (Azhar et al. 2010). serosurveillance has also been a helpful tool (Thiongane et al. 1994). At the start of an epidemic, often the only disease data available Over the past several years, geospatial and information technology has been acquired through case reports and targeted, or risk-based, has supported the development of innovative approaches for dis- surveys (Soumare et al. 2007; Munyua et al. 2010). The general lack ease surveillance and mapping. Mostly developed for human health of inter-epidemic surveillance means that awareness of RVF risks is applications, initiatives such as the Google trends project can pro- lowered, and the early detection of epidemics therefore less likely, vide early warning of epidemics based on the frequency of Google resulting in diminished timeliness and effectiveness of appropriate searches on the relevant disease-related terms (Ginsberg et al. 2008). mitigation measures (Jost et al. 2010). Widespread use of mobile technology in the developing world also has much to offer to disease-related information exchange be- Both main elements triggering RVF epidemics—reduced immu- tween stakeholders and formal authorities and is being piloted in nity and rainfall-induced vector increases (see section 2.2.3)— some studies in Africa (Aanensen et al. 2009). (See also LIDC 2010). can be targeted by longitudinal serological and entomological R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S C H A P T E R 2 — A C T I O N A B L E TO O L S TO R E D U C E C L I M AT E - S E N S I T I V E D I S E A S E R I S K S 17 surveillance. GIS-based approaches that use environmental and certain Bluetongue serotypes.” Other objectives may include the climatic data (Anyamba et al. 2009), data on historical outbreaks, seasonally vector-free period and identifying vector species (EC statistical sampling theory, or even expert knowledge (Clements, Regulation No 1108/2008). Outside the restriction zones, surveil- Pfeiffer, and Martin 2006) can be used to target these surveillance lance targets “detecting any possible incursions of the Bluetongue efforts and optimize the resources needed. Importantly, local virus and demonstrating the absence of that virus in a Bluetongue- knowledge can be polled: Jost et al. (2010) showed that pastoral- free Member State or epidemiologically relevant geographical ist livestock owners “were aware of the unusually heavy nature of area” (EC Regulation No 1108/2008) and is carried out effectively the rains and flooding before the outbreak of RVF in their areas, through passive clinical surveillance and active laboratory-based noticed mosquito swarms that were unusual because of their surveillance based on at least one annual serological/virological intensity and the physical characteristics of the species involved survey. Further details of the BT surveillance obligations are set (Aedes spp.), and noted unusually high morbidity and mortality in out in the EC Regulation No 1108/2008. EU member countries are their flocks consistent with RVF.” Much benefit can be gained for requested to submit to the EU monthly, biannual, or yearly reports, RVF surveillance from livestock owners’ knowledge and participa- depending on their restriction zone status. In practice, each mem- tory approaches, particularly if paired with effective data analysis ber state has the freedom to establish its own surveillance system, and dissemination. provided that it complies with the EC minimum regulation and that countries have developed several integrated information sys- Surveillance and mitigation measures are difficult to disentangle. tems to collect and disseminate Bluetongue disease and vector Recent decision-support systems reflect this, such as those devel- surveillance data. oped by ILRI and FAO, which strongly support a phased approach, unfurling a series of interventions that includes communication, Since BT was first detected in Italy in August 2000, authorities coordination, surveillance, early warning systems, and disease and have invested substantial resources to develop a structured sur- vector control, that are specific to each region and geography of veillance and early warning system for the disease (Giovannini outbreak (ILRI/FAO 2010). et al. 2004a, 2004b). The surveillance system is based on two main components. The first, periodic testing of unvaccinated sentinel Bluetongue cattle uniformly scattered throughout the country, aims to assess Current surveillance for Bluetongue in the developing world is rare; the incidence of infection in non-immunized strata of ruminant the current state of the art is illustrated by the surveillance concen- animals and to inform movement control planning. The second, a trated in regions in the developed world that have been afflicted network of permanent traps sampled weekly year-round, intends with the disease or are considered at risk for the future. Yet it is to define the geographical distribution of vectors and their sea- included here given that some of the lessons learned from devel- sonal population dynamics. All surveillance data are integrated into oped world experiences may have something to offer to develop- a GIS-based information system with a web interface that allows ing regions. the collection, management, and dissemination of data collected At the European level, Bluetongue surveillance is mandatory under by field veterinarians (Conte et al. 2005). The surveillance is based European Commission Regulation No 1266/2007 and must include upon daily “a) recordings of all suspected and confirmed BT clinical clinical, serological, and entomological components. The obliga- cases; b) recordings of the results of periodic testing of sentinel tions differ according to whether a given serotype is considered animals; c) reports on monitoring of the spread of vectors and to be present (see figure 1.6). Within restriction zones, surveillance their seasonal dynamics; d) recordings of all diagnostic results; is carried out through networks of sentinel unvaccinated, suscep- e)  recordings of the progress of vaccination campaigns” (Conte tible animals and through networks of vector surveillance. The et al. 2005). This system is beneficial to both centralized decision overall objective of these programs is “to detect the introduction makers and field veterinarians, providing national-level perspective of new Bluetongue serotypes and to demonstrate the absence of and information that can assist practitioners in daily tasks (that is, A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 18 C H A P T E R 2 — A C T I O N A B L E TO O L S TO R E D U C E C L I M AT E - S E N S I T I V E D I S E A S E R I S K S identifying municipalities in the infected zone or in buffer radii). 2.2 CLIMATE-SENSITIVE DISEASE RISK MAPS Here, too, the collection of surveillance data is tightly linked with Key Messages: control and mitigation operations (mainly movement control and vaccination), with the information systems ensuring a swift flow • Risk maps rely heavily on accurate collection of disease data. of information between the different levels of responsibilities. • Risk maps enable better prioritization of surveillance, Livestock owners can also benefit from the system through access prevention, and mitigation efforts. to a publicly available view of web-based information systems, for • Risk maps have been used extensively for BT and RVF instance to verify the BT status of geographical areas where they and to a lesser extent for ECF. want to move animals. • In many studies, risk maps have consisted of mapping the distribution of vectors; recent works have enabled Many other developed countries have developed BT surveil- modeling of diseases themselves. lance by using modeling and/or simulation target surveillance (Bonfanti et al. 2008; Racloz et al. 2008; Gubbins et al. 2010; 2.2.1 Risk Maps: General Principles Szmaragd, Gunn, and Gubbins 2010), by identifying new meth- The distinction between risk maps and early warning systems ods to survey vector populations (Meiswinkel et al. 2008), or by is nuanced. Early warning systems can best be thought of as developing innovative surveillance systems. For example, Hadorn the comprehensive set of information and actions that alert et al. (2009) presented a cost-effective surveillance system for BT decision makers to impending harm. Risk maps are merely one in Switzerland based on the combination of improved passive component of this in that they are the application of data to a clinical surveillance in cattle and sheep, relying on increasing visual media that facilitate the communication of threats. Risk awareness of BT symptoms by farmers (Stuber et al. 2009), con- mapping is a useful tool in disease mitigation in that it can be current with a targeted bulk milk-testing strategy of dairy cattle used to distinguish areas that experience epidemic and seasonal herds in high-risk areas. transmission from those with more stable or continuous transmis- sion patterns, for which EWS will be less useful (Kuhn et al. 2005). East Coast Fever Maps make it possible to visualize areas of greatest threat so that The surveillance of ECF is not systematic in the countries where disease mitigation efforts can be most effectively developed and the disease is present, and similar to RVF, surveillance tends to be implemented. replaced by case reports. Several specific projects have neverthe- less attempted to establish the distribution of ECF prevalence in Creating risk maps requires a variety of technical inputs that vary a number of countries: in domestic cattle in Zambia (Simuunza by region and disease. In the case of climate-sensitive diseases, it et al. 2011), in free-ranging buffaloes in Namibia (Pascucci et al. requires both environmental and disease data. Once the environ- 2011), and in longitudinal studies in Uganda (Rubaire-Akiiki et al. mental parameters that affect a disease are defined, data points can 2006; Ocaido, Muwazi, and Opuda 2009). ECF vectors have also be collected from a variety of environmental and health resources been subject to cross-sectional studies providing data that may (table 2.1). Multiple environmental indicators for disease can be support epidemiological surveillance at the country scale, for overlaid to produce the most comprehensive results. Disease data, example in Rwanda (Bazarusanga et al. 2007), and at the scale such as incidence and type of animal that it affects and historical of Africa as a region (Cumming 1999, 2000). An interesting pilot records of outbreaks, can then be added to the map so that correla- study for ECF is the disease surveillance program that was imple- tions can be identified. Comparisons in both spatial and temporal mented in East Africa, which explored the use of mobile phone dimensions can be made, enabling predictions for regions and collection of epidemiological data relevant to ECF, anthrax, periods of time. Other inputs, like vulnerability status, can also be rabies, Peste des Petits ruminants and foot-and-mouth disease added to the map, bolstering the robustness of the tool so that best (LIDC 2010). adaptations can be made. R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S C H A P T E R 2 — A C T I O N A B L E TO O L S TO R E D U C E C L I M AT E - S E N S I T I V E D I S E A S E R I S K S 19 TABLE 2.1: Early Warning System Data and Risk Mapping then the likelihood that it will spread from established pockets into Technical Resources nearby areas, with estimates for the last two stages of both distribu- Africa Real Time Environmental Monitoring System (ARTEMIS) tions and case numbers. Committee on Earth Observation Satellites (CEOS) Emergency Prevention System for Transboundary Animal Diseases (EMPRES) Programme European Space Agency Data Sets In reality, however, these ideal requirements are very rarely met, and FAO GeoNetwork the great majority of risk maps are based on the prediction of some Global Risk Identification Programme (GRIP) Group on Earth Observation (GEO) index of disease presence. In the case of vector-borne diseases, the International Research Institute for Climate and Society index of presence produced may not be for the disease itself but NASA Goddard Spaceflight Center OIE World Animal Health Information Database (WAHID) for its vector, as this may be easier to estimate than pathogen pres- RVF Activity Database in Kenya, Zimbabwe ence is, and the risk assessment thus produced becomes potential SERVIR Regional Visualization and Monitoring System U.S. DoD Global Emerging Infections Surveillance Program rather than actual presence. Risk (defined as an index of presence U.S. NOAA Climate Prediction Center (including ANHRR) of either vector or disease) can be assessed in the field either by di- World Animal Health Information Database (WAHID) rect measurement through surveillance and monitoring programs or by modeling prediction and projection. The former tends to be The production of disease risk maps requires a number of initial expensive and time-consuming, especially in remote areas and for decisions—namely, what sort of risk is to be mapped and what rare diseases, and is frequently replaced with reporting (which may methods are available to produce them (see figure 2.1). The defini- or may not be reliable) or the collection and analyses of disease tion of risk is a moveable feast, and while the simplest one may be records from medical facilities and hospitals. While these data are the straightforward presence of the disease, or perhaps the number certainly an improvement over no data at all, they can be mislead- of cases in a particular place, a proper understanding of the risk ing because the data sources may be biased or incomplete and posed by a disease requires estimating the risk of introduction, the only representative of certain areas or categories of host population. chance that the disease once introduced becomes established, and Surveillance is discussed further in the previous section. FIGURE 2.1: Inputs, Analytical Process, and Outputs Typically Involved in Disease Risk Mapping Model Knowledge development • Statistical • Process-based Predictors • Eco-climatic • Host distribution • Anthropogenic Prediction Analytical process • Interpolation • Extrapolation Surveillance data Validation • Accuracy assessment Inputs Applications • Prioritization of suveillance and control • Impact assessment Outputs A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 20 C H A P T E R 2 — A C T I O N A B L E TO O L S TO R E D U C E C L I M AT E - S E N S I T I V E D I S E A S E R I S K S Disease risk maps are produced by modeling in two ways, each with calibrate the models or to act as the skeleton framework within markedly different structures and assumptions: biological mecha- which new values can be interpolated or extrapolated. “Garbage nistic (or process-based) models and statistical models. Biological in, garbage out” is particularly relevant here, as the modeling will models are based on a detailed knowledge of the actual processes merely match any defect in the data with more. It is therefore abso- underlying the presence of a disease or its vector. These models are lutely essential to ensure that the input training data are as accurate clearly hypothesis- rather than data-driven, and if the hypotheses and representative as possible. In particular, it is important that a are wrong, then so will the model predictions be wrong. Data-driven robust sampling and surveillance strategy is defined, and this can or statistical models are essentially pattern matching procedures be very resource-intensive. The second important element is reli- based on the known presence of a disease (or vector) over space able host and, for vector-borne diseases, vector data. A vector-borne and/or time (Rogers and Randolph 2006). Temporal models usually disease cannot exist unless both are present, and their abundance imply time series analyses that look at trends in historical data and and distribution will determine the severity of any outbreak and its then project them into the future or that apply the trends found in potential for persistence and spread. one place to another (similar) place. They focus on looking at the Additional predictor variable datasets will be needed to apply and way disease occurrence has fluctuated in the past and assume the map the relationships identified during the modeling process. same is going to happen elsewhere or in the future. Spatial distribu- There is a wide range of potential predictor variables—essentially, tion models first find statistical relationships between a disease and any that might have a relationship to the presence or severity of a predictor data sets and then apply these relationships to the pre- disease or its vector(s), including climatic parameters such as tem- dictors of data to produce disease risk maps. There are many well- perature, precipitation, and humidity; environmental parameters established methods for this type of modeling (Elith et al. 2006), and like vegetation cover and type, soil moisture and water, host dis- they are widely available in specialized software packages. tribution; topographic variables like altitude or slope; and anthro- Statistical models are advantageous over biological models in that pogenic factors such as agriculture and use, population, and trade they are quite adaptable and can be used to model almost anything pressures. within a known distribution. On the other hand, they are difficult to Fortunately, the availability of spatial data sets has blossomed in use and require considerable supporting data, which are not always recent years and there are hundreds of datasets of global imagery available for the desired parameters. in the public domain for major climatic and environmental vari- The inputs needed for disease risk maps obviously depend on ables. With all such datasets there is an issue of scale and reso- the type of model. For biological models it may be necessary to lution in time and space, with “better” resolution requiring more know details of the population at risk—for example, the number processing but resulting in finer detail of prediction. There are also of susceptible and immune animals or people, the rate of devel- challenges in using appropriate means, averages, or summaries opment of the disease once caught, the percentage mortality, the of each covariate within the modeling procedures. Data reduc- contact rate between infected and susceptible animals/individuals, tion techniques like Fourier processing, however, can be used to the effects of environment on disease, host and vector distribution extract biologically meaningful summaries from large and long- and abundance, the movement and immigration of susceptible running datasets. and infected individuals, transmission routes and rates between In the context of climate-sensitive diseases and potential effect hosts (and, if relevant, vectors), and the role of alternative disease of climate change, it is of course necessary to acquire projected reservoirs. climate data sets, of which again there are a large number, For data-driven models, there are two primary disease-related representing a range of “scenarios” assuming different rates requirements. The first is reliable input disease (or vector) data to in carbon dioxide increase (and thus temperature and rainfall R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S C H A P T E R 2 — A C T I O N A B L E TO O L S TO R E D U C E C L I M AT E - S E N S I T I V E D I S E A S E R I S K S 21 change), and a number of global climate models (GCMs), which in the previous section concerning RVF and BT. One is the risk of predict a wide range of different outcomes. It has now become HPAI H5N1 that was mapped in Thailand based on intensive disease standard practice to use the average of a number of GCMs (so- surveillance surveys and that detailed a poultry census (Gilbert called ensembles) to try to produce a consensus set of projected et al. 2008). The data allowed identification of optimal allocation of outcomes. surveillance effort (McCarthy et al. 2010) in the countries that have succeeded in eradicating the disease, despite poor agricultural and The outputs required for a risk map may include disease incidence, environmental conditions. prevalence, presence or absence, case numbers, and vector abun- dance or presence, as well as the probability of introduction, Second, disease risk maps can also be used to evaluate how establishment, or spread or the spread rate of both vector and changes in one parameter of the model can affect the extent of disease. While there may be a preference for a particular output regions at risk. This principle underlines most studies that have in principle, the choices in practice are usually mostly dictated by looked at the potential effect of climate change on the distribu- what processes are best understood (for biological models) and tion of vector-borne disease, including BT (Guis et al. 2012) and what disease or vector data are actually available or can feasibly ECF (Olwoch et al. 2008). Disease risk maps can be combined be collected with the resources available (for data-driven mod- with socioeconomic parameters to assess and map the poten- els). The availability of such data depends on the disease and the tial economic benefit of trypanosomiasis control in West Africa location, though most frequently produced outputs are some (Shaw et al. 2006, submitted) for example, and this can be further index of presence or absence, usually expressed as a probability. developed to quantify the resulting cost benefit ratios for different Biological data-driven models can be refined by using “masks” to control techniques, as demonstrated for the Horn of Africa. The improve the outputs; for example, climate, environment, or land insights gleaned here can then be used to design and target inter- use are often used to restrict the areas where a disease or vector ventions most effectively. can be present. Third, risk maps can be useful in the development of movement In recent years it has become increasingly clear that a simple control maps. HPAI H5N1 in Thailand is also illustrative here, wherein prediction—a “one off ” model—is not enough, primarily because free-grazing ducks were identified as being the greatest risk associ- the models all have errors implicit in the relationships used to build ated with the presence of the disease (Gilbert et al. 2006). In the fol- them and because a single model may, simply by chance, be one lowing months, Thailand implemented movement control policies with a substantial error. As a result, it is important to use methods that prevented the transport of ducks over long distances, which that can provide some description of uncertainty or error in the had an immediate impact by reducing the number of outbreaks in predictions. This largely relies on producing a large number of the following year. Bovine tuberculosis is another example where models with automated software, each using a random subset of statistical data-driven approaches developed to map the risk of the the disease or vector data used to calibrate or “train” the models. disease in Great Britain quantified the impact of cattle movement The replicates are then averaged to produce a single output with on disease risk (Gilbert et al. 2005) and guided the introduction of an associated error. movement controls. Fourth, disease risk maps are an extremely efficient way of com- 2.2.2 Utility of Risk Maps municating information about diseases to decision makers and to a The development of a risk map serves many different purposes in the lay audience. Proper communication, recruitment of experts (both context of the mitigation of livestock diseases. First, risk maps help modelers and communicators), and in-country trainings will be in- better target surveillance and control in high-risk areas, for which dispensable steps in the implementation and use of this tool (see there are several other examples of uses than those highlighted also chapter 3). A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 22 C H A P T E R 2 — A C T I O N A B L E TO O L S TO R E D U C E C L I M AT E - S E N S I T I V E D I S E A S E R I S K S Although this work focuses on RVF in East Africa, the potential uses VOICES FROM THE FIELD for other diseases and regions are clear (Anyamba et al. 2012). For Using Risk Maps in Malaria Control in Africa example, in Senegal, high-resolution satellite imagery has been Dr. Judy Omumbo, Policy Impact Unit, Malaria Public Health used to map fine-scale distribution of water ponds (Tourre et al. Cluster (KEMRI/University of Oxford collaborative program) 2008; Vignolles et al. 2010) and has been combined with rainfall “We recently carried out a study to assess how risk maps are data to map the RVF risk (Tourre et al. 2009). Other studies have used in decision-making for malaria control in Africa. mapped suitable vector habitat for RVF directly using serological Climate is a key driver of malaria transmission in Africa, and data (Clements et al. 2007b). At the continental scale, Clements there are several risk maps that have been produced at the and colleagues have compiled the available data on RVF in Africa continental and national scale. So, we went through national (Clements 2007a) and used a knowledge-based approach to pro- malaria control policy documents, national strategies and applications for global funds to better understand what risk duce a comprehensive continental risk map (Clements, Pfeiffer, maps are effectively used. It was very interesting to find that and Martin 2006). risk maps had been identified for the large majority of coun- tries with endemic malaria in Africa. Those maps ranged from Bluetongue simple eco-climatic descriptions to more complex maps of modeled malaria-parasite prevalence. The majority of risk maps published for BT have intended to map the distribution of disease vectors as indicators of spread poten- However, only five countries used national malaria maps to de- sign, control, or make decisions on how to allocate resources. In tial. This has been carried out at continental and national levels in other words, a limiting factor is not necessarily the availability of Europe and North Africa (Baylis et al. 2001; Wittmann, Mellor, and risk maps, but rather is the development of the science to policy Baylis 2001; Tatem et al. 2003), Spain (Calvete et al. 2008; Acevedo and practice dimensions. There is a need to explore ways to im- et  al. 2010), Italy (Conte et al. 2003, 2007), Calabria (Calistri et al. prove how these types of data are used for more effective control. 2003), Sicily (Purse et al. 2004), France (Guis et al. 2007), Morocco One way to achieve this is to build platforms at the country (Baylis and Rawlings 1998; Baylis et al. 1998), and South Africa (Baylis, level to better link risk mapping with policy, strategic plan- ning and financing. In addition, ensuring country ownership Meiswinkel, and Venter 1999). The distribution maps were produced of epidemiological risk maps and research outputs can better using data-driven approaches. They used several remotely sensed enhance their value and application in the long-run.” indicators to identify the eco-climatic signatures characterizing locations where vectors were known to be present and applied the resultant statistical model to the predictor variables in order to 2.2.3 Risk Maps: Knowledge and Applications map the areas with similar conditions. These models have proved Rift Valley Fever particularly useful in mapping areas at risk of BT transmission and Risk maps generated by Ken Linthicum and Assaf Anyamba have in identifying key variables influencing the distribution of vectors. successfully predicted RVF outbreaks in the Horn of Africa. Using sea However, the predictions of data-driven models depend heavily on surface temperatures (SSTs), rainfall, and the Normalized Difference the data that have been used to “train” them. Consequently, differ- Vegetation Index (NDVI), they have been able to show high risk cor- ent models based on different training sets may produce different relations with El Niño–Southern Oscillation (ENSO) and to map these outcomes for the same region, as in the cases of Italy and Sardinia areas for use in early warning systems (Witt et al. 2011; Anyamba (Calistri et al. 2003; Pili et al. 2006). Additionally, since these mod- et  al. 2009). The maps are then transmitted to international orga- els only attempt to predict the distribution of known and identi- nizations and governments to warn them of the high disease risk fied vectors, they cannot predict some events like the spread of resulting from the ripe environmental conditions. BT through new vector groups, as was the case during the 2006 This work has been instrumental in illustrating how remote BTV-8 epidemic in the Netherlands, Belgium, France, and Germany. sensing can be used to predict vector-borne disease epidemics. Here, BT was introduced well beyond the distributional limits of the R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S C H A P T E R 2 — A C T I O N A B L E TO O L S TO R E D U C E C L I M AT E - S E N S I T I V E D I S E A S E R I S K S 23 traditional vector of Bluetongue in the Mediterranean basin, and When BTV-8 was spreading in northwestern Europe, countries it was in fact later demonstrated to be mediated by entirely differ- such as France and Germany used GIS-based risk maps to prioritize ent groups of Culicoides vectors with different habitat and climate surveillance, targeting the fringe of vector distribution, looking preferences (Purse et al. 2007, 2008). for expansion (Racloz et al. 2008; Koslowsky et al. 2004). Risk maps based on wind-modeling were also used in a number of cases to Over the last few decades, BT has definitively expanded its range assess BTV risk in the United Kingdom (Gloster et al. 2007a) and to northward in Europe—the likely result of climate change caus- define (and subsequently modify) movement restriction zones and ing the latitudinal shift of its primary Mediterranean basin vector, areas where vaccination was recommended, as well as to predict its C. imicola (Purse et al. 2005). Yet with the BTV-8 epidemic that spread northward spread (Ducheyne et al. 2011). Such techniques enabled through other communities of vectors in northwestern Europe, it veterinary officers in Belgium to decide whether there was a risk of became clear that this mechanism could only partly explain the introduction of BTV from France and thus whether there was a need shift of BT and that other mechanisms, such as changes in vector for preventative measures. identity, vectorial capacity, extrinsic incubation period, or biting rate possibly linked to climate change could have also had some caus- East Coast Fever ative effect (Guis et al. 2012). Compared with RVF and BT, there has been relatively little work to map the risk of ECF. Two studies have mapped the distribution The distribution of vectors mapped through data-driven models has of ECF vectors in Africa (Randolph 1999; Cumming 2000), and risk not been the only approach used to map the risk of BT. Hartemink maps integrating vector, host, and/or climatic data have been pro- et  al. (2009) developed a process-based mathematical model duced in a limited number of studies in the central highlands of that integrates data on the distribution of vectors, animal hosts, Kenya (Diaz et al. 2003), in Zimbabwe (Pfeiffer et al. 1997), and at the and transmission parameters to map the distribution of R0 in the scale of Africa (Lessard et al. 1988). Recently, a study by Olwoch et al. Netherlands. (R0 is the average number of secondary infections aris- (2008) used tick, climate, and cattle data to predict the possible ef- ing from one single infected in a totally susceptible population; an fect of climate change (as predicted by the nested regional climate epidemic dies out if R0 is < 1 and may spread if R0 > 1.) Other authors model DARLAM) on ECF in sub-Saharan Africa. have used and analyzed case data to investigate the factors associ- ated with BT presence directly (Allepuz et al. 2010; Silbermayr et al. 2.3 CLIMATE-SENSITIVE DISEASE OUTLOOKS 2011), or to quantify the rate of disease spread (Gerbier et al. 2008; Pioz et al. 2011). Recently, Guis et al. (2012) published a study based Key Messages: on an integrated mechanistic model of BT transmission risk that quantified the potential role of climate change in the northward • Disease outlooks aim to provide long-term projections of disease trends so that disease control and mitigation expansion of Bluetongue in Europe over the past several decades. efforts can be integrated into long-term planning. The model was then applied to an ensemble of 11 regional climate • Few disease outlooks are yet available for any diseases. models to project the distribution of BT risk in the future. It has also been noted that BT can spread over long distances 2.3.1 Disease Outlooks: General Principles though wind-aided dispersal of its vectors, and several studies Disease outlooks apply in two general situations: the introduc- have found spatio-temporal correlations between the wind and BT tion and establishment of a disease in a region where it was pre- spread (Ducheyne et al. 2007; Kedmi et al. 2010; García-Lastra et al. viously absent and the increase in incidence and/or prevalence, 2012; Sedda et al. 2012), establishing the basis of research efforts to or the occurrence of more epidemics, within regions where produce maps quantifying the wind-resultant BT risk (Gloster et al. diseases are already present. Developing long-term disease 2007b; Hendrickx et al. 2008; Ducheyne et al. 2011). outlooks or projections is challenging because different factors A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 24 C H A P T E R 2 — A C T I O N A B L E TO O L S TO R E D U C E C L I M AT E - S E N S I T I V E D I S E A S E R I S K S may influence introduction and spread. For example, in a region 2.3.2 Disease Outlook: Knowledge and Applications where a disease is absent, climate may become suitable as a re- Few disease outlooks or projections have been published in the sult of climate change, but the disease may never spread if it is literature regarding any climate-sensitive diseases that might affect never introduced through trade, tourism, wind patterns, or some rural pastoralist farmers. other transplantation event. Conversely, repeated introduction Outlooks have been produced for some diseases in the European may never lead to establishment and spread if the environmental Union, and RVF is frequently cited in risk assessment as being a pos- conditions are not exactly right, given the subtle sensitivities of sible contender for introduction to Europe (López-Vélez and Molina both diseases and vectors. Furthermore, projections often ignore Moreno 2005; Martin et al. 2008) or the United States (Konrad, Miller, the evolutionary capacity of disease and vectors to adapt to new and Reeves 2011; Hartley et al. 2011), although the precise mecha- conditions, given the complexity in incorporating this into a nisms supporting those increased risks are diverse. These outlooks model. Adding to the difficulty of definitively making long-term are further supported by studies that gauge expert opinion (Gale climate-related projections is the temporal uncertainty and spa- et  al. 2010), also identifying RVF as one of the diseases believed tial heterogeneity associated with climate change projections on to have an increased risk of introduction into Europe (EFSA 2013). global and regional scales. Climate matching approaches and projections according to climate Health impact models are also constrained in that there are a num- change scenarios have been produced to map the projected dis- ber of inputs besides just environmental variables that determine tribution of the vectors for BT in Spain (Acevedo et al. 2010) and their outcomes. A multiplicity of socioeconomic factors and policy for ECF in Africa (Olwoch et al. 2008), and they found respectively a decisions determines diseases and can often be too complicated low impact and a noticeable impact in vector distribution. Perhaps to include in long-term health models, regardless of whether these the most elaborate model used to make predictions on disease has are for animals or humans (IPCC 2007). For example, failure to imple- been produced at the European level by Guis et al. (2012), who used ment appropriate and timely mitigation measures in the United an epidemiological model of BT transmission, adapted for two hosts Kingdom resulted in an excess of bovine TB, a previously control- and two vectors to quantify the possible impact of climate change lable disease. Regarding the environmental determinants alone, im- on BT R0 in northwestern (a 4.3 percent increase per decade) and provements or expansion of control activities may prevent spread, southwestern Europe (a 1.3 percent increased per decade). Most of or indeed eliminate a disease from areas where it is historically pres- this increase is mediated by the effect of temperature on the extrin- ent, as in the cases of malaria in Europe and North America and of sic incubation period (the period between infection of the vector rinderpest in Africa. and its ability to infect the next host). Although there are currently no models for these outlooks in Asia, Africa, or South America, les- In considering the economic impacts of diseases in the future, it is sons learned from the European studies can be applied to diseases important to note that the link between GDP and burden of dis- in these regional contexts. eases is confounded by social, environmental, and climate factors (Arnell et al. 2004; van Lieshout et al. 2004). In this context, global Establishing outlooks is important when building long-term disease macroeconomic trends can drastically affect resources directed to mitigation plans as it provides a framework for governments to in- disease mitigation or the research needed to develop new control vest in research in order to reduce uncertainties and to develop dis- methods, and such factors can dwarf the likely impact of climate ease mitigation efforts. Although it is true that climate change will and environmental change of disease levels (Gething et al. 2010). affect diseases in regions differently, current data indicate that, on Discerning the economic impacts of these future disease threats average, emerging climate patterns—like increased temperatures is therefore virtually impossible because economic scenarios can- and precipitation—will lead to increased geographic distribution not be directly linked to disease burdens and most attempts to of certain diseases. It may also be that “what if” scenarios can be do so are often constructed as “what if” scenarios rather than firm effective tools for planning, particularly if they incorporate socio- projections. economic or policy-related factors. R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S C H A P T E R 2 — A C T I O N A B L E TO O L S TO R E D U C E C L I M AT E - S E N S I T I V E D I S E A S E R I S K S 25 2.4 EARLY WARNING SYSTEMS VOICES FROM THE FIELD Key Messages: The Liver Fluke Climate Forecast in the United Kingdom • Early warning systems aim to provide short- or mid- Prof. Matthew Baylis, Head of the Department of Epidemiology term disease forecasting so that appropriate interven- and Population Health Institute of Infection and Global Health, tions and mitigation efforts can reduce the impact of an University of Liverpool. epidemic. • Climate-based EWS have been developed for RVF in “An interesting example of early warning system applied to live- East Africa and have proved useful in predicting recent stock diseases in the United Kingdom is the liver fluke forecast. outbreaks. What drives the forecast is a monthly ‘fluke index’ based on the relative levels of rainfall and potential transpiration. If rainfall is higher than potential transpiration, then there is a net accu- 2.4.1 Early Warning Systems: General Principles mulation of moisture on grass, leading to a higher index. This Because the geographic and seasonal distributions of many infec- is done for those months (May to October), which are warm tious diseases are linked to climate, the notion of using climate enough for flukes to survive. variables to predict disease and establish EWS has long been an According to Dave William, from NADIS (http://www.nadis. area of academic, practical, and political interest. Many of the major org.uk/), the forecast is emailed every month, in slightly differ- ent formats depending on the recipients to i) all the UK vet- climate-sensitive human diseases are associated with some sort of erinary practices who registered into the system, ii) farmers EWS research or development activity. Climate-sensitive animal dis- associations of the ruminant sector (English Beef and Sheep eases are also increasingly being explored. Capabilities in building Meat Industry EBLEX, Hybu Cig Cymru Meat Promotion Wales efficient disease EWS blossomed in the 1990s, coinciding with the HCC, Quality Meat Scotland QMS), iii) to the Animal Medicine Training Regulatory Authority to be distributed through their widespread availability of relevant spatially explicit environmental to Continuous Professional Development training and iv) to data, improvements in data storage and epidemiological modeling farm businesses. The forecast is also featured on the NADIS technology, and increased awareness of anthropogenic climate web site. change (Kuhn et al. 2005). An interesting feature of the forecast is that it stimulates discus- sion between the farmer and his vet about individual farm con- Using climate data to predict disease occurrence or outbreak dates ditions. It highlights sustainable control of parasites in sheep in from the first half of the twentieth century. Researchers in India de- a seasonal context and the vets central role in parasite control. veloped an early warning system for malaria based on rainfall, the So, the system is not directly targeted at farmers themselves, prevalence of enlarged spleens, economic conditions (such as the but at veterinarian practitioners as a way to improve their ser- price of grains), and a coefficient for epidemic potential. The system vice and interactions with their customers, which in return im- predicted epidemics from 1921–49 in Punjab, and retrospective proves the overall control of those parasitic diseases.” analyses have revealed the probable accuracy of the projections The current forecast can be viewed at http://www.nadis.org. uk/parasite-forecast.aspx. (Swaroop 1949). Additional work was done to explore the associa- tions between pneumonia, smallpox, tuberculosis, and leprosy and various climatic variables such as temperature, humidity, rainfall, and EWS. Disease diagnostic tools to monitor incidence and transmis- wind. Datasets were kept on decadal scales for thousands of sites, sion are globally available; environmental monitoring systems, such demonstrating the potential feasibility and utility of such methods as satellites and meteorological stations, are accessible online; and for widespread surveillance, even when using few variables, primi- advances in statistics and epidemiology allow for more accurate tive models of measurement, and incomplete knowledge of the measurement of climate-disease associations. At present, many effects of climate on all aspects of disease (Kuhn et al. 2005). early warning systems exist for a range of climate-sensitive impacts. Today the health sector (animal or human) is better positioned to Famine, for example, though not explicitly disease-related, has clear provide the required inputs and utilize the potential benefits of association with climate through its effects on crop production, A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 26 C H A P T E R 2 — A C T I O N A B L E TO O L S TO R E D U C E C L I M AT E - S E N S I T I V E D I S E A S E R I S K S and it was one of the early health impacts to be rigorously invested capacities of indigenous social organizations so that communica- in, with operational programs on global, regional, national, subna- tion and any necessary technologic components of EWS can be tional, and local levels (Kuhn et al. 2005). Many diseases and pests successfully rolled out (IFAD 2009). with major impacts on humans, such as malaria, dengue, cholera, WHO has published specific recommendations regarding the devel- influenza, and locusts, for example, are all widely supported by EWS opment of climate-sensitive disease EWS (see figure 2.2). Although programs. formulated for climate-sensitive disease risks in humans, the similar- Early warning systems are also under development for a number ity to diseases in animals allows considerable transferability of ideas of climate-sensitive animal health impacts. In 2004–09, the first and applications, hastening the speed with which animal EWS can phase of an EWS was implemented by the International Fund for be generated and implemented. Agricultural Development (IFAD) and the World Bank to protect A number of WHO Special Program for Research and Training in herders in Ethiopia. With a goal of strengthening the resilience of Tropical Diseases pilot studies have been in progress in seven cen- the rural poor and increasing their ability to cope with external tral Asian countries, which have followed this scheme. Throughout, shocks (such as the impact of drought on livestock), the project has it has become apparent that proper vulnerability assessment by col- in these initial stages focused on strengthening the institutional laborative national and international agencies is key to the success- ful design of regional and effective transboundary mitigation pro- CASE STUDY grams, and that this stage may take substantially longer than initially IFAD/World Bank Ethiopian Pastoralist Community envisaged, in part because it involves significant in-country capacity Development Project building and training prior to implementation. Some of these ini- The project aims to improve the prospects of achieving sus- tial vulnerabilities and capacities with regard to animal health can tainable livelihoods among herders living in arid and semi-arid potentially be determined through a tool like the OIE Performance lowlands. It seeks to harmonize the development of Ethiopia’s of Veterinary Services, as described in more detail in chapter 3. lowlands and its more fertile highlands, and reduce vulnerabil- ity to drought and the risk of local conflict. In evaluating the potential utility of early warning systems, WHO The first phase of the project (2004–09) was a response to recommends that disease EWS be developed only if the disease is drought and to the need to create sustainable livelihoods for epidemic-prone. That is, “an occurrence in a community or region herders. In partnership with the World Bank, the project estab- lished early warning systems and disaster preparedness plans, of cases of an illness … in excess of normal expectancy.” Outbreaks through a participatory approach to programming, implemen- also qualify under this designation when they are epidemics “lim- tation and monitoring. The objective was to strengthen the re- ited to localized increase in the incidence of a disease, for example, silience of the rural poor and increase their ability to cope with in a village, town or closed institution” (Last 2001). Because external shocks, while making them less vulnerable to drought epidemics and outbreaks differ only in the scale of their effects, and other natural disasters, thus indirectly promoting climate change adaptation. Initial activities included strengthening a climate-sensitive disease EWS will be effective for both (Kuhn the institutional capacity of indigenous social organizations. et al. 2005). Using this criteria, EWS can be targeted to regions and The disaster-preparedness and contingency fund (DPCF) will resources maximized. be created in the second phase, with separate “windows” for early response and disaster-preparedness investment financ- The primary aim of EWS is to predict the occurrence of an epidemic ing. Through the disaster-preparedness strategy and invest- with a sufficient lead time that allows actions to be taken to mitigate ment program (DPSIP) subcomponent, the project will identify its extent and impact. Including risk maps as part of an EWS allows local needs for long-term regional disaster-preparedness and prioritization of the surveillance and control activities, improving mitigation. Under the DPCF, each region will receive DPSIP grants to finance disaster-preparedness investments. their efficiency. For any EWS, it is thus essential to characterize the lead time, the spatial and temporal scale of the predictions, and the Source: IFAD 2009. uncertainties of outputs, and to ensure that adequate contingency R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S C H A P T E R 2 — A C T I O N A B L E TO O L S TO R E D U C E C L I M AT E - S E N S I T I V E D I S E A S E R I S K S 27 FIGURE 2.2: Framework for Developing Early Warning Systems for Climate-Sensitive Diseases (Adapted from: Kuhn et al. 2005) Vulnerability assessment • Evaluate epidemic potential of the disease • Identify geographical location of epidemic-prone population • Identify climatic and non-climatic disease risk factors • Quantify the link between climate variability and epidemics Data requirements Implementation measures Weekly or monthly Develop national and incidence data Early warning and detection components district epidemic response • Seasonal climate forecasts (lead-time in months-low geographical resoultion) plans, define range of Frequently updated data • Monitoring of disease risk factors (lead-time in weeks or months—higher control interventions, on rainfall, temperature, assign clear roles and humidity, stream-flow, geographical resolution) responsibilities vegetation indices • Disease surveillance (lead-time negligible—confirmation of epidemic in process) Regional and national Identify data sources seasonal climate forecasts, and indicators drought and flood surveys Control response Identify case definitions • Assess opportunities for timely vector control and act accordingly and confounders Population migrations and displaced persons • Raise community awareness and call for greater personal protection • Ensure prompt and effective case management Identify key informants Supplementary data (these may be in other (as capacity allows): sectors, e.g. food security, Post-epidemic assessment drought/flood monitoring) Entomological indices • Was the early warning system useful? Carry out cost-effectiveness Parasitological indices • Were the indicators sufficiently sensitive/specific? analysis of timely preventive • Were effective preventive/treatment control opportunities enabled? control and treatment Drug resistance testing • What were the strengths/weakness in control operations? options • Does the epidemic-preparedness plan need to be modified? plans associated with each step of the warning are clearly identified. Rift Valley Fever Those principles are elaborated on in chapter 3. Of the three diseases detailed in this report, RVF provides the best example of an EWS tool. As previously noted, RVF epidemics in East 2.4.2 Early Warning Systems: Knowledge and Applications Africa have been linked with inter-annual climate variability and At present, there is little integration of EWS into disease control deci- above normal rains and floods triggered by ENSO. Coupling climate sion making (Kuhn et al. 2005). Yet there are some positive examples and disease data is thus possible by collection of the right sources. to draw from. An RVF EWS provided decision support to interna- For example, detection of SST and thus the ENSO event itself is the tional organizations during epidemics in 2007 and 2010 in East Africa necessary first step establishing this kind of seasonally determined (Anyamba et al. 2009). Some animal health information systems also early warning system. These data alone, however, do not provide exist at the global scale, such as the FAO Empres-I information sys- spatially detailed information about where the excessive rain will tem (http://empres-i.fao.org/eipws3g/) and the World Animal Health occur within regions, and so looking to historic patterns and identi- Information System of the OIE (http://web.oie.int/wahis/public fying anomalies in rainfall and vegetation (via the NDVI) from satel- .php?page=home), although these systems rely largely on reported lite imagery (available, for example, from NASA/NOAA) can be used confirmed cases, and thus provide little lead time or early warning. to refine spatial predictions (Anyamba et al. 2009). Coordinating In addition to the official notification of disease or infection out- these data with disease sensitivities to these environmental vari- breaks, FAO, OIE, and WHO systematically collect, verify, analyze, and ables can provide public health officials with warnings on multiple respond to information from a variety of sources, including unof- levels: general warnings sent when an ENSO is detected and proxi- ficial media reports and informal networks. In 2006, these heuristics mal early warnings once severity and weather patterns emerge. were combined to launch the Global Early Warning System for Major During the 2006–07 and 2010 outbreaks, a six-week lead time was Animal Diseases, including Zoonoses, or GLEWS, forming a new and provided, which and had the ultimate effect of diminishing overall wide-reaching collaborative. RVF impact. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 28 C H A P T E R 2 — A C T I O N A B L E TO O L S TO R E D U C E C L I M AT E - S E N S I T I V E D I S E A S E R I S K S East Coast Fever VOICES FROM THE FIELD A strong correlation between the sero-prevalence of T. parva Experiences with RVF Early Warning in Kenya in 2010 and the presence of ENSO has been noted in southern Zambia Dr. Peter Ithondeka, Director of Veterinary Services, Kenya (Fandamu et al. 2005, 2006). Although ECF is more an endemic than epidemic disease, the results—if confirmed in other countries and “Decision-making in RVF outbreak cycles is always difficult be- cause it involves balancing the lack of perfect information with regions—could confirm the ENSO correlations and enable better the need to make a decision to prevent losses. If a decision is disease predictions. As of yet, however, there is little literature or taken too early with scant information, the likelihood of taking experience with early warning systems for ECF. a wrong decision is increased and unnecessary costs will result from inappropriate activities. If a decision is taken too late, the opportunity to intervene effectively may be lost, leading to un- 2.5 COMPLEMENTARY NATURE OF SURVEILLANCE mitigated impacts. The decision-maker has to balance the risks SYSTEMS, RISK MAPS, AND DISEASE of over-reacting against those of under-reacting. OUTLOOKS WITHIN EARLY WARNING SYSTEMS In 2010, we received an early warning that due to an ENSO Monitoring and surveillance detect early signs of a growing situation, we could experience abnormal rainfall resulting in epidemic in the field and provide the network for facilitating the floods in high risk areas for RVF. Within the areas identified as early dissemination of recommendations. These disease data are high risk with early warning maps, we had worrying evidence from the field: there was an upsurge in food rot, many areas complemented by climatic data (observed in and transmitted by were flooded, and within those, there were many sites where weather stations or remote sensing modalities) and ancillary data we knew that RVF outbreaks had occurred in the past and that (that is, spatial distribution of livestock and breeds). These three we hence considered as high risk. types of surveillance data can then be used to build reliable risk So, we combined the forecasted risk map with our local knowl- maps and disease outlooks, which can in turn be aggregated to edge to conclude that there was a very high risk of RVF out- establish a base of information for dissemination through early breaks, and that those outbreaks would be concentrated in those areas. So, we took the decision to concentrate our distri- warning systems (see figure 2.3.) In effect, EWS are the overarch- bution of vaccines in those areas, such as to prevent as much ing tool that enables threat alerts early enough for preventative as possible the extent of the outbreak. action; surveillance systems provide the inputs for the visualiza- It is always difficult to retrospectively assess what would have tion and perception of risk; risk maps aid in the discernment and happened if we had taken another decision, but we believed that this was quite successful.” FIGURE 2.3: Different Components of Early Warning Bluetongue Systems and Their Relationship Although several systems have been described in the literature for EARLY WARNING SYSTEMS BT EWS (Giovannini et al. 2004b; Racloz et al. 2006), they are not by Short-term definition truly EWS because they only embody the disease surveil- Monitoring and surveillance data collection lance component, offering little lead time ahead of an epidemic. Analysis and monitoring of wind patterns have, however, been used successfully in the United Kingdom to ensure early detection by tar- Time Risk maps geting surveillance in areas of high risk from introduction of the dis- data synthesis and communication ease from continental Europe, and to define the potential for spread within the country as a whole. In this regard, the lessons learned here may be more useful for long-term disease outlooks through Outlooks data projection the analysis of wind patterns. Long-term R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S C H A P T E R 2 — A C T I O N A B L E TO O L S TO R E D U C E C L I M AT E - S E N S I T I V E D I S E A S E R I S K S 29 communication of these associated threats; and disease outlooks VOICES FROM THE FIELD provide an opportunity for temporal understanding and long- Why Risk Mapping and Early Warning Systems Are Not term planning. Used More Widely Unfortunately, many gaps still exist in the development and imple- Dr. Diarmid Campbell-Lendrum, Team Leader for Climate mentation of each of these tools. Considering them together as Change and Human Health, World Health Organization integrated components under the early warning system framework “There are many scientific and policy rationale for using cli- will optimize their utility and potential for impact, regardless of the mate-driven risk maps and early warning systems in climate- sensitive diseases. Climate is intrinsically variable in space and disease that they address. Preparing these approaches now will pro- time, and the overall future impact of climate change further tect the health, livelihood, and financial futures of those most at risk emphasizes the need for a better understanding of its influ- for generations to come. ence on diseases. In addition, there is a strong demand from the countries themselves for such systems. However, there are today few climate-based early warning systems used to influ- ence disease control decisions. There are many reasons for this. In some instances, climate is one of the many determinants of diseases, and the complex- ity of integrating all factors becomes a real obstacle. In other instances, systems are in place, but address no real stakehold- ers who could convert the system outputs into actions. Other problems can be a lack of added value of the system over the knowledge of the users in the field. I think that in the future, one should pay attention to embed risk mapping and EWS into operational decision support systems, to share lessons and systems with other hazards for which EWS have been developed successfully, and to evaluate the benefit of those systems against criteria that are relevant to decision-makers.” A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R C H A P T E R 3 — I N V E S T M E N T S A N D A P P R O A C H E S F O R E S TA B L I S H I N G E A R LY A C T I O N C L I M AT E - S E N S I T I V E D I S E A S E R I S K R E D U C T I O N TO O L S 31 Chapter 3 INVESTMENTS AND APPROACHES FOR ESTABLISHING EARLY ACTION CLIMATE-SENSITIVE DISEASE RISK REDUCTION TOOLS This chapter proposes investments that will address the under- Proposed Priority Investment Needs lying requirements for establishing climate-sensitive disease The project components in this box have been highlighted risk reduction tools. It contains a synthesis of reviewed litera- as priorities for investment based on their basic utility in the ture and outcomes achieved during an expert-level consulta- development of climate-sensitive disease risk reduction tools. The knowledge infrastructure is the obligatory first stage of tion on climate-sensitive disease risk reduction. Steps to assess investment, and EWS messages is the last. The other project baseline capacities are also included. components can be concurrently enacted throughout an inte- grated process of investment. Key Messages: Phase 1: • Preparing climate-sensitive disease risk reduction tools •Knowledge: Needs assessments and baseline surveys of requires minimum investment in a number of areas: basic capacities of institutions, individuals, and technical knowledge, policy, human resources, information and and physical infrastructures communication technology, and physical building. Phase 2: • Investment in individual project components in each of •ICT: Climate-sensitive disease web-portals inclusive of these infrastructure areas will help build the capacities integrated EWS information, risk maps, disease of countries so that they can effectively implement and outlooks use risk management tools. •ICT: Mapping, GIS, and modeling software • Many of the actions and project components leading •ICT: New and/or integration with current hydro-met to the strengthening of these capacities are interrelated information systems and co-dependent, necessitating investment packages •Human Resources: Workforce trainings (policy that address a portfolio of needs. makers, veterinarians, physicians, environmental • The actions required to bolster these areas are not scientists, communication experts, others) through necessarily specific to any one disease. short courses, workshops, and sponsored advanced • Investment in project components can have degree programs on general climate-sensitive disease co-benefits for a variety of non-disease-related information as well as specialized technical aspects of development needs. the work (such as disease diagnostics, GIS, computer programming) 3.1 REQUIREMENTS FOR EARLY ACTION TOOLS •Policy: Coordinated animal health–human health col- laboration mechanisms through, for example, com- The development and implementation of the tools highlighted in mittees and cross-sectoral working groups at national/ regional levels chapter 2—surveillance systems, risk maps, disease outlooks, and Phase 3: early warning systems—require basic levels of underlying infrastruc- ture. This infrastructure can be deconstructed into five main cat- • ICT: EWS messages disseminated through new media: websites, mobile phones, social media egories: baseline knowledge, policy, human resources, information A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 32 C H A P T E R 3 — I N V E S T M E N T S A N D A P P R O A C H E S F O R E S TA B L I S H I N G E A R LY A C T I O N C L I M AT E - S E N S I T I V E D I S E A S E R I S K R E D U C T I O N TO O L S TABLE 3.1: Climate-Sensitive Disease Investments INFRASTRUCTURE REQUIREMENT INVESTMENT FAMILY PROJECT COMPONENTS REQUIRING INVESTMENT Baseline Knowledge Information Product and Knowledge • Needs assessments and baseline surveys of basic capacities of institutions, individuals, and technical/physical Generation infrastructures • Climate-sensitive disease risk catalogues and impact assessments at national and regional levels • Feasibility studies for risk management tools, such as EWS messaging Policy and Human Resources Institutional Strengthening and • Workforce trainings (policy makers, veterinarians, physicians, environmental scientists, communication Professional Capacity Building experts, others) through short courses, workshops, and sponsored advanced degree programs on general climate-sensitive disease research as well as specialized technical aspects of the work (such as disease diagnostics, disease risk mapping (GIS and spatiotemporal modeling), computer programming) • Environment, disease, and ICT workforce recruitment • Coordinated animal-human health collaboration committees and cross-sectoral working groups at national/ regional levels • Early warning protocols for specific climate-sensitive diseases Community Capacity Building • Climate-sensitive disease and ICT user trainings at local and subnational levels • Community support groups and knowledge exchanges Information and Communication Information Dissemination • Climate-sensitive disease publications disseminated to professional and lay audiences • Climate-sensitive disease and EWS messages to be disseminated through traditional media resources: print, television, radio, community theatre • EWS messages disseminated through new media: websites, mobile phones, social media ICT Capacity Building • Digital climate-sensitive disease libraries at regional/national level • Climate-sensitive disease web-portals inclusive of integrated EWS information, risk maps, disease outlooks • Mapping, GIS, and modeling software • New and/or integrated with current hydro-met information systems • Innovative data collection approaches Physical Building and Construction • New or retrofitting of current facilities to create coordinated animal-human health–environmental data collection and collaboration centers at national/regional levels; to include meeting facilities, high speed Internet, resource libraries, and computers equipped with mapping, modeling, climate, and disease monitoring software • Rapid diagnostic laboratories equipped to process climate-sensitive diseases • ICT networks and communication, and physical buildings (see table  3.1). Each 3.2 REQUIREMENT 1: BASELINE KNOWLEDGE necessitates investment, although the type and amount of invest- Information products are important for enabling understanding ment and the level at which it is targeted will vary by region and on multiple levels and across disciplines. Not only will they help country. This chapter establishes what these underlying require- users stay up to date with the most recent knowledge, they can ments are and describes possible approaches by which investment encourage further engagement in regional contexts and can link needs can be determined and met so that the risk reduction tools practitioners with others around the world. can be widely, rapidly, and effectively adopted. Many of the require- ments depend on other ones, and so it is important to develop a Project Component: Performance of Baseline Surveys, Assessments, range of investment to cultivate a healthy operational environment and Feasibility Studies and Creation of Catalogues for climate-sensitive disease preparedness. The requirements and Achieving baseline knowledge is a fundamental requirement for approaches in this chapter are not specific to any of the diseases the development of any of the risk reduction tools detailed in this detailed throughout the report; where specific disease consider- report. Strong region and nation-specific information products will ations are necessary, they are noted in the text. Although broadly help target investment in the areas where it is needed the most. applicable, it is worth noting that disease prevention efforts should In this regard, a number of initial studies are necessary to set the be focused on the diseases that are most economically and medi- stage for project investment and implementation. Key actions that cally costly to the populations that suffer them, and they should be will facilitate this include the following: conducting needs assess- considered on a country- and region-specific basis. ments and baseline surveys of the basic capacities of institutions, R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S C H A P T E R 3 — I N V E S T M E N T S A N D A P P R O A C H E S F O R E S TA B L I S H I N G E A R LY A C T I O N C L I M AT E - S E N S I T I V E D I S E A S E R I S K R E D U C T I O N TO O L S 33 individuals, and technical and physical infrastructures (for example, Disease Catalogues on veterinary services performance, see box below); creating The creation of climate-sensitive disease catalogues is important climate-sensitive disease catalogues and impact assessments at because it will alert health specialists to the potential threats a re- national and regional levels; and performing feasibility studies for gion is facing and/or those that are likely to increase with climate the proposed risk management tools within current infrastructures. change. The availability of such catalogues will increase overall awareness of climate-sensitive diseases and enable better pre- Surveys paredness and faster response time in the event of an outbreak. Coordinated deployment of surveys and assessments may yield the Determination of the specific disease threats will also enable the best and most unified results. In some cases, current resources can enaction of targeted measures to prevent disease spread, such as be drawn upon to achieve outcomes; in others, unique interviews maintaining vaccines for certain diseases on hand or engaging in and research will need to be conducted. Specific assessment mo- mitigation efforts like spraying for insects, which apply to specific dality will vary by the subject of the survey and will use a number diseases. of methods. Assessing institutional capability, for example, could include an inventory of the number of relevant departments within Feasibility Studies an institution, the number of policies in place that address climate- Conducting feasibility studies presumes there is an underlying sensitive disease, and the history of institutions able to successfully level of infrastructure sufficient to support some degree of risk deal with health threats. Individual assessments might include, on management tool implementation. This method might include the the professional side, tallying the number of trained persons who pilot launch of one of the tools to determine how rapidly and ef- could serve a role in the implementation of risk management tools, ficaciously it could be enacted, providing an applied check on the interviewing them about their availability to engage in this work, results determined in the baseline knowledge assessments. and gauging their perceptions of general success of this initiative. For the community component, interviews could be conducted to In each of the following sections, questions are provided that understand general awareness for these climate-sensitive disease can be used to guide program managers in efforts to establish threats and local and traditional approaches to managing envi- this infrastructure project inception phase. ronmental and disease threats. Regarding technical and physical infrastructures, hydro-meteorological and disease surveillance data 3.3 REQUIREMENTS 2 AND 3: POLICY AND HUMAN can be quantified, ICT capabilities measured, and abundance and RESOURCES integrity of scientific facilities assessed. The two requirements in the institutional strengthening and capaci- ty building category are difficult to disentangle, as policy relies upon Highlighted Approach human resources to be enacted, while, conversely, human resources The OIE PVS Pathway (PVS stands for Performance of Veterinary often require policy for organization and structure. Investment in Services) is an internationally recognized system of measure- these two areas will ensure a nourishing socio-political environment ment and evaluation of national veterinary services based within which the risk reduction tools can develop and thrive. The on OIE standards, helping countries identify their deficien- specific approaches incorporate a combination of capacity building cies, prescribe solutions, and undertake strategic actions. Assessments with these tools have been completed in around and policy actions that taken collectively will create strong institu- 120 countries, comparing performance to international stan- tional and human capacity to support and use the risk management dards of veterinary service quality and guiding assistance with tools. investment decision making. The results of these assessments provide comprehensive and objective information for donors Project Component: Workforce Capacity Building and Recruitment and partners willing to help countries strengthen their animal Risk reduction tools are useless without a capable workforce health systems efficiently (OIE 2014). to employ them. Once the range of stakeholders is identified, A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 34 C H A P T E R 3 — I N V E S T M E N T S A N D A P P R O A C H E S F O R E S TA B L I S H I N G E A R LY A C T I O N C L I M AT E - S E N S I T I V E D I S E A S E R I S K R E D U C T I O N TO O L S specialized and directed training programs can be developed to INCEPTION PHASE QUESTIONS: WORKFORCE CAPACITY BUILDING train workers at all levels, including pastoralist farmers, veterinar- What are the current capacities of the workforce in each sector? Which sectors need training? ians, physicians, health care extension workers, environmental Which sectors have priority training needs? specialists, ICT specialists, epidemiologists, and government Which sector training needs correspond with other project components (for example, technical infrastructure)? officials involved in decision making. The nature of this train- What kind of training do professionals in each sector need? ing will coincide with the technical requirements in the subse- Where can individuals achieve the training/who will provide it? What resources are available for the training? quent section, so that the technology and user capabilities will Are there on-going training modalities that can integrate climate-sensitive disease trainings? co-develop. Project Component: Workforce Recruitment The first step in this process, as noted in the previous section, would Another way to improve the overall quality of the workforce is to be to conduct baseline surveys that determine the core competen- develop incentives for training and recruiting technical specialists. cies of different types of workers. Tiered and directed training can Attractive salary packages and healthy programmatic budgets for then be optimally targeted to those most in need. developing innovative climate change and disease approaches The approaches to this are many. Client country-driven train- can both harness top talent within a country as well as lure leading ing programs can be established where capacity is greatest, professionals to countries and regions and encourage collaboration fostering ownership and autonomy over climate-sensitive dis- among local and international workforces. The development of ease initiatives. Intergovernmental agencies (WHO, OIE, the UN in-country talent will raise the profile for the issue nationally, while Development Programme, the WMO) and development banks the attraction of external experts can raise the profile of climate- can offer training, providing the advantage of institutional cen- sensitive diseases on an international level and create opportunities tralization and regional congruency in training received. Training for expert and political engagement. Joint programming through modalities currently exist on climate change and disease at both other universities and institutions, staff exchanges, professional WHO and WMO, which have recently partnered to address cli- fellowships, and other professional development programs that mate change and health threats. Third-party experts can also be establish international knowledge exchange can also be effective. called upon to provide the expertise and training; for example, the International Research Institute for Climate and Society at INCEPTION PHASE QUESTIONS: WORKFORCE RECRUITMENT Columbia University provides training modules to build a range What types of professionals are lacking? From where can these professionals be attracted? of climate change and disease technical competencies, such as What kinds of resources will it take to attract the right professionals? Once recruited, where will the professionals work? access to environmental and climate data, epidemiologic and en- vironmental data integration, disease transmission and climatic Project Component: Coordinated Animal-Human Health analysis, and disease outbreak forecasting. Collaboration Committees and Cross-Sectoral Working Groups at National/Regional Levels The content of these trainings will depend largely on the capa- Establishing overall governance and accountability mechanisms bilities and needs of the training audience. Yet there are certain is imperative to ensuring the successful implementation and con- must-have training needs that will ensure the effective use of tinuation of a climate-sensitive disease initiative. Without high-level the climate-sensitive disease tools. Topics for these include basic government support for policies on climate risk management, it will understanding of how climate change and disease are related, be difficult to achieve the resources necessary to keep interest and using risk maps and early warning systems, rapidly responding to action alive. Obviously, the fact that a disease is climate-sensitive disease outbreaks, integrating health and environmental services, does not automatically justify the implementation of national sur- and communicating health and climate concepts. veillance and control programs. Conducting a prioritization exercise R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S C H A P T E R 3 — I N V E S T M E N T S A N D A P P R O A C H E S F O R E S TA B L I S H I N G E A R LY A C T I O N C L I M AT E - S E N S I T I V E D I S E A S E R I S K R E D U C T I O N TO O L S 35 at the national or regional level would provide solid socioeconomic documents such as EWS protocols, needs assessments, and/or grounds to advocate investment in major diseases, whether they contingency plans; and overseeing training and education efforts. are climate-sensitive or not. Some regional committees are also Regular communications and interactions with regional bodies well positioned to address transboundary diseases. All the major and international technical organizations in charge of coordinating diseases of climate-sensitive importance fall under this category. and assisting countries on this subject (such as WHO, FAO, OIE) can Targeting them collectively in a way that transcends national be ensured through provision of technical advice, equipment, and boundaries, rather than through a piecemeal, country-by-country software. approach, offers the best chance to overcome these diseases. Approaches to developing these kinds of committees will require To address this critical aspect of coordination and collaboration, a investment in dedicated human resources—in other words, indi- first step should be to understand the various existing formal and viduals who are positioned and capable of bringing together key informal relevant collaborative mechanisms at national and regional members of leadership communities to ensure deliverance of these levels that could serve as entry points to further build upon and integrated networks and committees. Further, there must be articu- consolidate performing ones. lated reasons to participate that incentivize leadership and buy-in. Additional resources will then be required to maintain the commit- If the establishment of a formal committee would be deemed most tees through general staffing and logistical needs. appropriate, the next steps could include convening and establish- ing a multi-sectoral national or regional steering committee for Information on recommended actions (prevention, control, surveil- control of climate-sensitive disease. This should include, at a mini- lance, spreading the word, increased awareness, and so on) should mum, representatives from decision-making bodies (the Ministries always accompany information communicated to stakeholders, and of Health, Agriculture, Environment/Meteorological Agencies); na- the resource needed to follow recommended actions should be in tional and international research institutions and technical agencies place. Information should be available on the benefit of mitigation and universities; end-users of climate-sensitive disease plans, such actions in terms of relevant criteria (prevention of productivity or as farmers; and the people who work with these groups, such as trade-related losses, number of cases prevented, and so on). veterinarians and community communication specialists. To avoid overlap and inefficiency, this may make use of existing commit- INCEPTION PHASE QUESTIONS: COLLABORATIVE HEALTH COLLABORATION tees, such as the task team for implementation of the Libreville COMMITTEES What are the existing mechanisms of cross-sectoral collaboration at subnational, national, and Declaration on Health and Environment in Africa (Libreville 2008), regional levels, if any, that can be further developed to address climate-sensitive disease issues? any committees devised during and after avian influenza outbreaks, If a committee should be developed, are there any committees in place that could serve as a model? and any standing committees that are currently addressing One Which agencies are active in the relevant fields and should be included on the committee? What professional level should the committee represent? Health challenges. The committee should also include a link to na- Should the committee be regional or national? tional bodies covering related functions, such as national climate What tasks should the committee be responsible for and what will the outputs look like? How often and where should the committee meet? change committees and strategic planning committees responsible What will be the relationship of this committee to other committees and international for human and animal health. If pre-existing institutions are capable organizations? What power will this committee have? of taking on a climate-health element, they should be used first. This committee would serve multiple functions including, but not Project Component: Early Warning Protocols for Climate-Sensitive limited to, being an entry point for international climate-sensitive Diseases disease efforts (along with local authorities); coordinating interven- The development of early warning protocols provides the best hope tions aimed at responding to any kind of climate-sensitive disease for unified and consistent responses by policy makers and practi- related event; developing and implementing pre-operational tioners on a range of levels. Given the complexity of information A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 36 C H A P T E R 3 — I N V E S T M E N T S A N D A P P R O A C H E S F O R E S TA B L I S H I N G E A R LY A C T I O N C L I M AT E - S E N S I T I V E D I S E A S E R I S K R E D U C T I O N TO O L S coming into decision makers and the need for rapid action, it is health and production services, and human health and well-being. imperative to carefully design protocols that outline what actions With the needs assessment results in hand, program implementers need to be taken and by whom. Best practices for the development can begin designing outreach and education efforts that meet the of protocols of this sort can be learned from disaster response col- demands of individual communities. leagues who similarly have long-standing and constantly evolving As with the professional training, community efforts will need to fo- protocols to meet their demands. Architects of these protocols cus on key themes, namely basic understanding of climate-sensitive should be a combination of higher-level decision makers who can diseases and why they are important to them and their livelihoods, ensure the protocols will make it into the appropriate hands, as well measures that can be taken to mitigate disease risk (such as elimi- as on-the-ground practitioners who are familiar with the capacities nating standing water to reduce vector breeding grounds and vac- of those who will be responsible for implementing them. The “nuts cinating animals), what EWS are and how they can use them, and and bolts” of early warning protocols—data flow, interdisciplinary avenues for engagement with centralized authorities. Resources partnerships, calibration, pilot-testing—are imperative to identify for community outreach efforts such as these currently exist within early on so that the most comprehensive systems can be developed. similar institutions as those detailed in the professional training sec- tion. NGOs such as Vétérinaires Sans Frontières and international INCEPTION PHASE QUESTIONS: EARLY WARNING PROTOCOLS organizations such as UNICEF also have capabilities in dealing with What diseases will these protocols cover? community outreach and communication. Contracting NGOs or Who will be responsible for generating and updating the protocols? Who will be responsible for enacting and implementing the protocols? third-party consultancies will be an important part of this process Who will be the beneficiaries of the protocols? and will require skilled project managers with an awareness of When should the protocols be used? Should the protocols be nationally/regionally implemented and should they be universal? which actors are best positioned to deliver results. What kinds of response protocols currently exist? INCEPTION PHASE QUESTIONS: COMMUNITY TRAININGS Project Component: Community Climate-Sensitive Disease and ICT What is the current community understanding of climate-sensitive diseases? User Trainings at Local and Subnational Levels What is the current capacity of the community for understanding abstract concepts like climate-sensitive disease? Similar to professional training and education, the benefits of a well- What is the literacy of the current population? What languages do members of the community speak? informed, non-specialist community are many. Outreach and un- How can messages be best targeted to reach most community members? derstanding campaigns that engage the public can help them pre- Which members of the community are most important to target? How does gender or age affect the communication strategy? pare for and respond to disease outbreaks, build trust, and enable How much trust do community members have for government authorities? community members to connect with appropriate authorities to Who can lead these trainings? What community resources exist in place for trainings? hasten response time and effectively reduce disease risk. In-school What new resources will be required? courses, community workshops, educational radio and television Where does the community typically get their information? What kind of ICT access and competencies do members of the community have? programming, and community theater can all be used as avenues for communication. Project Component: Community Support Groups and Knowledge Conducting a participatory needs assessment can be a helpful first Exchanges step, as it will identify community interests and priorities and acute Given the complex concepts associated with climate-sensitive dis- needs and the capabilities of the end users’ communities. To fully eases and the introduction of new technology and modalities for earn the trust of a community on something as abstract as “reduc- reducing their risk, appropriate forums for addressing the issues that ing climate-sensitive disease risks,” program implementers will have arise around them will need to be established. Building community to prove they are there for the community in less abstract, more support groups or tapping already existing support groups will meaningful ways—for example, by describing outreach efforts in aid in this effort, improve usability of the risk reduction tools, and terms locally understood, such as water/land problems, animal hasten the speed at which knowledge and early warning messages R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S C H A P T E R 3 — I N V E S T M E N T S A N D A P P R O A C H E S F O R E S TA B L I S H I N G E A R LY A C T I O N C L I M AT E - S E N S I T I V E D I S E A S E R I S K R E D U C T I O N TO O L S 37 can be communicated. Additionally, these support groups will Project Component: Climate-Sensitive Disease Information enable two-way communication—allowing for messages and in- Dissemination and EWS Messaging formation to come from centralized authorities while providing an This component will dovetail closely with the professional commu- opportunity for community members to convey experiences and nity capacity building components listed above. Rather than train- understandings so that improvements can be made and success ings and workshops, this component will focus on the generation monitored. and dissemination of the information literature itself: research and “how-to” publications to professional audiences, and pamphlets and Outlining the purpose and aims of each group is key and will maxi- lay resources to the community. To ensure the broadest reach, dis- mize the usefulness both to group members and to the project in- semination of up-to-date literature and EWS messages will rely both terlocutors who work with them. Different groups will be able to of- on human resources and well-functioning ICT systems. Information fer different kinds of information that can be used in risk reduction. can be channeled through traditional means such as town halls, Farmer support groups, for example, will harness the resources of a school training, television, radio, and theater, as well as through the demographic that is highly knowledgeable about their animals, the new technologic infrastructures that are detailed in the next section nuances of their local environment, and the interplay of disease and (cell phones, websites, and social media). environmental change. Much in the same way, medical profession- als and extenders can contribute understandings of demographic INCEPTION PHASE QUESTIONS: INFORMATION DISSEMINATION and disease outbreak that could be crucial in the development of To whom does this information need to be disseminated? early warning systems in humans. The community support groups What are the best avenues for dissemination? can therefore both benefit from and contribute knowledge to cen- How can new technologies best be used to disseminate information? How do the target audiences typically consume information? tralized agencies. What technologies need to be in place in order for the best dissemination practices to occur? How do centralized authorities currently disseminate information? Can any of this information be disseminated through crowd-sourcing equivalents to dissemination? INCEPTION PHASE QUESTIONS: COMMUNITY SUPPORT GROUPS Are there current community support groups or knowledge exchanges in place? How do farmers typically communicate with one another? How do farmers typically communicate with governing authorities? Project Component: Climate-Sensitive Disease Libraries What is the literacy rate of pastoralists? Data collected in the field and properly transferred should be or- What kind of ICT access do pastoralists have? ganized in a database management system, where data from dif- ferent origin and type can be centralized, organized according to 3.4 REQUIREMENT 4: INFORMATION AND international standards in meta-data, checked for integrity, made COMMUNICATION accessible for analysis, and stored in backups on a regular basis. The Information and communication infrastructure is critical for both data will need to be properly geo-referenced and uniform so they harnessing environmental and disease data resources and making can be applied seamlessly across databases for analyses with data them available so that they may be used in the development and from other sectors, such as data on herd population distribution, use of disease risk management tools. There are two components other disease risk maps, socioeconomic and welfare indicators, to this: the action of conveying communication and the technical administrative units at varying decision-making levels, and climate hardware and software capacities that underlie it. In most countries information. The IT should allow remote access to data through the this will exist to some degree. The specialized functions required by network to facilitate analysis and modeling by local scientists. The the tools outlined in this report will require additional investment. IT should also allow, through a sufficient bandwidth, easy access to It is important to note that “more data” is not always the solution. international sources of data and scientific journals. Then the infor- Translating, analyzing, interpreting, and using the data requires sig- mation must be prepared so that it is most useful to the recipient nificant investment as well—investment without which the better audiences, such as policy and decision makers; this may mean it is collection of data will be ineffectual. summarized according to decision-making administrative units. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 38 C H A P T E R 3 — I N V E S T M E N T S A N D A P P R O A C H E S F O R E S TA B L I S H I N G E A R LY A C T I O N C L I M AT E - S E N S I T I V E D I S E A S E R I S K R E D U C T I O N TO O L S It is worth noting that there are, however, drawbacks to open- Examples of Data Partners source data. Ownership, need for historical records, capability, server maintenance costs, and ethical oversight can slow or halt the International Organizations: development of such systems. National Health Information systems Food and Agriculture Organization (FAO) in many countries offer lessons to deal with these challenges that Global Early Warning Systems (GLEWS – OIE/FAO/WHO) International Fund for Agricultural Development (IFAD) must be taken into consideration before launching this component International Livestock Research Institute (ILRI) of a project. World Organisation for Animal Health (OIE) World Health Organization (WHO) INCEPTION PHASE QUESTIONS: CLIMATE-SENSITIVE DISEASE LIBRARIES Who should be responsible for collecting and collating the data? Governmental and Nongovernmental Agencies, Academia, What system or repository will be used to store the data? Thematic Groups, and Institutions: What institution will be responsible for storing the data? How will these libraries be made accessible to users? USAID PREDICT How can these libraries best be kept up to date? Adapting Livestock to Climate Change, Collaborative Research How will these libraries integrate with other libraries and modalities in a region and globally? What data sources can be drawn from currently existing resources to create these libraries? Support Program at Colorado State University Climate Change, Agriculture, and Food Security at CGIAR GALVMed Project Component: Climate-Sensitive Disease Web-Portals Global Initiative for Food Systems Leadership at University of It is important to develop a web-based interface with capacity to Minnesota support weekly, expert-issued disease risk bulletins inclusive of rel- Regional Center for Mapping of Resources for Development evant information for climate-sensitive diseases relevant to a given (RCMRD) Trust in Animals and Food Safety (TAFS) Forum country X (weather forecasts, satellite images, monitoring products, Climate, livestock, and disease researchers risk models, and so on). Interconnectivity and scalability of this interface is of paramount importance, as it will ideally need to be deployed in multiple countries within a given region. The Famine The modalities through which these data can be transmitted will Early Warning System Network developed for the U.S. Agency for depend on the type of data that is flowing. Short messages and International Development is an example. Software engineers and warnings, for example, can be transmitted through basic commu- local users will need to be trained so that they can effectively use nication technologies like mobile phones. Complex disease profiles and maintain such a system. If the capacity and the data exist, which and climate data, however, will require more sophisticated software they may well in many regions, it will be important to acquire, assess systems like those described earlier. the quality, and use these data. In many instances, there is sufficient existing data on environment INCEPTION PHASE QUESTIONS: CLIMATE-SENSITIVE DISEASE WEB-PORTALS and diseases to inform the development of each of our proposed Who will be responsible for moderating and maintaining these portals? tools. Global research studies and agencies have been collecting dis- Who will have access to them? How can access be assured for a wide user audience? ease and climate data for years and in many cases have made the data What technology will be required to access the portals? freely available to the public; the unfortunate caveat is that the data Can the portals be accessed with mobile technology? What are the software/hardware needs of the server? are not universally accessible due to existing policies (or lack thereof ) What are the software/hardware needs of the users? and structures that restrict their availability in the public domain. How will information that passes through this portal be regulated and ensured accurate? Developing these kinds of data-sharing modalities requires first the engagement of global climate and disease communities. Those that Project Component: Mapping, GIS, and Modeling Software have the data are the gatekeepers to the information; achieving their Risk mapping will take many forms and serve a multiplicity of pur- buy-in is essential to creating networks where data can freely flow. poses. As detailed in chapter 2, maps will be developed that depict R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S C H A P T E R 3 — I N V E S T M E N T S A N D A P P R O A C H E S F O R E S TA B L I S H I N G E A R LY A C T I O N C L I M AT E - S E N S I T I V E D I S E A S E R I S K R E D U C T I O N TO O L S 39 potential risk areas, health care providers, health facilities, commu- conducted to ensure the right kinds of data for the right kinds of nity capacities, disease incidence, environmental variables, and oth- users are collected and integrated. ers. Before this can occur, countries must acquire appropriate soft- Currently, there are World Bank investments, such as the Agriculture ware to map and integrate these inputs. This can be accomplished Risk Management Information System, that are integrating this in- through a combination of regional, national, and local efforts so that formation. One approach to the applied utility of hydro-met services the most comprehensive visual depictions can be delivered. to climate-sensitive diseases is to piggyback on the already estab- In recent years, the development of open-source software has lished services generated under this project. Doing so would save provided a wealth of computer programs that can be used free of time and resources while facilitating integration of related efforts. charge to store, manage, and process spatial data (such as MySQL: database; Quantum GIS: Geographical Information System; R: statis- INCEPTION PHASE QUESTIONS: HYDRO-MET DATA SERVICES tical analysis and modeling). These software are however somewhat What hydro-met services currently exist in the region/country? difficult to use for non-experts, and several software and hardware Who has access to these services? Is there currently any integration of these services with health/veterinary/agriculture services? solutions are available off-the-shelf for collecting and mapping What is the current physical infrastructure to support hydro-met services? vector-borne diseases field data, such as VecMap (ESA/Avia-GIS What is the current software/hardware infrastructure to support hydro-met data services? 2012) or EpiCollect (Aanensen et al. 2009), including the possibil- ity to use these input data to model and predict the distribution Project Component: Innovative Data Collection Approaches of vectors based on a sample dataset (ESA/Avia-GIS 2012). In order Recent advances in technology and geo-positioning—many recent to maximize usability of these modalities and safeguard intellectual mobile phones now have a GPS chip—has resulted in easier data property rights, data sharing protocols and agreements between collection. This applies to epidemiological data where the avail- the producers and users and spatial data must be in place and ability of rapid lab kits, or even biosensors in the future, may allow enforceable. One approach to this would be embedding a digital molecular diagnostics to be carried out in the field and epidemio- object identifier with data repository systems so that it can be cited logical data to become easier to obtain. as a scientific publication and encourage the producers of data to share their data sets and keep the corresponding credit (for exam- INCEPTION PHASE QUESTIONS: INNOVATIVE DATA COLLECTION ple, http://datacite.org). What are the current innovation capabilities in a given country/region? What are the technical capabilities to support innovative approaches? Who are the potential users and actors of these collection practices? INCEPTION PHASE QUESTIONS: MAPPING, GIS, MODELING SOFTWARE What are the software/hardware requirements for hosts/developers of this information? What are the software/hardware requirements for users of this information? 3.5 REQUIREMENT 5: PHYSICAL INFRASTRUCTURE What institutions will be responsible for hosting/moderating these needs? How will these resources be kept up to date? Developing new physical infrastructure or re-appropriating cur- What kinds of personnel needs are required to maintain these systems? rent infrastructure is an important step in building a capacity and Project Component: Hydro-Met Data Services community of practice. Permanent or semi-permanent and mobile spaces establish a forum for a range of social, operational, and Hydro-met services are those that provide meteorological informa- research-based activities that can contribute to disease risk reduction. tion to users. In many cases, these will be pre-established at a coun- try level. Attaining targeted and needed information, however, will Project Component: Coordinated Human and Animal Health require integration of these services with disease data so that the Collaboration Centers most relevant values can be achieved. For example, many diseases Enhancing collaboration among disciplines is of fundamental im- are susceptible to overnight temperatures and relative humidity; portance to reducing climate-sensitive disease risks. Eco-climatic assuring variables are delivered is essential to successful delivery of conditions that favor climate-sensitive livestock disease often the risk management tools. Sufficient background research must be also favor diseases directly affecting people. For example, climatic A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 40 C H A P T E R 3 — I N V E S T M E N T S A N D A P P R O A C H E S F O R E S TA B L I S H I N G E A R LY A C T I O N C L I M AT E - S E N S I T I V E D I S E A S E R I S K R E D U C T I O N TO O L S conditions favoring RVF also tend to support bursts in malaria vector The OIE PVS Pathway evaluations reports offer a roadmap to populations, and RVF epidemics have often coincided with malaria strengthen both this surveillance capacity and the laboratory ca- outbreaks. There is much mutual benefit in sharing information on pacities to better diagnose diseases. Modernized laboratory facili- risk mapping, early warning systems, and disease outlooks of CSD-IS ties, the provision of equipment and consumables, and concurrent with other sectors. Conversely, the cost of entomological surveil- training on sample taking, packing, and sending outbreak investi- lance or maintaining meteorological ground observation networks gation and laboratory techniques will greatly improve the overall can be shared among sectors. capacity of the biological components of climate-sensitive disease reduction. One approach to fostering this work and dialogue is through the establishment or strengthening of current One Health partnerships INCEPTION PHASE QUESTIONS: RAPID DIAGNOSTIC CENTERS or collaboration centers that bridge the gap between animal, hu- What are the best locations for these centers? man, and environmental health. Establishing centers at regional Can existing facilities be retrofitted or is there a need for new centers? How big should they be? or national levels asserts the recognition that collaborative health What kinds of spaces should they include? systems are important and effectively offers recurrent political Do they need to be of a particular biosafety level? What kinds of technical capacities should they include? support for the issue if well-executed and delivering good results. What kinds of building materials are needed for them? Functionally, they offer a forum for collaboration and education Where can these building materials be sourced? What are the human resource needs required to build these centers? within which many stakeholders can engage. Purpose-built physi- What are the human resource needs required to staff the centers? cal space dedicated to collaborative health systems work will enable What kind of environmental impact will these centers have? the intellectual exchange necessary to launch the necessary cross- Project Component: ICT Networks disciplinary collaborations. This effort, however, will necessarily need to coincide with capacity building of human and policy resources so Network availability—both mobile and broadband—will affect the that there will be sufficient support and staffing. ultimate users of any early warning system. Verifying the presence of networks, the number and types of users, and average costs for INCEPTION PHASE QUESTIONS: COLLABORATIVE HEALTH CENTERS network access can determine the ultimate usage and success of What are the best locations for these centers? a system. Where networks are not available, investment can be an Can existing facilities be retrofitted or is there a need for new centers? How big should they be? important first step in establishing the absolute basis upon which What kinds of spaces should they include? an entire EWS can be built, as it will enable the exchange of raw What kinds of technical capacities should they include? What kinds of building materials are needed for them? meteorological and disease data as well as the related messages Where can these building materials be sourced? that can warn of outbreaks and epidemics. If network connections What are the human resource needs required to build these centers? What are the human resource needs required to staff the centers? are non-existent or weak, farmers or extension agents will have to What kind of environmental impact will these centers have? travel to centralized locations to receive data and messages, delay- ing overall responsiveness and increasing disease risk. Project Component: Rapid Diagnostic Laboratories The construction of new or the enhancement of existing rapid di- agnostic laboratories will increase the capacity for vector/pathogen INCEPTION PHASE QUESTIONS: ICT NETWORKS surveillance and characterization. Currently, many countries lack What is the present telecommunication and mobile coverage in the country or region? the ability to identify vectors associated with particular diseases Are the telecommunication networks reliable? and also the pathogens that they transmit. In-country laboratories What is the level of broadband penetration? What is the speed of data transfer through the network? will facilitate cataloguing of vector and disease ecosystems, so that Does the region have reliable access to electricity? monitoring and surveillance can be done accurately. Additionally, Are there alternative sources for energy generation? How much of the population has mobile or broadband coverage and devices to access it? they enable important measurements that can hasten the speed of Do farmers have access to networks? rapid response mechanisms, thus reducing overall exposures. What are the average usage costs? R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S C H A P T E R 3 — I N V E S T M E N T S A N D A P P R O A C H E S F O R E S TA B L I S H I N G E A R LY A C T I O N C L I M AT E - S E N S I T I V E D I S E A S E R I S K R E D U C T I O N TO O L S 41 3.6 CO-BENEFITS OF IMPROVING Climate change and disease, deforestation and biodiversity loss, INFRASTRUCTURE FOR CLIMATE-SENSITIVE ocean degradation and depleted fisheries: each is an example of DISEASE RISK REDUCTION TOOLS how humans are affecting the global environment in profound Improving basic infrastructures for climate-sensitive disease risk ways, which then in turn affects the collective human ability to live reduction tools can lead to co-benefits for a number of other related and thrive. Ours is an era of complex global challenges that tran- sectoral development goals. This is due in large part to the basic sect geography and sector. Siloed solutions are no long sufficient. requirements that these risk reduction approaches target and the Diverse voice and expertise brought together through partnership breadth of sectors that they incorporate. Examples of potential and shared understanding is the only path forward. Recognizing co-benefits include: this and preparing to act, we can then move at scale and speed § Improved agricultural management systems. toward healthier and more sustainable futures for all. § Strengthened hydro-meteorological services. § Improved livestock productivity and outputs due to healthier stocks. Spotlighting Other World Bank Resources § Reduced zoonotic diseases impact on human health. Best practices for any of these tools and requirements need § Job creation and increased per capita GDP. not be exclusively derived from this paper. There is consider- § Decreased long-term veterinary and human health costs. able ongoing World Bank work that incorporates many of these tools in separate capacities. Searching World Bank da- § Increased transboundary disease management. tabases for any of the key terms will yield resources that can § Sustainable agricultural practices. contribute to any component of a climate-sensitive disease § Improved disaster preparedness. risk reduction program. § Improved veterinary services delivery to farmers. § Improved surveillance and control of vector-borne diseases that affect humans. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R G LO S S A R Y 43 GLOSSARY Active surveillance: Method by which special effort is expended Disease catalogue: Database containing information about type, to discover disease cases, for example through surveying and biologic profile, and incidence of endemic and epidemic diseases. searches. Includes purposeful gathering of information. Cf. passive Disease outlook: Reports and collections of data that provide long- surveillance. term projection of disease trends for control and mitigation efforts. Biological model: Hypothesis-driven, mathematical model based Early warning system (EWS): Comprehensive set of information on a detailed knowledge of the actual processes underlying the and actions that alert decision makers of impending harm; inclusive presence of a disease or its vector. of surveillance; aims to provide short- or mid-term disease forecast- ing so that appropriate interventions and mitigation efforts can Bluetongue (BT): Vector-borne viral disease that affects primarily reduce the impact of an epidemic. sheep, occasionally goats and deer, and cattle; transmitted by vari- ous Culicoides species of biting midge and can result in severe clini- East Coast fever (ECF): Vector-borne cattle disease endemic to cal symptoms, sometimes leading to death. regions of southern Sudan to South Africa and west to eastern Democratic Republic of Congo; transmitted by several species of Climate: In a narrow sense is usually defined as the “average Ixodid ticks; caused by the parasite Theileria parva, one of six species weather,” but more rigorously as the statistical description in terms of Theileria that infect cattle. of the mean and variability of relevant quantities over a period of time ranging from months to thousands or millions of years. These Endemic: Situation in which a disease is present or established in a quantities are most often surface variables such as temperature, country or area over consecutive time periods. precipitation, and wind (IPCC 2007). Epidemic: Situation in which new cases of a particular disease in Climate change: Any change in climate over time, whether due to a given population during a given period are significantly higher natural variability or as a result of human activity (IPCC 2007). than baseline. Hydro-meteorological data: Data that focus on water and associ- Climate-sensitive disease: A disease whose incidence or trans- ated energy in the atmosphere. mission is affected, positively or negatively, by climate. Normalized Difference Vegetation Index (NDVI): Numerical Climate variability: Variations in the mean state and other statis- indicator that uses visible (VIS) and near infrared bands (NIR) of the tics (such as standard deviations, statistics of extremes, and so on) electromagnetic spectrum to assess whether observed target con- of the  climate  on all temporal and spatial scales beyond that of tains green vegetation; as defined by (NIR – VIS)/(NIR + VIS). individual weather events. Variability may be due to natural internal processes within the climate system (internal variability) or to varia- Outbreak: Occurrence of one or more animals infected by a tions in natural or anthropogenic external forcing (external variabil- pathogenic agent in a group sharing a common environment (for ity). See also climate change (IPCC 2007). example, a farm or village). A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 44 G LO S S A R Y Passive surveillance: Method by which disease cases are uncov- Statistical model: Mathematical model that uses pattern match- ered through routine report. No special effort is extended to dis- ing procedures based on known presence of a disease (or vector) cover cases. Cf. active surveillance. over space and/or time. Rift Valley fever (RVF): Vector-borne viral zoonosis transmitted by Theileriosis: Parasitic disease caused by any species of genus mosquitoes that primarily affects animals, though sometimes also Theileria; can infect humans and animals. humans; transmitted by a broad range of mosquitoes, although cer- Transmission: Passing of an infectious disease from one infected tain Aedes species can act as reservoirs during inter-epidemic years. individual or group to another. In animals, it primarily affects sheep, cattle, goats, camels, and wild ruminants, resulting in high rates of abortion and neonatal mortality. Vector-borne disease: Infectious disease transmitted from one host to another by an insect or any other living carrier. Risk map: Application of data to a visual media that facilitates the communication of disease threats. Zoonosis: Infectious disease transmissible from animals to humans. R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S REFERENCES 45 REFERENCES Aanensen, David M., Derek M. Huntley, Edward J. Feil, Fada’a al-Own, Arnell, N. W., M. J. L. Livermore, S. Kovats, P. E. Levy, R. Nicholls, M. L. Parry, and Brian G. Spratt. 2009. “EpiCollect: Linking Smartphones to and S. R. Gaffin. 2004. “Climate and Socio-economic Scenarios for Web Applications for Epidemiology, Ecology and Community Data Global-scale Climate Change Impacts Assessments: Characterising Collection.” PLoS ONE 4 (9): e6968. doi:10.1371/journal.pone.0006968. the SRES Storylines.” Global Environmental Change 14 (1): 3–20. doi:10.1016/j.gloenvcha.2003.10.004. Abd el-Rahim, I. H., U. Abd el-Hakim, and M. Hussein. 1999. “An Epizootic of Rift Valley Fever in Egypt in 1997.” Revue Scientifique et Technique de Arthur, R. R., M. S. el-Sharkawy, S. E. Cope, B. A. Botros, S. Oun, J. C. Morrill, L’Office International des Epizooties 18 (3): 741–48. R. E. Shope, R. G. Hibbs, M. A. Darwish, and I. Z. Imam. 1993. “Recurrence of Rift Valley Fever in Egypt.” Lancet 342 (8880): 1149–50. Abdo-Salem, S., G. Gerbier, P. Bonnet, M. Al-Qadasi, A. Tran, E. Thiry, G. Al-Eryni, and F. Roger. 2006. “Descriptive and Spatial Epidemiology Azhar, M., Ade S. Lubis, Elly Sawitri Siregar, Robyn G. Alders, Eric Brum, of Rift Valley Fever Outbreak in Yemen 2000–2001.” Annals of the James McGrane, Ian Morgan, and Peter Roeder. 2010. “Participatory New York Academy of Sciences 1081 (October): 240–42. doi:10.1196/ Disease Surveillance and Response in Indonesia: Strengthening annals.1373.028. Veterinary Services and Empowering Communities to Prevent and Control Highly Pathogenic Avian Influenza.” Avian Diseases 54 Acevedo, P., F. Ruiz-Fons, R. Estrada, A. L. Márquez, M. A. Miranda, C. Gortázar, (1 Suppl): 749–53. and J. Lucientes. 2010. “A Broad Assessment of Factors Determining Culicoides Imicola Abundance: Modelling the Present and Forecasting Barrett, R., C. W. Kuzawa, T. McDade, and G. J. Armelagos. 1998. “Emerging Its Future in Climate Change Scenarios.” PLoS ONE 5 (12). and Re-emerging Infectious Diseases: The Third Epidemiologic Ahmad, K. 2000. “More Deaths from Rift Valley Fever in Saudi Arabia and Transition.” Annual Review of Anthropology: 247–71. Yemen.” Lancet 356 (9239): 1422. doi:10.1016/S0140-6736(05)74068-X. Baylis, M., H. Bouayoune, J. Touti, and H. El Hasnaoui. 1998. “Use of Climatic Allepuz, A., I. García-Bocanegra, S. Napp, J. Casal, A. Arenas, M. Saez, Data and Satellite Imagery to Model the Abundance of Culicoides and M. A. González. 2010. “Monitoring Bluetongue Disease (BTV-1) Imicola, the Vector of African Horse Sickness Virus, in Morocco.” Epidemic in Southern Spain During 2007.” Preventive Veterinary Medical and Veterinary Entomology 12 (3): 255–66. Medicine 96 (3–4): 263–71. Baylis, M., and A. K. Githeko. 2006. “The Effects of Climate Change on Antia, R., R. R. Regoes, J. C. Koella, and C. T. Bergstrom. 2003. “The Role of Infectious Diseases of Animals.” Report for the Foresight Project on Evolution in the Emergence of Infectious Diseases.” https://digital.lib. Detection of Infectious Diseases, Department of Trade and Industry, UK washington.edu/xmlui/handle/1773/1985. Government: 35. Andriamandimby, S. F., A. E. Randrianarivo-Solofoniaina, E. M. Jeanmaire, Baylis, M., R. Meiswinkel, and G. J. Venter. 1999. “A Preliminary Attempt L. Ravolomanana, L. T. Razafimanantsoa, T. Rakotojoelinandrasana, to Use Climate Data and Satellite Imagery to Model the Abundance J. Razainirina, et al. “Rift Valley Fever Virus in Madagascar, 2008–2009.” and Distribution of Culicoides Imicola (Diptera: Ceratopogonidae) in 2010. Emerging Infectious Diseases 16: 963–70. Southern Africa.” Journal of the South African Veterinary Association 70 (2): 80–89. Anyamba, A., J.-P. Chretien, J. Small, C. J. Tucker, P. B. Formenty, J.  H.  Richardson, S. C. Britch, D. C. Schnabel, R. L. Erickson, and Baylis, M., P. S. Mellor, E. J. Wittmann, and D. J. Rogers. 2001. “Prediction of K.  J. Linthicum. 2009. “Prediction of a Rift Valley Fever Outbreak.” Areas Around the Mediterranean at Risk of Bluetongue by Modelling Proceedings of the National Academy of Sciences of the United States of the Distribution of Its Vector Using Satellite Imaging.” Veterinary America 106 (3): 955–59. Record 149 (21): 639–43. Anyamba, A., K. J. Linthicum, J. L. Small, K. M. Collins, C. J. Tucker, E. W. Pak, Baylis, M., and P. Rawlings. 1998. “Modelling the Distribution and S. C. Britch, J. R. Eastman, J. E. Pinzon, and K. L. Russell. 2012. “Climate Abundance of Culicoides Imicola in Morocco and Iberia Using Climatic Teleconnections and Recent Patterns of Human and Animal Disease Data and Satellite Imagery.” Archives of Virology. Supplementum Outbreaks.” PLoS Neglected Tropical Diseases 6 (1). 14: 137–53. Archer, Brett N., Jacqueline Weyer, Janusz Paweska, Deliwe Nkosi, Patricia Bazarusanga, Thomas, Dirk Geysen, Jozef Vercruysse, and Maxime Leman, Khin San Tint, and Lucille Blumberg. 2011. “Outbreak of Rift Madder. 2007. “An Update on the Ecological Distribution of Ixodid Valley Fever Affecting Veterinarians and Farmers in South Africa, Ticks Infesting Cattle in Rwanda: Countrywide Cross-sectional Survey 2008.” South African Medical Journal = Suid-Afrikaanse Tydskrif Vir in the Wet and the Dry Season.” Experimental & Applied Acarology Geneeskunde 101 (4): 263–66. 43 (4): 279–91. doi:10.1007/s10493-007-9121-y. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 46 REFERENCES Bendali, F. 2006. “The Design and Implementation of Effective Conte, A., A. Giovannini, L. Savini, M. Goffredo, P. Calistri, and R. Meiswinkel. Epidemiological Surveillance Programmes in Sub-Saharan Africa.” 2003. “The Effect of Climate on the Presence of Culicoides Imicola in Revue Scientifique et Technique de L’Office International des Epizooties Italy.” Journal of Veterinary Medicine, Series B 50 (3): 139–47. 25 (1): 199–209. Conte, A., M. Goffredo, C. Ippoliti, and R. Meiswinkel. 2007. “Influence Black, Peter and Nunn, Mike. Impact of Climate Change and Environmental of Biotic and Abiotic Factors on the Distribution and Abundance of Change on Emerging and Re-Emerging Animal Diseases and Animal Culicoides Imicola and the Obsoletus Complex in Italy.” Veterinary Production. Department of Agriculture, Fisheries and Forestry. Parasitology 150 (4): 333–44. Canberra, Australia: OIE, 2009. Cumming, G. S. 1999. “Host Distributions Do Not Limit the Species Bonfanti, L., M. Cecchinato, G. Ciaravino, F. Montarsi, M. Lorenzetto, Ranges of Most African Ticks (Acari: Ixodida).” Bulletin of Entomological A. Natale, L. Bortolotti, L. Ceglie, and S. Marangon. 2008. “Management Research 89 (4): 303–27. doi:10.1017/S0007485399000450. and Control of Bluetongue Virus Serotype 8 (BTV-8) in Veneto Region.” Large Animal Review 14 (6): 259–66. ———. 2000. “Using Habitat Models to Map Diversity: Pan-African Species Richness of Ticks (Acari: Ixodida).” Journal of Biogeography 27 Cagnolati, V., S. Tempia, and A. M. Abdi. 2006. “Economic Impact of Rift (2): 425–40. doi:10.1046/j.1365-2699.2000.00419.x. Valley Fever on the Somali Livestock Industry and a Novel Surveillance Approach in Nomadic Pastoral Systems.” In 11th International Daszak, P., A. A. Cunningham, and A. D. Hyatt. 2000. “Emerging Infectious Symposium on Veterinary Epidemiology and Economics. Cairns, Australia. Diseases of Wildlife—Threats to Biodiversity and Human Health.” Science 287 (5452): 443. Calistri, P., M. Goffredo, V. Caporale, and R. Meiswinkel. 2003. “The Distribution of Culicoides Imicola in Italy: Application and Evaluation Davies, F. G. 2006. “Risk of a Rift Valley Fever Epidemic at the Haj in Mecca, of Current Mediterranean Models Based on Climate.” Journal of Saudi Arabia.” Revue Scientifique et Technique de L’Office International Veterinary Medicine, Series B 50 (3): 132–38. des Epizooties 25 (1): 137–47. Calvete, C., R. Estrada, M. A. Miranda, D. Borrás, J. H. Calvo, and J. Lucientes. ———. 2010. “The Historical and Recent Impact of Rift Valley Fever in 2008. “Modelling the Distributions and Spatial Coincidence of Africa.” American Journal of Tropical Medicine and Hygiene 83 (2 Suppl): Bluetongue Vectors Culicoides Imicola and the Culicoides Obsoletus 73–74. doi:10.4269/ajtmh.2010.83s2a02. Group Throughout the Iberian Peninsula.” Medical and Veterinary de la Rocque, Stéphane. 2008.“Climate Change: Impact on the Epidemiology Entomology 22 (2): 124–34. and Control of Animal Diseases. Introduction.” Revue Scientifique et Chevalier, V., S. de la Rocque, T. Baldet, L. Vial, and F. Roger. 2004. Technique de L’Office International des Epizooties. 27 (2): 303–08. “Epidemiological Processes Involved in the Emergence of Vector-borne Delgado, C. L. 1999. Livestock to 2020: The Next Food Revolution. Vol. 28. Diseases: West Nile Fever, Rift Valley Fever, Japanese Encephalitis and Washington, DC: International Food Policy Research Institute. Crimean-Congo Haemorrhagic Fever.” Revue Scientifique et Technique de L’Office International des Epizooties 23 (2): 535–55. Diamond, J. 2002. “Evolution, Consequences and Future of Plant and Animal Domestication.” Nature 418 (6898): 700–07. Chevalier, V., Y. Thiongane, and R. Lancelot. 2009. “Endemic Transmission of Rift Valley Fever in Senegal.” Transboundary and Emerging Diseases Diaz, C. M., S. C. Massawe, A. Clemence, G. K. Gitau, H. K. Kiara, G. R. 56 (9–10): 372–74. doi:10.1111/j.1865-1682.2009.01083.x. Muraguri, C. J. O’Callaghan, and B. D. Perry. 2003. “Risk Mapping of East Coast Fever in Coastal and Highland Regions of Kenya Based Cleaveland, S., M. K. Laurenson, and L. H. Taylor. 2001. “Diseases of on Predicted Mortality and Morbidity Incidences.” In ISVEE 10: Humans and Their Domestic Mammals: Pathogen Characteristics, Proceedings of the 10th Symposium of the International Society for Host Range and the Risk of Emergence.” Philosophical Transactions Veterinary Epidemiology and Economics, Vina del Mar, Chile, 17–21 of the Royal Society of London. Series B, Biological Sciences 356 (1411): November 2003. 991–99. doi:10.1098/rstb.2001.0889. Ducheyne, E., R. De Deken, S. Bécu, B. Codina, K. Nomikou, O. Mangana- Clements, A. C. A., D. U. Pfeiffer, and V. Martin. 2006. “Application of Vougiaki, G. Georgiev, B. V. Purse, and G. Hendickx. 2007. “Quantifying Knowledge-driven Spatial Modelling Approaches and Uncertainty the Wind Dispersal of Culicoides Species in Greece and Bulgaria.” Management to a Study of Rift Valley Fever in Africa.” International Geospatial Health 1 (2): 177–89. Journal of Health Geographics 5 (1): 57. Ducheyne, E., M. Lange, Y. Van der Stede, E. Meroc, B. Durand, and G. Clements, Archie C. A., Dirk U. Pfeiffer, Vincent Martin, and M. Joachim Hendrickx. 2011. “A Stochastic Predictive Model for the Natural Otte. 2007a. “A Rift Valley Fever Atlas for Africa.” Preventive Veterinary Spread of Bluetongue.” Preventive Veterinary Medicine 99 (1): 48–59. Medicine 82 (1–2): 72–82. doi:10.1016/j.prevetmed.2007.05.006. Dufour, B., P. Hendrikx, and B. Toma. 2006. “The Design and Establishment Clements, Archie C. A., Dirk U. Pfeiffer, Vincent Martin, Claudia Pittliglio, of Epidemiological Surveillance Systems for High-risk Diseases in Nicky Best, and Yaya Thiongane. 2007b. “Spatial Risk Assessment Developed Countries.” Revue Scientifique et Technique de L’Office of Rift Valley Fever in Senegal.” Vector Borne and Zoonotic Diseases International des Epizooties 25 (1): 187–98. (Larchmont, N.Y.) 7 (2): 203–16. doi:10.1089/vbz.2006.0600. Ebi, K., P. Berry, D. Campbell-Lendrum, C. Corvalán, and J. Guillemot. 2011. Conte, A., P. Colangeli, C. Ippoliti, C. Paladini, M. Ambrosini, L. Savini, F. Protecting Health from Climate Change: Vulnerability and Adaptation Dall’Acqua, and P. Calistri. 2005. “The Use of a Web-based Interactive Assessment. Geneva: World Health Organization. Geographical Information System for the Surveillance of Bluetongue in Italy.” Revue Scientifique et Technique de L’Office International des EFSA Panel on Animal Health and Welfare (AHAW). 2013. “Scientific Epizooties 24 (3): 857–68. Opinion on Rift Valley Fever.” EFSA Journal 11(4): 3180 [48 pp]. R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S REFERENCES 47 El Mamy, Ahmed B. O., Mohamed Ould Baba, Yahya Barry, Katia Isselmou, Gilbert, M., A. Mitchell, D. Bourn, J. Mawdsley, R. Clifton-Hadley, and W. Mamadou L. Dia, Mohamed O. B. El Kory, Mariam Diop, et al. 2011. Wint. 2005. “Cattle Movements and Bovine Tuberculosis in Great “Unexpected Rift Valley Fever Outbreak, Northern Mauritania.” Britain.” Nature 435 (7041): 491–96. doi:10.1038/nature03548. Emerging Infectious Diseases 17 (10): 1894–96. doi:10.3201/ Gilbert, Marius, Xiangming Xiao, Dirk U. Pfeiffer, M. Epprecht, Stephen eid1710.110397. Boles, Christina Czarnecki, Prasit Chaitaweesub, et al. 2008. “Mapping Elith, J., C. H. Graham, et al. 2006. “Novel methods improve prediction of H5N1 Highly Pathogenic Avian Influenza Risk in Southeast Asia.” species’ distributions from occurrence data.” Ecography 29, 129–151. Proceedings of the National Academy of Sciences of the United States of America 105 (12): 4769–74. doi:10.1073/pnas.0710581105. Epstein, Paul R. 2001. “Climate Change and Emerging Infectious Diseases.” Microbes and Infection 3 (9): 747–54. doi:10.1016/S1286- Ginsberg, Jeremy, Matthew H. Mohebbi, Rajan S. Patel, Lynnette Brammer, 4579(01)01429-0. Mark S. Smolinski, and Larry Brilliant. 2008. “Detecting Influenza Epidemics Using Search Engine Query Data.” Nature 457 (7232): ESA/Avia-GIS. 2012. “VECMAP—Demonstration Study on Services for 1012–14. doi:10.1038/nature07634. Mosquito Habitat Mapping, ESA—Integrated & Telecommunications Applications.” http://artes-apps.esa.int/projects/vecmap. Giovannini, A. P. Calistri, A. Conte, L. Savini, D. Nannini, C. Patta, U. Santucci, and V. Caporale. 2004a. “Bluetongue Virus Surveillance in a Fadiga, M., C. Jost, and J. Ihedioha. 2011. “Financial Costs of Disease Newly Infected Area.” Veterinaria Italiana 40 (3): 188–97. Burden, Morbidity and Mortality from Priority Livestock Diseases in Nigeria: Disease Burden and Cost-benefit Analysis of Targeted Giovannini, A., C. Paladini, P. Calistri, A. Conte, P. Colangeli, U. Santucci, Interventions.” Nigeria Integrated Animal and Human Health D. Nannini, and V. Caporale. 2004b. “Surveillance System of Bluetongue Management Project Final Report. Nairobi, Kenya: International in Italy.” Veterinaria Italiana 40 (3): 369–84. Livestock Research Institute. Gloster, J., P. S. Mellor, L. Burgin, C. Sanders, and S. Carpenter. 2007a. “Will Fandamu, P., L. Duchateau, N. Speybroeck, T. Marcotty, V. Mbao, J. Mtambo, Bluetongue Come on the Wind to the United Kingdom in 2007?” M. Mulumba, and D. Berkvens. 2005. “Theileria Parva Seroprevalence Veterinary Record 160 (13): 422–26. in Traditionally Kept Cattle in Southern Zambia and El Nino.” Gloster, J., P. S. Mellor, A. J. Manning, H. N. Webster, and M. C. Hort. 2007b. International Journal for Parasitology 35 (4): 391–96. doi:10.1016/j. “Assessing the Risk of Windborne Spread of Bluetongue in the 2006 ijpara.2004.12.011. Outbreak of Disease in Northern Europe.” Veterinary Record 160 (2): Fandamu, P., L. Duchateau, N. Speybroeck, M. Mulumba, and D. Berkvens. 54–56. 2006. “East Coast Fever and Multiple El Niño Southern Oscillation Gonzalez, J. P., B. Le Guenno, M. J. Some, and J. A. Akakpo. 1992. Ranks.” Veterinary Parasitology 135 (2): 147–52. doi:10.1016/j. “Serological Evidence in Sheep Suggesting Phlebovirus Circulation in vetpar.2005.09.008. a Rift Valley Fever Enzootic Area in Burkina Faso.” Transactions of the Royal Society of Tropical Medicine and Hygiene 86 (6): 680–82. FAO (Food and Agriculture Organization). 2009. FAO Statistical Database. Available at http://faostat.fao.org/. Grace, D., F. Mutua, P. Ochungo, R. Kruska, K. Jones, L. Brierley, L. Lapar, et al. 2012. “Mapping of Poverty and Likely Zoonoses Hotspots: Zoonoses Gale, P., A. Brouwer, V. Ramnial, L. Kelly, R. Kosmider, A. R. Fooks, and Project 4.” Report to the Department for International Development. E. L. Snary. 2010. “Assessing the Impact of Climate Change on Nairobi: International Livestock Research Institute. Vector-borne Viruses in the EU Through the Elicitation of Expert Opinion.” Epidemiology and Infection 138 (2): 214–25. doi:10.1017/ Gubbins, S., C. Szmaragd, L. Burgin, A. Wilson, V. Volkova, J. Gloster, and S0950268809990367. G. J. Gunn. 2010. “Assessing the Consequences of an Incursion of a Vector-borne Disease. I. Identifying Feasible Incursion Scenarios for García-Lastra, R., I. Leginagoikoa, J. M. Plazaola, B. Ocabo, G. Aduriz, Bluetongue in Scotland.” Epidemics 2 (3): 148–54. T.  Nunes, and R. A. Juste. 2012. “Bluetongue Virus Serotype 1 Outbreak in the Basque Country (Northern Spain) 2007–2008. Guis, H., C. Caminade, C. Calvete, A. P. Morse, A. Tran, and M. Baylis. Data Support a Primary Vector Windborne Transport.” PLoS 2012. “Modelling the Effects of Past and Future Climate on the Risk ONE 7 (3). of Bluetongue Emergence in Europe.” Journal of the Royal Society Interface 9 (67): 339–50. Gerbier, G., T. Baldet, A. Tran, G. Hendrickx, H. Guis, K. Mintiens, A. R. W. Elbers, and C. Staubach. 2008. “Modelling Local Guis, H., A. Tran, S. de La Rocque, T. Baldet, G. Gerbier, B. Barragué, Dispersal of Bluetongue Virus Serotype 8 Using Random Walk.” F. Biteau-Coroller, F. Roger, J.-F. Viel, and F. Mauny. 2007. “Use of Preventive Veterinary Medicine 87 (1–2): 119–30. doi:10.1016/j. High-resolution Satellite Imagery to Characterize Landscapes at prevetmed.2008.06.012. Risk for Bluetongue in Corsica, France.” Veterinary Research 38: 669–83. Gething, Peter W., David L. Smith, Anand P. Patil, Andrew J. Tatem, Robert W. Snow, and Simon I. Hay. 2010. “Climate Change and the Hadorn, D. C., V. Racloz, H. Schwermer, and K. D. C. Stärk. 2009. “Establishing Global Malaria Recession.” Nature 465 (7296): 342–45. doi:10.1038/ a Cost-effective National Surveillance System for Bluetongue Using nature09098. Scenario Tree Modelling.” Veterinary Research 40 (6): 57. Gilbert, M., P. Chaitaweesub, T. Parakarnawongsa, S. Premashthira, Harper, Kristin, and George Armelagos. 2010. “The Changing Disease- T. Tiensin, W. Kalpravidh, H. Wagner, and J. Slingenbergh. 2006. “Free- Scape in the Third Epidemiological Transition.” International Journal of grazing Ducks and Highly Pathogenic Avian Influenza, Thailand.” Environmental Research and Public Health 7 (2): 675–97. doi:10.3390/ Emerging Infectious Diseases 12 (2): 227–34. ijerph7020675. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 48 REFERENCES Hartemink, N. A., B. V. Purse, R. Meiswinkel, H. E. Brown, A. de Koeijer, Kedmi, M., Y. Herziger, N. Galon, R. M. Cohen, M. Perel, C. Batten, Y. Braverman, A. R. W. Elbers, G.-J. Boender, D. J. Rogers, and J. A. P. Heesterbeek. Y. Gottlieb, N. Shpigel, and E. Klement. 2010. “The Association of Winds 2009. “Mapping the Basic Reproduction Number (R0) for Vector-borne with the Spread of EHDV in Dairy Cattle in Israel During an Outbreak in Diseases: A Case Study on Bluetongue Virus.” Epidemics 1 (3): 153–61. 2006.” Preventive Veterinary Medicine 96 (3–4): 152–60. Hartley, David M., Jennifer L. Rinderknecht, Terry L. Nipp, Neville P. Clarke, Kivaria, F. M. 2006. “Estimated Direct Economic Costs Associated with and Gary D. Snowder. 2011. “Potential Effects of Rift Valley Fever in Tick-borne Diseases on Cattle in Tanzania.” Tropical Animal Health and the United States.” Emerging Infectious Diseases 17 (8): e1. doi:10.3201/ Production 38 (4): 291–99. eid1708.101088. Konrad, S. K., S. N. Miller, and W. K. Reeves. 2011. “A Spatially Explicit Häsler, B., K. S. Howe, E. Di Labio, H. Schwermer, and K. D. C. Stärk. Degree-day Model of Rift Valley Fever Transmission Risk in the 2012. “Economic Evaluation of the Surveillance and Intervention Continental United States.” GeoJournal 76 (3): 257–66. Programme for Bluetongue Virus Serotype 8 in Switzerland.” Koslowsky, S., C. Staubach, M. Kramer, and L. H. Wieler. 2004. “Risk Preventive Veterinary Medicine 103 (2–3): 93–111. Assessment of Bluetongue Disease Incursion into Germany Using Hassan, Osama Ahmed, Clas Ahlm, Rosemary Sang, and Magnus Evander. Geographic Information System (GIS).” Berliner Und Munchener 2011. “The 2007 Rift Valley Fever Outbreak in Sudan.” PLoS Neglected Tierarztliche Wochenschrift 117 (5–6): 214–25. Tropical Diseases 5 (9): e1229. doi:10.1371/journal.pntd.0001229. Kovats, R. S., K. Ebi, and B. Menne. 2003. “Methods of Assessing Human Heffernan, C., M. Salman, and L. York. 2012. “Livestock Infectious Disease Health Vulnerability and Public Health Adaptation to Climate Change.” and Climate Change: A Review of Selected Literature.” CAB Reviews: Health and Global Environmental Change Series No. 1. Copenhagen: Perspectives in Agriculture, Veterinary Science, Nutrition and Natural World Health Organization Europe. Resources 7. Kuhn, K., D. Campbell-Lendrum, A. Haines, J. Cox, C. Corvalán, M. Anker, Hendrickx, G., M. Gilbert, C. Staubach, A. Elbers, K. Mintiens, G. Gerbier, and and R. B. Malaria. 2005. “Using Climate to Predict Infectious Disease E. Ducheyne. 2008. “A Wind Density Model to Quantify the Airborne Epidemics.” Geneva: World Health Organization. Spread of Culicoides Species During North-western Europe Bluetongue Epidemic, 2006.” Preventive Veterinary Medicine 87 (1–2): 162–81. LaBeaud, A. Desiree, James W. Kazura, and Charles H. King. 2010. “Advances in Rift Valley Fever Research: Insights for Disease Hughes-Fraire, R., A. D. Hagerman, B. A. McCarl, and H. Gaff. 2011. “Rift Prevention.” Current Opinion in Infectious Diseases 23 (5): 403–08. Valley Fever: An Economic Assessment of Agricultural and Human doi:10.1097/QCO.0b013e32833c3da6. Vulnerability.” In Southern Agricultural Economics Association, 2011 Annual Meeting, February 5–8, 2011, Corpus Christi, Texas. IFAD Last, J. M. 2001. A Dictionary of Epidemiology. New York: Oxford University (International Fund for Agricultural Development). 2009. Livestock Press USA. and Climate Change. Livestock Thematic Papers: Tools for Project LeBreton, Matthew, Sally Umlauf, Cyrille F. Djoko, Peter Daszak, Donald S. Design. Rome: IFAD. Burke, Paul Yemgai Kwenkam, and Nathan D. Wolfe. 2006. “Rift Valley ILRI /FAO (International Livestock Research Institute/ Food and Agriculture Fever in Goats, Cameroon.” Emerging Infectious Diseases 12 (4): 702–03. Organization). 2010. “Decision-support Tool for Prevention and Lessard, P., R. L’Eplattenier, R. A. Norval, B. D. Perry, T. T. Dolan, A. Burrill, Control of Rift Valley Fever Epizootics in the Greater Horn of Africa.” H. Croze, M. Sorensen, J. G. Grootenhuis, and A. D. Irvin. 1988. “The American Journal of Tropical Medicine and Hygiene 83: 75–85. Use of Geographical Information Systems in Estimating East Coast IOM (Institute of Medicine). 2007. “Summary and Assessment—Global Fever Risk to African Livestock.” Acta Veterinaria Scandinavica. Infectious Disease Surveillance and Detection—NCBI Bookshelf.” Supplementum 84: 234–36. Washington, DC: IOM. LIDC (London International Development Centre). 2010. “Maasai Vets IPCC (Intergovernmental Panel on Climate Change). 2007. Climate Change Carry Out Disease Surveillance of 86,000 Animals With Google Mobile 2007: Synthesis Report. Contribution of Working Groups I, II and III to Phones.” London: LIDC. the Fourth Assessment Report of the Intergovernmental Panel on Linthicum, K. J., F. G. Davies, A. Kairo, and C. L. Bailey. 1985. “Rift Valley Climate Change. Geneva: IPCC. Fever Virus (family Bunyaviridae, Genus Phlebovirus). Isolations from Jones, K. E., N. G. Patel, M. A. Levy, A. Storeygard, D. Balk, J. L. Gittleman, Diptera Collected During an Inter-epizootic Period in Kenya.” The and P. Daszak. 2008. “Global Trends in Emerging Infectious Diseases.” Journal of Hygiene 95 (1): 197–209. Nature 451 (7181): 990–93. Little, P. 2009. “Hidden Value on the Hoof: Cross-border Livestock Trade Jongejan, F., and G. Uilenberg. 2004. “The Global Importance of Ticks.” in Eastern Africa.” Common Market for Eastern and Southern Africa Parasitology 129 (S1): S3–S14. Comprehensive African Agriculture Development Program, Policy Brief 2. London: International Institute for Environment and Development. Jost, C. C., J. C. Mariner, P. L. Roeder, E. Sawitri, and G. J. Macgregor-Skinner. 2007. “Participatory Epidemiology in Disease Surveillance and López-Vélez, Rogelio, and Ricardo Molina Moreno. 2005. “[Climate Change Research.” Revue Scientifique et Technique de L’Office International des in Spain and Risk of Infectious and Parasitic Diseases Transmitted by Epizooties 26 (3): 537–49. Arthropods and Rodents].” Revista Española De Salud Pública 79 (2): 177–90. Jost, Christine C., Serge Nzietchueng, Simon Kihu, Bernard Bett, George Njogu, Emmanuel S. Swai, and Jeffrey C. Mariner. 2010. Malak, A. K., L. Mpoke, J. Banak, S. Muriuki, R. A. Skilton, D. Odongo, J. “Epidemiological Assessment of the Rift Valley Fever Outbreak Sunter, and H. Kiara. 2012. “Prevalence of Livestock Diseases and Their in Kenya and Tanzania in 2006 and 2007.” American Journal of Impact on Livelihoods in Central Equatoria State, Southern Sudan.” Tropical Medicine and Hygiene 83 (2 Suppl): 65–72. doi:10.4269/ Preventive Veterinary Medicine 104 (3–4): 216–23. doi:10.1016/j. ajtmh.2010.09-0290. prevetmed.2011.12.001. R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S REFERENCES 49 Mariner, J. C., J. Morrill, and T. G. Ksiazek. 1995. “Antibodies to Hemorrhagic Ocaido, M. R. T. Muwazi, and J. Asibo Opuda. 2009. “Disease Incidence Fever Viruses in Domestic Livestock in Niger: Rift Valley Fever and in Ranch and Pastoral Livestock Herds Around Lake Mburo National Crimean-Congo Hemorrhagic Fever.” American Journal of Tropical Park, in South Western Uganda.” Tropical Animal Health and Production Medicine and Hygiene 53 (3): 217–21. 41 (7): 1299–1308. doi:10.1007/s11250-009-9315-x. Martin, V., V. Chevalier, P. Ceccato, A. Anyamba, L. De Simone, J. Lubroth, OIE (World Organisation for Animal Health). 2014. Animal Health S. de La Rocque, and J. Domenech. 2008. “The Impact of Climate Information Department. Change on the Epidemiology and Control of Rift Valley Fever.” Revue ———. 2014. PVS Pathway. www.oie.int/en/support-to-oie-members/ Scientifique et Technique de L’Office International des Epizooties 27 (2): pvs-pathway/. 413–26. Olaleye, O. D., O. Tomori, and H. Schmitz. 1996. “Rift Valley Fever in Nigeria: McCarthy, Michael A., Colin J. Thompson, Cindy Hauser, Mark A. Infections in Domestic Animals.” Revue Scientifique et Technique de Burgman, Hugh P. Possingham, Melinda L. Moir, Thanawat Tiensin, L’Office International des Epizooties 15 (3): 937–46. and Marius Gilbert. 2010. “Resource Allocation for Efficient Environmental Management.” Ecology Letters 13 (10): 1280–89. Olwoch, J. M., B. Reyers, F. A. Engelbrecht, and B. F. N. Erasmus. 2008. doi:10.1111/j.1461-0248.2010.01522.x. “Climate Change and the Tick-borne Disease, Theileriosis (East Coast Fever) in Sub-Saharan Africa.” Journal of Arid Environments 72 (2): McDermott, J. J., P. M. Kristjanson, R. L. Kruska, R. S. Reid, T. P. Robinson, P. G. 108–20. Coleman, P. G. Jones, and P. K. Thornton. 2002. “Effects of Climate, Human Population and Socio-economic Changes on Tsetse-transmitted Otte, M. J., R. Nugent, and A. McLeod. 2004. Transboundary Animal Trypanosomiasis to 2050.” The African Trypanosomes: 25–38. Diseases: Assessment of Socio-economic Impacts and Institutional Responses. Livestock Policy Discussion Paper No. 9. Rome: Food and Meegan, J. M., H. Hoogstraal, and M. I. Moussa. 1979. “An Epizootic of Rift Agriculture Organization. Valley Fever in Egypt in 1977.” The Veterinary Record 105 (6): 124–25. Ouagal, M., P. Hendrikx, D. Berkvens, A. Ncharé, B. Cissé, P.Y. Akpeli, K. Sory, and Meiswinkel, R., M. Goffredo, P. Leijs, and A. Conte. 2008. “The Culicoides C. Saegerman. 2008. “Epidemiological Surveillance Networks for Animal ‘Snapshot’: A Novel Approach Used to Assess Vector Densities Widely Diseases in French-speaking West and Central Africa.” Revue Scientifique and Rapidly During the 2006 Outbreak of Bluetongue (BT) in The et Technique de L’Office International des Epizooties (3): 689–702. Netherlands.” Preventive Veterinary Medicine 87 (1–2): 98–118. Pascucci, Ilaria, Andrea Capobianco Dondona, Cesare Cammà, Maurilia Mellor, P. S., S. Carpenter, L. Harrup, M. Baylis, and P. P. C. Mertens. 2008. Marcacci, Marco Di Domenico, Rossella Lelli, Massimo Scacchia, et al. “Bluetongue in Europe and the Mediterranean Basin: History of 2011. “Survey of Ixodid Ticks and Two Tick-borne Pathogens in African Occurrence Prior to 2006.” Preventive Veterinary Medicine 87 (1–2): 4–20. Buffaloes, Syncerus Caffer, from the Caprivi Strip, Namibia.” Journal of Minjauw, B., and A. McLeod. 2003. Tick-borne Diseases and Poverty: The Zoo and Wildlife Medicine. 42 (4): 634–40. Impact of Ticks and Tick-borne Diseases on the Livelihoods of Small- Patz, Jonathan A., Peter Daszak, Gary M. Tabor, A. Alonso Aguirre, scale and Marginal Livestock Owners in India and Eastern and Southern Mary Pearl, Jon Epstein, Nathan D. Wolfe, et al. 2004. “Unhealthy Africa. Edinburgh: DFID Animal Health Programme. Landscapes: Policy Recommendations on Land Use Change and Mohamed, Mohamed, Fausta Mosha, Janeth Mghamba, Sherif R. Infectious Disease Emergence.” Environmental Health Perspectives 112 Zaki, Wun-Ju Shieh, Janusz Paweska, Sylvia Omulo, et al. 2010. (10): 1092–98. doi:10.1289/ehp.6877. “Epidemiologic and Clinical Aspects of a Rift Valley Fever Outbreak in Perrin, J.-B., C. Ducrot, J.-L. Vinard, E. Morignat, A. Gauffier, D. Calavas, and Humans in Tanzania, 2007.” American Journal of Tropical Medicine and P. Hendrikx. 2010. “Using the National Cattle Register to Estimate the Hygiene 83 (2 Suppl): 22–27. doi:10.4269/ajtmh.2010.09-0318. Excess Mortality During an Epidemic: Application to an Outbreak of Mukhebi, A. W., B. D. Perry, and R. Kruska. 1992. “Estimated Economics of Bluetongue Serotype 8.” Epidemics 2 (4): 207–14. Theileriosis Control in Africa.” Preventive Veterinary Medicine 12 (1–2): Perry, B. D., T. F. Randolph, and P. K. Thornton. 2002. Investing in Animal 73–85. Health Research to Alleviate Poverty. Nairobi: International Livestock Research Institute. Munyua, Peninah, Rees M. Murithi, Sherrilyn Wainwright, Jane Githinji, Allen Hightower, David Mutonga, Joseph Macharia, et al. 2010. “Rift Pfeiffer, D. U., L. Duchateau, R. L. Kruska, U. Ushewokunze-Obatolu, and Valley Fever Outbreak in Livestock in Kenya, 2006–2007.” American B. D. Perry. 1997. “A Spatially Predictive Logistic Regression Model for Journal of Tropical Medicine and Hygiene 83 (2 Suppl): 58–64. Occurrence of Theileriosis Outbreaks in Zimbabwe.” Epidemiologie et doi:10.4269/ajtmh.2010.09-0292. Santé Animale 31–32: 12.12.1–3. Murithi, R. M., P. Munyua, P. M. Ithondeka, J. M. Macharia, A. Hightower, Pili, E., S. Ciuccé, J. Culurgioni, V. Figus, G. Pinna, and A. Marchi. 2006. E. T. Luman, R. F. Breiman, and M. Kariuki Njenga. 2011. “Rift Valley “Distribution and Abundance of Bluetongue Vectors in Sardinia: Fever in Kenya: History of Epizootics and Identification of Vulnerable Comparison of Field Data with Prediction Maps.” Journal of Veterinary Districts.” Epidemiology and Infection 139 (3): 372–380. doi:10.1017/ Medicine Series B: Infectious Diseases and Veterinary Public Health S0950268810001020. 53 (7): 312–16. Norval, R. A., J. A. Lawrence, A. S. Young, B. D. Perry, T. T. Dolan, and J. Pioz, Maryline, Hélène Guis, Didier Calavas, Benoît Durand, David Abrial, Scott. 1991. “Theileria Parva: Influence of Vector, Parasite and Host and Christian Ducrot. 2011. “Estimating Front-wave Velocity of Relationships on the Epidemiology of Theileriosis in Southern Africa.” Infectious Diseases: A Simple, Efficient Method Applied to Bluetongue.” Parasitology 102 (Pt 3): 347–56. Veterinary Research 42 (1): 60. doi:10.1186/1297-9716-42-60. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R 50 REFERENCES Pourrut, Xavier, Dieudonné Nkoghé, Marc Souris, Christophe Paupy, Rushton, J., C. L. Heffernan, and D. Pilling. 2002. “A Literature Review of Janusz Paweska, Cindy Padilla, Ghislain Moussavou, and Eric M. Livestock Diseases and Their Importance in the Lives of Poor People.” Leroy. 2010. “Rift Valley Fever Virus Seroprevalence in Human Rural In Mapping Poverty and Livestock in the Developing World. ed. P. K. Populations of Gabon.” PLoS Neglected Tropical Diseases 4 (7): e763. Thornton, R. L. Kruska, N. Henninger, P. M. Kristjanson, R. S. Reid, doi:10.1371/journal.pntd.0000763. F. Atieno, A. N. Odero, and T. Ndegwa, 129. Nairobi: International Livestock Research Institute. Prüss-Ustün, Annette, and Carlos Corvalán. 2007. “How Much Disease Burden Can Be Prevented by Environmental Interventions?” Saegerman, Claude, Dirk Berkvens, and Philip S Mellor. 2008. “Bluetongue Epidemiology 18 (1): 167–78. doi:10.1097/01.ede.0000239647.26389.80. Epidemiology in the European Union.” Emerging Infectious Diseases 14 (4): 539–44. doi:10.3201/eid1404.071441. Purse, B. V., H. E. Brown, L. Harrup, P. P. C. Mertens, and D. J. Rogers. 2008. “Invasion of Bluetongue and Other Orbivirus Infections into Europe: Samui, K. L., S. Inoue, A. S. Mweene, A. M. Nambota, J. E. Mlangwa, P. The Role of Biological and Climatic Processes.” Revue Scientifique et Chilonda, M. Onuma, and C. Morita. 1997. “Distribution of Rift Valley Technique de L’Office International des Epizooties 27 (2): 427–42. Fever Among Cattle in Zambia.” Japanese Journal of Medical Science & Biology 50 (2): 73–77. Purse, B. V., B. J. J. Mccormick, P. S. Mellor, M. Baylis, J. P. T. Boorman, D. Borras, I. Burgu, et al. 2007. “Incriminating Bluetongue Virus Vectors with Sedda, Luigi, Heidi E. Brown, Bethan V. Purse, Laura Burgin, John Gloster, Climate Envelope Models.” Journal of Applied Ecology 44 (6): 1231–42. and David J. Rogers. 2012. “A New Algorithm Quantifies the Roles of Wind and Midge Flight Activity in the Bluetongue Epizootic in Purse, B. V., P. S. Mellor, D. J. Rogers, A. R. Samuel, P. P. C. Mertens, and Northwest Europe.” Proceedings. Biological Sciences/The Royal Society M. Baylis. 2005. “Climate Change and the Recent Emergence of 279 (1737): 2354–62. doi:10.1098/rspb.2011.2555. Bluetongue in Europe.” Nature Reviews Microbiology 3 (2): 171–81. Shaman, Jeffrey, and Marc Lipsitch. 2012. “The El Nino-Southern Oscillation Purse, B. V., A. J. Tatem, S. Caracappa, D. J. Rogers, P. S. Mellor, M. Baylis, (ENSO)-pandemic Influenza Connection: Coincident or Causal?” and A. Torina. 2004. “Modelling the Distributions of Culicoides Proceedings of the National Academy of Sciences of the United States of Bluetongue Virus Vectors in Sicily in Relation to Satellite-derived America 110 (Suppl. 1): 3689–91. doi:10.1073/pnas.1107485109. Climate Variables.” Medical and Veterinary Entomology 18 (2): 90–101. Racloz, V., R. Straver, M. Kuhn, B. Thur, T. Vanzetti, K. D. C. Stärk, C. Griot, Shaw, A., Hendrickx, G., M. Gilbert, R. Mattioli, V. Codjia, B. Dao, O. Diall, and A. Cagienard. 2006. “Establishment of an Early Warning System C. Mahama, I. Sidibé, and W. Wint. 2006. Mapping the Benefits: a New Against Bluetongue Virus in Switzerland.” Schweizer Archiv Fur Decision Tool for Tsetse and Trypanosomiasis Interventions. Edinburgh: Tierheilkunde 148 (11): 593–98. DFID Animal Health Programme and FAO Programme Against African Trypanosomiasis. Racloz, V., G. Venter, C. Griot, and K. D. C. Stärk. 2008. “Estimating the Temporal and Spatial Risk of Bluetongue Related to the Incursion of Shaw, A., W. Wint, G. Cecchi, S. J. Torr, R. Mattioli, and T. P. Robinson. Infected Vectors into Switzerland.” BMC Veterinary Research 4. Submitted. “Benefit-cost Mapping of Bovine Trypanosomosis Control in Eastern Africa.” Agricultural Systems. Randolph, S. 1999. “Epidemiological Uses of a Population Model for the Tick Rhipicephalus Appendiculatus.” Tropical Medicine & International Silbermayr, K., K. Hackländer, C. Doscher, J. Koefer, and K. Fuchs. 2011. “A Health 4 (9): A34–42. Spatial Assessment of Culicoides Spp. Distribution and Bluetongue Disease Risk Areas in Austria.” Berliner Und Munchener Tierarztliche Randolph, T. F., E. Schelling, D. Grace, C. F. Nicholson, J. L. Leroy, D. C. Cole, M. Wochenschrift 124 (5–6): 228–35. W. Demment, A. Omore, J. Zinsstag, and M. Ruel. 2007. “Invited Review: Role of Livestock in Human Nutrition and Health for Poverty Reduction Simuunza, Martin, William Weir, Emily Courcier, Andy Tait, and Brian in Developing Countries.” Journal of Animal Science 85 (11): 2788–800. Shiels. 2011. “Epidemiological Analysis of Tick-borne Diseases in Zambia.” Veterinary Parasitology 175 (3–4): 331–42. doi:10.1016/j. Rich, K. M., and F. Wanyoike. 2010. “An Assessment of the Regional and vetpar.2010.09.027. National Socio-economic Impacts of the 2007 Rift Valley Fever Outbreak in Kenya.” American Journal of Tropical Medicine and Hygiene Sindato, Calvin, Esron Karimuribo, and Leonard E. G. Mboera. 2012. 83 (2 Suppl): 52–57. “The Epidemiology and Socio-economic Impact of Rift Valley Fever in Tanzania: A Review.” Tanzania Journal of Health Research 13 (5). Ringot, David, Jean-Paul Durand, Hugues Toulou, Jean-Paul Boutin, and doi:10.4314/thrb.v13i5.2. Bernard Davoust. 2004. “Rift Valley Fever in Chad.” Emerging Infectious Diseases 10 (5): 945–47. Smil, V. 2002. “Worldwide Transformation of Diets, Burdens of Meat Production and Opportunities for Novel Food Proteins.” Enzyme and Rogers, D. J., S. I. Hay, and M. J. Packer. 1996. “Predicting the Distribution Microbial Technology 30 (3): 305–11. of Tsetse Flies in West Africa Using Temporal Fourier Processed Meteorological Satellite Data.” Annals of Tropical Medicine and Soumare, Baba, Stefano Tempia, Vittorio Cagnolati, Abdullatif Mohamoud, Parasitology 90 (3): 225–42. Guido Van Huylenbroeck, and Dirk Berkvens. 2007. “Screening for Rift Valley Fever Infection in Northern Somalia: A GIS Based Survey Rogers, D. J., and S. E. Randolph. 2006. “Climate Change and Vector-borne Method to Overcome the Lack of Sampling Frame.” Veterinary Diseases. Advances in Parasitology, 62: 345–81. Microbiology 121 (3–4): 249–56. doi:10.1016/j.vetmic.2006.12.017. Rubaire-Akiiki, Christopher M., Joseph Okello-Onen, David Musunga, Sperlova, A., and D. Zendulkova. 2011. “Bluetongue: A Review.” Veterinarni Edmond K. Kabagambe, Mettee Vaarst, David Okello, Charles Opolot, Medicina 56 (9): 430–52. et al. 2006. “Effect of Agro-ecological Zone and Grazing System on Incidence of East Coast Fever in Calves in Mbale and Sironko Districts Steinfeld, H., P. Gerber, T. D. Wassenaar, V. Castel, and C. de Haan. 2006. of Eastern Uganda.” Preventive Veterinary Medicine 75 (3–4): 251–66. Livestock’s Long Shadow: Environmental Issues and Options. Rome: doi:10.1016/j.prevetmed.2006.04.015. Food and Agriculture Organization. R E D U C I N G C L I M AT E - S E N S I T I V E D I S E A S E R I S K S REFERENCES 51 Stuber, B., H. Ochs, A. Tschuor, P. Zanolari, J. Danuser, and L. Perler. 2009. Vignolles, C., Y. M. Tourre, O. Mora, L. Imanache, and M. Lafaye. “Livestock Owners with Specifically Increased Disease Awareness 2010. “TerraSAR-X High-resolution Radar Remote Sensing: An for Bluetongue: A New Approach to Disease Surveillance.” Schweizer Operational Warning System for Rift Valley Fever Risk.” Geospatial Archiv Fur Tierheilkunde 151 (7): 317–21. Health 5 (1): 23–31. Swaroop, S. 1949. “Forecasting of Epidemic Malaria in the Punjab, India.” WHO (World Health Organization). 2009. Libreville Declaration on Health American Journal of Tropical Medicine and Hygiene 29 (1): 1–17. and Environment in Africa. Libreville, 29 August 2008. Szmaragd, C., G. J. Gunn, and S. Gubbins. 2010.“Assessing the Consequences ———. 2010. “WHO: Rift Valley Fever.” Fact Sheet No. 207. Geneva: of an Incursion of a Vector-borne Disease. II. Spread of Bluetongue in WHO. Scotland and Impact of Vaccination.” Epidemics 2 (3): 139–47. Wilson, A. J., and P. S. Mellor. 2009. “Bluetongue in Europe: Past, Present Tabachnick, W. J., C. T. Smartt, and C. R. Connelly. 2008. “Bluetongue.” and Future.” Philosophical Transactions of the Royal Society B: Biological Entomology and Nematology Department, University of Florida IFAS Sciences 364 (1530): 2669–81. Extension, Gainesville, Florida. Wilson, Mark L. 1994. “Rift Valley Fever Virus Ecology and the Tatem, A. J., M. Baylis, P. S. Mellor, B. V. Purse, R. Capela, I. Pena, and D. J. Rogers. Epidemiology of Disease Emergencea.” Annals of the New York 2003. “Prediction of Bluetongue Vector Distribution in Europe and North Academy of Sciences 740 (1): 169–80. doi:10.1111/j.1749-6632.1994. Africa Using Satellite Imagery.” Veterinary Microbiology 97 (1–2): 13–29. tb19867.x. Taylor, L. H., S. M. Latham, and M. E. Woolhouse. 2001. “Risk Factors for Witt, Clara J., Allen L. Richards, Penny M. Masuoka, Desmond H. Foley, Human Disease Emergence.” Philosophical Transactions of the Royal Anna L. Buczak, Lillian A. Musila, Jason H. Richardson, et al. 2011. Society of London. Series B, Biological Sciences 356 (1411): 983–89. “The AFHSC-Division of GEIS Operations Predictive Surveillance doi:10.1098/rstb.2001.0888. Program: A Multidisciplinary Approach for the Early Detection and Thiongane, Y., H. Zeller, M. M. Lo, N. A. Fati, J. A. Akakpo, and J. P. Gonzalez. 1994. Response to Disease Outbreaks.” BMC Public Health 11 (Suppl 2): S10. “[Decrease of Natural Immunity Against Rift Valley Fever in Domestic doi:10.1186/1471-2458-11-S2-S10. Ruminants of the Senegal River Basin After the Epizootic Outbreak of 1987].” Bulletin De La Société De Pathologie Exotique (1990) 87 (1): 5–6. Wittmann, E. J., P. S. Mellor, and M. Baylis. 2001. “Using Climate Data to Map the Potential Distribution of Culicoides Imicola (Diptera: Tiensin, Thanawat, Prasit Chaitaweesub, Thaweesak Songserm, Arunee Ceratopogonidae) in Europe.” Revue Scientifique et Technique de Chaisingh, Wirongrong Hoonsuwan, Chantanee Buranathai, Tippawon L’Office International des Epizooties 20 (3): 731–40. Parakamawongsa, et al. 2005. “Highly Pathogenic Avian Influenza H5N1, Thailand, 2004.” Emerging Infectious Diseases 11 (11): 1664–72. Wolfe, N. D., C. P. Dunavan, and J. Diamond. 2007. “Origins of Major Human Infectious Diseases.” Nature 447 (7142): 279–83. Tourre, Yves M., Jean-Pierre Lacaux, Cécile Vignolles, and Murielle Lafaye. 2009. “Climate Impacts on Environmental Risks Evaluated from Space: Woods, Christopher W., Adam M. Karpati, Thomas Grein, Noel McCarthy, A Conceptual Approach to the Case of Rift Valley Fever in Senegal.” Peter Gaturuku, Eric Muchiri, Lee Dunster, et al. 2002. “An Outbreak of Global Health Action 11: 2. doi:10.3402/gha.v2i0.2053. Rift Valley Fever in Northeastern Kenya, 1997–98.” Emerging Infectious Diseases 8 (2): 138–44. Tourre, Yves M., Jean-Pierre Lacaux, Cecile Vignolles, Jacques-André Ndione, and Murielle Lafaye. 2008. “Mapping of Zones Potentially Woodward, A. J. 2011. “Global Environmental Change and Disease Occupied by Aedes Vexans and Culex Poicilipes Mosquitoes, the Main Dynamics.” In Climate Change and Disease Dynamics in India, 1–10. Vectors of Rift Valley Fever in Senegal.” Geospatial Health 3 (1): 69–79. New Delhi: TERI Press. van Lieshout, M., R. S. Kovats, M. T. J. Livermore, and P. Martens. 2004. World Bank. 2010a. People, Pathogens, and Our Planet: Volume 1, Towards “Climate Change and Malaria: Analysis of the SRES Climate and Socio- a One Health Approach for Controlling Zoonotic Diseases. Washington, economic Scenarios.” Global Environmental Change 14 (1): 87–99. DC: World Bank. doi:10.1016/j.gloenvcha.2003.10.009. ———. 2010b. World Development Report 2010: Development and Climate Velthuis, A. G. J., H. W. Saatkamp, M. C. M. Mourits, A. A. de Koeijer, and A. Change. Washington DC: World Bank. R. W. Elbers. 2010. “Financial Consequences of the Dutch Bluetongue Serotype 8 Epidemics of 2006 and 2007.” Preventive Veterinary Medicine 93 (4): 294–304. doi:10.1016/j.prevetmed.2009.11.007. A G R I C U LT U R E A N D E N V I R O N M E N TA L S E R V I C E S D I S C U S S I O N PA P E R Agriculture and Environmental Services (AES) 1818 H Street, NW Washington, D.C. 20433 USA Telephone: 202-473-1000 Internet: www.worldbank.org/agriculture