Assembling evidence for identifying reservoirs of infection

Abstract

•We review the problem of identifying reservoirs of infection for multihost pathogens and provide an overview of current approaches and future directions.•We provide a conceptual framework for classifying patterns of incidence and prevalence.•We review current methods that allow us to characterise the components of reservoir-target systems.•Ecological theory offers promising new ways to prioritise populations when designing interventions.•We propose using interventions as quasi-experiments embedded in adaptive management frameworks.•Integration of data and analysis provides powerful new opportunities for studying multihost systems. Many pathogens persist in multihost systems, making the identification of infection reservoirs crucial for devising effective interventions. Here, we present a conceptual framework for classifying patterns of incidence and prevalence, and review recent scientific advances that allow us to study and manage reservoirs simultaneously. We argue that interventions can have a crucial role in enriching our mechanistic understanding of how reservoirs function and should be embedded as quasi-experimental studies in adaptive management frameworks. Single approaches to the study of reservoirs are unlikely to generate conclusive insights whereas the formal integration of data and methodologies, involving interventions, pathogen genetics, and contemporary surveillance techniques, promises to open up new opportunities to advance understanding of complex multihost systems. Many pathogens persist in multihost systems, making the identification of infection reservoirs crucial for devising effective interventions. Here, we present a conceptual framework for classifying patterns of incidence and prevalence, and review recent scientific advances that allow us to study and manage reservoirs simultaneously. We argue that interventions can have a crucial role in enriching our mechanistic understanding of how reservoirs function and should be embedded as quasi-experimental studies in adaptive management frameworks. Single approaches to the study of reservoirs are unlikely to generate conclusive insights whereas the formal integration of data and methodologies, involving interventions, pathogen genetics, and contemporary surveillance techniques, promises to open up new opportunities to advance understanding of complex multihost systems. Most disease-causing organisms, including many important human, livestock, and wildlife pathogens, are capable of infecting multiple hosts [1Cleaveland S. et al.Diseases of humans and their domestic mammals: pathogen characteristics, host range and the risk of emergence.Philos. Trans. R. Soc. Lond. B: Biol. Sci. 2001; 356: 991-999Crossref PubMed Scopus (759) Google Scholar, 2Taylor L.H. et al.Risk factors for human disease emergence.Philos. Trans. R. Soc. Lond. B: Biol. Sci. 2001; 356: 983-989Crossref PubMed Scopus (1852) Google Scholar, 3Woolhouse M.E. et al.Population biology of multihost pathogens.Science. 2001; 292: 1109-1112Crossref PubMed Scopus (572) Google Scholar]. Therefore, determining how hosts enable persistence [4Streicker D.G. et al.Differential sources of host species heterogeneity influence the transmission and control of multihost parasites.Ecol. Lett. 2013; 16: 975-984Crossref PubMed Scopus (87) Google Scholar] and which hosts are crucial for the persistence of these multihost pathogens [5Haydon D.T. et al.Identifying reservoirs of infection: a conceptual and practical challenge.Emerg. Infect. Dis. 2002; 8: 1468-1473Crossref PubMed Scopus (598) Google Scholar] is essential for the design of effective control measures. Failure to establish this understanding can hamper policy formulation and lead to ineffective or counter-productive control measures with costly implications for socially, economically, or ecologically important populations. Reservoirs of infection can be ecologically complicated structures comprising one or more interacting populations or species (Box 1 [5Haydon D.T. et al.Identifying reservoirs of infection: a conceptual and practical challenge.Emerg. Infect. Dis. 2002; 8: 1468-1473Crossref PubMed Scopus (598) Google Scholar]). Although a range of developments has led to better theoretical conceptualisation of reservoirs [5Haydon D.T. et al.Identifying reservoirs of infection: a conceptual and practical challenge.Emerg. Infect. Dis. 2002; 8: 1468-1473Crossref PubMed Scopus (598) Google Scholar, 6Ashford R.W. What it takes to be a reservoir host.Belg. J. Zool. 1997; 127: 85-90Google Scholar, 7Ashford R.W. When is a reservoir not a reservoir?.Emerg. Infect. Dis. 2003; 9: 1495-1496Crossref PubMed Scopus (69) Google Scholar, 8Nishiura H. et al.How to find natural reservoir hosts from endemic prevalence in a multi-host population: a case study of influenza in waterfowl.Epidemics. 2009; 1: 118-128Crossref PubMed Scopus (35) Google Scholar, 9Dobson A.P. of pathogens with multiple host PubMed Scopus Google their a this we review methods to characterise of the components that a reservoir to the framework in we present a conceptual for classifying patterns of incidence and prevalence that from the and populations in in We review methods that allow us to or populations or in how are in and the role that of these populations has in the pathogen reservoir reservoirs study of is the to control disease in a host or a of a et [5Haydon D.T. et al.Identifying reservoirs of infection: a conceptual and practical challenge.Emerg. Infect. Dis. 2002; 8: 1468-1473Crossref PubMed Scopus (598) Google we to this as the that are sources of infection for the are of is with to a as or more populations or in which a pathogen can be and from which infection is to the [5Haydon D.T. et al.Identifying reservoirs of infection: a conceptual and practical challenge.Emerg. Infect. Dis. 2002; 8: 1468-1473Crossref PubMed Scopus (598) Google Scholar]. reservoirs can be and a host can a more of host of these populations can the pathogen whereas infection reservoirs can be in a of Reservoirs can be wildlife species as a reservoir of in in or as a reservoir of for in species as a reservoir of for humans in many as a reservoir of for humans in the or of the species as a reservoir of for as an of the reservoir of human for of reservoirs have R.W. When is a reservoir not a reservoir?.Emerg. Infect. Dis. 2003; 9: 1495-1496Crossref PubMed Scopus (69) Google Scholar, et host PubMed Scopus Google Scholar]. Although R.W. When is a reservoir not a reservoir?.Emerg. Infect. Dis. 2003; 9: 1495-1496Crossref PubMed Scopus (69) Google Scholar] is for and et et host PubMed Scopus Google Scholar] for we et [5Haydon D.T. et al.Identifying reservoirs of infection: a conceptual and practical challenge.Emerg. Infect. Dis. 2002; 8: 1468-1473Crossref PubMed Scopus (598) Google Scholar] to not the for designing interventions. study of is the to control disease in a host or a of a et [5Haydon D.T. et al.Identifying reservoirs of infection: a conceptual and practical challenge.Emerg. Infect. Dis. 2002; 8: 1468-1473Crossref PubMed Scopus (598) Google we to this as the that are sources of infection for the are of is with to a as or more populations or in which a pathogen can be and from which infection is to the [5Haydon D.T. et al.Identifying reservoirs of infection: a conceptual and practical challenge.Emerg. Infect. Dis. 2002; 8: 1468-1473Crossref PubMed Scopus (598) Google Scholar]. reservoirs can be and a host can a more of host of these populations can the pathogen whereas infection reservoirs can be in a of Reservoirs can be wildlife species as a reservoir of in in or as a reservoir of for in species as a reservoir of for humans in many as a reservoir of for humans in the or of the species as a reservoir of for as an of the reservoir of human for of reservoirs have R.W. When is a reservoir not a reservoir?.Emerg. Infect. Dis. 2003; 9: 1495-1496Crossref PubMed Scopus (69) Google Scholar, et host PubMed Scopus Google Scholar]. Although R.W. When is a reservoir not a reservoir?.Emerg. Infect. Dis. 2003; 9: 1495-1496Crossref PubMed Scopus (69) Google Scholar] is for and et et host PubMed Scopus Google Scholar] for we et [5Haydon D.T. et al.Identifying reservoirs of infection: a conceptual and practical challenge.Emerg. Infect. Dis. 2002; 8: 1468-1473Crossref PubMed Scopus (598) Google Scholar] to not the for designing interventions. data multihost are and to et of a pathogen and wildlife host Trans. R. Soc. Lond. B: Biol. Sci. Scopus Google Scholar, et of transmission host using Biol. Sci. PubMed Scopus Google Scholar, et heterogeneity Biol. Sci. PubMed Scopus Google with the of identifying reservoirs of that data or in is unlikely to in a to control Here, we how to this the to understanding of the of reservoir-target systems. is that interventions that are to the to the and the understanding of how reservoirs with a formal integration of data and methods et for data in methods PubMed Scopus Google can provide powerful new opportunities for studying complex multihost et reservoir a case study of in the PubMed Scopus Google Scholar]). patterns of incidence and prevalence provide the and populations in the framework and framework for classifying disease Infect. Dis. PubMed Scopus Google patterns can be to 1 and in to the of the of infection from one or more in of in and the of the pathogen the of the and of disease in 1 of infection with transmission in the incidence with incidence with that the and infection disease in to a that to transmission it is incidence with in with incidence can in from with incidence from transmission in humans and in in in humans transmission of incidence with that are in the to from in to to a are present in the incidence with that are and transmission from to from in the in et the Trans. R. Soc. Lond. B: Biol. Sci. 2013; PubMed Scopus Google that in and to of is important to a of the for incidence with endemic for can be to a or a of in in in and with a of to from to in multiple in the with new from the that it is more to of disease in in in in a new the is a host from which transmission not a of the in the host is the of and are not the of infection from sources we in the with can in and as as in the pathogen is populations in which transmission can when the of infection from the is the in which an S. of and transmission heterogeneity from the of Biol. 2013; 9: PubMed Scopus Google Scholar]. these can as the the of we a in which the can to be a when it is not in which is to 1 present in their the pathogen or in the can to 1 and lead to an endemic R. et role of in the of 2003; PubMed Scopus Google Scholar]. can to is 1 the or are the a which the pathogen in the in and the a which the pathogen an endemic in the the is a in control measures the to a is and the pathogen in the and the of infection from the reservoir is is unlikely of disease in in from to an from to which for a of infection from the are to be to using patterns of incidence and prevalence data and pathogen with as can provide of the to these patterns the the of pathogens from wildlife patterns of incidence and prevalence are from or have to infection of multihost systems, as in of the S. et and for domestic as a of infection for PubMed Scopus Google Scholar, et of the patterns of infection and in Scopus Google et and in and domestic in are domestic the 9: PubMed Scopus Google and et of disease in 2009; PubMed Scopus Google in wildlife hosts in the et of reservoir species from the Dis. PubMed Scopus Google Scholar] and in in et in and in in PubMed Scopus Google Scholar]. their with to to and and with the of these the and infection a not and the of for pathogens that in the et to the of in 2013; Scopus Google Scholar]. as methods and have to of prevalence from in the of infection S. et of endemic in a population: of and in transmission PubMed Scopus Google Scholar]. Although in of techniques, as can enable of of from for et to of a wildlife from Scopus Google transmission J. et the of and for persistence in Sci. 2013; Scopus Google to which host species is the of to reservoirs that populations as or or in Therefore, we as an of persistence that can be to the reservoir framework et [5Haydon D.T. et al.Identifying reservoirs of infection: a conceptual and practical challenge.Emerg. Infect. Dis. 2002; 8: 1468-1473Crossref PubMed Scopus (598) Google Scholar]. can be as the host which a pathogen persist and R. Soc. Google Scholar, and for of the and of Scholar]. a can be as a host in which a pathogen the is whereas a is one [5Haydon D.T. et al.Identifying reservoirs of infection: a conceptual and practical challenge.Emerg. Infect. Dis. 2002; 8: 1468-1473Crossref PubMed Scopus (598) Google Scholar]. are to the study of in is to the in which persistence is to be that persistence is to the complex and it is to in the of et of in PubMed Scopus Google Scholar, J. et for infection in Scopus Google is in the of these are in natural systems. is in of is to be to the of persistence et the of the persistence of R. Soc. PubMed Scopus Google Scholar]. as the of a in which as as not a and for of the and of Scholar]. persistence be from an to the endemic of in patterns of persistence and Biol. 2001; PubMed Scopus Google Scholar, the to in R. Soc. Scopus Google Scholar]. et et we for wildlife PubMed Scopus Google Scholar] that the persistence and is not as a in a can be of in to the of a or the a of have of to in to of Scopus Google Scholar]. are the is approaches which of incidence data and R. Soc. Google Scholar, for in the R. Soc. Google Scholar, of in Biol. Sci. PubMed Scopus Google Scholar, measures of persistence and in Lett. 2013; 16: PubMed Scopus Google the to in R. Soc. Scopus Google Scholar, new the for Biol. PubMed Scopus Google Scholar, for endemic Scholar, of and many to as and new the for Biol. PubMed Scopus Google Scholar, et of in the PubMed Scopus Google and in which are to generate of persistence for populations of and from which can be of in patterns of persistence and Biol. 2001; PubMed Scopus Google Scholar, the to in R. Soc. Scopus Google Scholar]. these studies a and which is in natural et we for wildlife PubMed Scopus Google Scholar]. studies of J. of 1997; PubMed Scopus Google Scholar, et of and with the 2002; Scopus Google Scholar, and the persistence of 1997; PubMed Scopus Google has to methods have to reservoir from endemic prevalence a that is to the effective in the endemic for host populations a populations new for the to control an Biol. Sci. 2003; PubMed Scopus Google Scholar]. using and to be of the reservoir of influenza in the H. et al.How to find natural reservoir hosts from endemic prevalence in a multi-host population: a case study of influenza in waterfowl.Epidemics. 2009; 1: 118-128Crossref PubMed Scopus (35) Google Scholar]. this persistence is the and the and that are to of the have the that infection prevalence is endemic in host which populations the reservoir understanding how the populations are to Here, we approaches to transmission the reservoir and the reservoir and a population: and and to the that a is a of infection can be interventions that prevalence of disease in the or transmission incidence in the are in a is to sources of of the of is that can be for et for and in R. Soc. 2009; PubMed Scopus Google Scholar, in and Scopus Google et of and the of Biol. Sci. PubMed Scopus Google and et of and the of Biol. Sci. PubMed Scopus Google Scholar, et for disease in and populations as a case R. Soc. PubMed Scopus Google Scholar, et for Sci. PubMed Scopus Google an data that can be and the which be infection et et of and the of Biol. Sci. PubMed Scopus Google Scholar] an of persistence in the in that of humans to with a of a the of transmission to the of and the of populations that it more that from domestic in or from in the from wildlife the their the in a of transmission reservoirs of infection can be using et of in the PubMed Scopus Google Scholar, M.E. et transmission in the of R. Soc. 8: PubMed Scopus Google Scholar, et of in of the PubMed Scopus Google Scholar]. are from or and can be to the of the reservoir to these et et of in of the PubMed Scopus Google Scholar] an of interacting populations and to the to which from in in the that in the to 1 the to in that led to transmission from to in of transmission has to approaches from and et host PubMed Scopus Google Scholar, et of in 2013; PubMed Scopus Google Scholar, D.G. et and of in PubMed Scopus Google Scholar] (Box methods have for many pathogens, the of takes the as transmission are can provide to incidence and prevalence pathogen in the data should and transmission in the from involving the incidence to a of infection from the and and and D.G. et and of in PubMed Scopus Google Scholar, et multiple host of from to in 8: PubMed Scopus Google pathogen to an to the problem of multihost transmission to pathogen data et takes a in and Biol. PubMed Scopus Google Scholar]. these are from the and host the of the are to the of with a host host populations the of the reservoir to in the can be et of in 2013; PubMed Scopus Google Scholar, in of Biol. PubMed Scopus Google Scholar]. more formal of and of and of PubMed Scopus Google Scholar, J. of a and effective in using a Sci. 2001; PubMed Scopus Google can be to with pathogen et and persistence of human influenza PubMed Scopus Google approaches data open up their has the more it is to of the reservoir from transmission the using can for to the of pathogen of to transmission be pathogen from the the reservoir R. et transmission patterns of in and 8: PubMed Scopus Google of has formal is to that can have a influence the sources of infection not their be to populations in the problem to that in of populations the of and PubMed Scopus Google Scholar]). transmission can the and of host populations. the can the of and in the of this can lead to of transmission can generate for the of to the incidence in is Therefore, ways to with to in the of and the of transmission an important for future an to the problem of multihost transmission to pathogen data et takes a in and Biol. PubMed Scopus Google Scholar]. these are from the and host the of the are to the of with a host host populations the of the reservoir to in the can be et of in 2013; PubMed Scopus Google Scholar, in of Biol. PubMed Scopus Google Scholar]. more formal of and of and of PubMed Scopus Google Scholar, J. of a and effective in using a Sci. 2001; PubMed Scopus Google can be to with pathogen et and persistence of human influenza PubMed Scopus Google Scholar]. Although approaches data open up their has the more it is to of the reservoir from transmission the using can for to the of pathogen of to transmission be pathogen from the the reservoir R. et transmission patterns of in and 8: PubMed Scopus Google Scholar]. of has formal is to that can have a influence the sources of infection not their be to populations in the problem to that in of populations the of and PubMed Scopus Google Scholar]). transmission can the and of host populations. the can the of and in the of this can lead to of transmission can generate for the of to the incidence in is Therefore, ways to with to in the of and the of transmission an important for future reservoir can multiple populations of the or species in can be as a this is capable of pathogen we can a with the reservoir as a we can the of of a PubMed Scopus Google Scholar] to that of reservoir (Box is a of the of a host to pathogen persistence in the of can be to a a is that the of to can be to prioritise populations when designing interventions. framework transmission and and is to one of these when it is to the of a of of a PubMed Scopus Google Scholar] in a the of a comprising to the persistence of a species in the of we as the of a to the persistence of a can be as a of effective host to as and that influence populations and transmission of a are in which and are of infection of as as factors as and transmission in the is the of to disease and these are for this is in a persistence this a pathogen in a and to the which has a Although this of is and et of in the Scopus Google Scholar] can allow of of this is the to which are measures of the of populations to the persistence that is to interventions. the of in the to reservoir from the et et of and the of Biol. Sci. PubMed Scopus Google Scholar, et implications of the design of PubMed Scopus Google Scholar] in the Although the in is the in of and for with the range from to their to for in of the and their and are to of of a PubMed Scopus Google Scholar] in a the of a comprising to the persistence of a species in the of we as the of a to the persistence of a can be as a of effective host to as and that influence populations and transmission of a are in which and are of infection of as as factors as and transmission in the is the of to disease and these are for this is in a persistence this a pathogen in a and to the which has a Although this of is and et of in the Scopus Google Scholar] can allow of of this is the to which are measures of the of populations to the persistence that is to interventions. the of in the to reservoir from the et et of and the of Biol. Sci. PubMed Scopus Google Scholar, et implications of the design of PubMed Scopus Google Scholar] in the Although the in is the in of and for with the range from to their to one of is to the and of a reservoir are interventions embedded adaptive management and data are to the of disease control management for Sci. 2002; Google Scholar]. be with an understanding of the disease We argue that using interventions as quasi-experiments can provide opportunities to more the of a reservoir in adaptive management and Scholar, et of and of disease control can be and enriching the to future control and studies host of species a from hosts for et of complex from in PubMed Scopus Google in to in of in the and of in and to of and in prevalence in and incidence in et of control of of infection in and populations from J. PubMed Scopus Google of for in et of the natural to the control of in Biol. Sci. 2013; PubMed Scopus Google and in a control in in et prevalence of in to control disease in Infect. PubMed Scopus Google and control in and prevalence in et al.Identifying disease reservoirs in complex as reservoirs of and pathogens of 2003; Scopus Google in in in human incidence reservoir hosts and control of 2009; PubMed Scopus Google R. et of the transmission of in Infect. Dis. 1997; PubMed Scopus Google and of from from et of from in Dis. 2001; PubMed Scopus Google and of J. review of in livestock, and of in of the et of in with Dis. PubMed Scopus Google of with to in and et the Trans. R. Soc. Lond. B: Biol. Sci. 2013; PubMed Scopus Google in a new that generate more to the are the to provide interventions can be to or or that are to for Google or transmission reservoir and is can be from or analysis H. and Scholar]. that can be for allow us to a transmission or What can be from interventions is enable to be is important to that these interventions can the transmission to in and of the the and a et and of in PubMed Scopus Google Scholar] a of data and analysis that should be in determining the in which to the effective control of in the it not be to in the disease transmission a or to and of the in using adaptive management in that future control with the current Scopus Google Scholar]. be that an that the one in the of disease prevalence in the is the one that provides the to the of can be a the of disease control and the of the to future is the of to the to control et not PubMed Scopus Google and the of a PubMed Scopus Google for the et of and of Scholar, reservoir hosts and control of 2009; PubMed Scopus Google Scholar]. with the of disease in the of an adaptive management to from the et to from of Scholar]. a of the the of and from the and to the of and provide insights understanding reservoirs the of multiple data of can the analysis or study design J. and and Scopus Google Scholar] or using as or reservoir hosts and control of 2009; PubMed Scopus Google Scholar]). of multiple sources should understanding of the and of et and a for 16: PubMed Scopus Google Scholar]. of et et reservoir a case study of in the PubMed Scopus Google Scholar] for the that domestic the for in integration of case data that can persist in domestic data that a a range of and analysis of incidence patterns that from domestic populations of transmission in with in we a formal integration of sources of that can be to characterise reservoir systems. approaches are powerful methods are and data can be to enable of transmission and host populations et the and of in domestic as a PubMed Scopus Google Scholar, et patterns of PubMed Scopus Google and can be with and data in a framework that of the and of transmission et of in 2013; PubMed Scopus Google Scholar, et transmission and identifying the Trans. R. Soc. Lond. B: Biol. Sci. 2013; PubMed Scopus Google Scholar]). and et et transmission and identifying the Trans. R. Soc. Lond. B: Biol. Sci. 2013; PubMed Scopus Google Scholar] the transmission of and and transmission that and are to data when are et to of a wildlife from Scopus Google Scholar]. methods that multiple data is a and to future in disease et of disease and Biol. PubMed Scopus Google Scholar, et for the of endemic from Biol. Sci. PubMed Scopus Google Scholar]. in the study of reservoirs of are powerful to characterise components of systems, are that to be determining persistence for host and the of we in which in effective control in the of data and data including their integration with theory as reservoir promising new ways to these of interventions as quasi-experiments can provide approaches to are as PubMed Scopus Google Scholar] that multiple to host populations as to a this how to design interventions that allow us to the of the of control to infection in one and how to the of the with with and Although are powerful to characterise components of systems, are that to be determining persistence for host and the of we in which in effective control in the of data and data including their integration with theory as reservoir promising new ways to these of interventions as quasi-experiments can provide approaches to are as PubMed Scopus Google Scholar] that multiple to host populations as to a this how to design interventions that allow us to the of the of control to infection in one and how to the of the with with and of the approaches provides important of these are of a more when in a understanding of how reservoirs interventions can provide of disease control the and important opportunities that can advance understanding of reservoir to these we with their range of from the and the of the reservoir and function not an integration of approaches to data and a in the that their with and human has in a the of the and of the of and of We are for and that We and for this is a from is is is and the in is the and is the and of or is for and not the of a in a host which a disease persist in the of in a an infection is endemic in a when it is the for of infection from a to host in a of new in a of host that can a pathogen the is a of which are a is a host capable of a pathogen the of populations that are can populations of the species in populations of or a of the of populations to the reservoir of a of in a a of the of a host to pathogen persistence in the of one or more populations or in which a pathogen can be and from which infection is to the of in the of when transmission hosts is to endemic or