Intervention Strategies for Carcass Disposal: Pareto Analysis of Exposures for Exotic Disease...

Intervention Strategies for CarcassDisposal: Pareto Analysis ofExposures for Exotic DiseaseOutbreaksJ O A O D E L G A D O , † P H I L L O N G H U R S T , †

G O R D O N A . W . H I C K M A N , ‡

D A N I E L M . G A U N T L E T T , ‡

S I M O N F . H O W S O N , ‡ P H I L I R V I N G , §

A L W Y N H A R T , | A N DS I M O N J . T . P O L L A R D * , †

Cranfield University, Collaborative Centre of Excellence inUnderstanding and Managing Natural and EnvironmentalRisks, Cranfield, Bedfordshire, U.K., MK43 0AL, AnimalHealth Agency, Contingency Planning Division, Departmentfor Food, Environment & Rural Affairs, Nobel House, 17Smith Square, London, U.K., SW1P 3JR, and EnvironmentAgency, King’s Meadow House, King’s Meadow Road, Reading,Berkshire, U.K., RG1 8DQ, and Olton Court, 10 WarwickRoad, Solihull, U.K., B92 7HX

Received January 6, 2010. Revised manuscript receivedApril 23, 2010. Accepted May 4, 2010.

An enhanced methodology for the policy-level prioritizationof intervention options during carcass disposal is presented.Pareto charts provide a semiquantitative analysis of opportunitiesfor multiple exposures to human health, animal health, andthe wider environment during carcass disposal; they identifycritical control points for risk management and assist in wastetechnology assessment. Eighty percent of the total availabilityof more than 1300 potential exposures to human, animal, orenvironmental receptors is represented by 16 processes, thesebeing dominated by on-farm collection and carcass processing,reinforcing the criticality of effective controls during earlystages of animal culling and waste processing. Exposures duringmass burials are dominated by ground- and surface-waterexposures with noise and odor nuisance prevalent for masspyres,consistentwithU.K.experience.Paretochartsarediscussedin the context of other visualization formats for policy officialsand promoted as a communication tool for informing the site-specific risk assessments required during the operational phasesof exotic disease outbreaks.

IntroductionRationale. Increased prevalence of exotic animal disease (e.g.,foot and mouth disease, avian influenza, Newcastle disease,classical swine fever, African swine fever, swine vesiculardisease) is a strategic threat to industrialized and developingcountries and their economies. Most controls strategies arebased on a “stamping out” policy and involve culling ofaffected livestock and the safe disposal of the carcasses. Theextent of carcass disposal operations during outbreaks

challenges the waste management infrastructure of thecountry concerned, posing risks to animal and public healthand to the environment (1-3). The policy level analysis ofthese disposal risks (evaluated as potential exposures tohuman, animal, and environmental health) adopts a gen-eralized, qualitative tone, in which hazardous agents inherentto carcass disposal are tracked through a series of unitprocesses (and associated exposure pathways) to sensitivereceptors, be these animal stock, the public or the widerenvironment (Figure 1). Our research (4) has sought toenhance the utility of generalized policy-level exposureassessments (referred to by Andrews et al. (5) as “macrostudies”) and, by informing contingency planning, improvethe pollution control strategies employed during outbreaks,thus reducing the environmental, social, and economicburden of these events.

We have previously described the policy and internationalcontext for our research and published a methodology forgeneralized exposure assessment, illustrating it for the avianinfluenza virus (4). Exotic animal disease outbreaks featureas a key strategic risk within Great Britain (6), an island nationwith a high population density where carcass disposal occursin close proximity to communities within close public andmedia view. When outbreaks do occur, large numbers ofcarcasses require rapid and responsible disposal to avoidthe onward transmission of disease to animals and humansand to prevent environmental harm. Given the challenges ofmanaging this in practice, Great Britain has developed ahierarchy of disposal options that retains some flexibilityover the specific choice of waste processing options. Thissaid, without targeted risk management, social and economiccosts escalate rapidly. The Anderson (7) report, for example,estimated the costs to the U.K. government alone of the 2001foot and mouth disease outbreak to be £3 billion, employing1800 vets and 2000 military personnel at its peak. Duringdisposal, the range of hazardous agents requiring activemanagement extends well beyond the agents of disease toinclude all hazards associated with a carcass, its byproducts,and the infrastructure of disposal (detergents, disinfectants,veterinary medicines, other pathogens, odor, and noise). Tobe cost-effective, risk management must tackle the mostpotent hazardous agents, be focused upon exposure pathwaysthat exhibit the greatest availability to the significant hazards,and address the most sensitive receptors, targeted at criticalcontrol points (CCPs) where harm reduction can be mosteffective (Figure 1). When used as part of good contingencyplanning and a preventative risk management approach, suchan approach can improve the direction of resources and thequality of management responses on the ground whenoutbreaks occur.

* Corresponding author e-mail: [email protected]; phone:+44(0)1234 754101; fax: +44(0)1234 751671.

† Cranfield University.‡ Animal Health Agency.§ Environment Agency, King’s Meadow House.| Environment Agency, Olton Court.

FIGURE 1. Key determinants considered in this generalizedexposure assessment for animal carcasses. Interventions mayprevent release or intercept secondary pathways to receptors.Generalized exposure assessment at the policy level cannotevaluate harm (receptor components) in that exposures are notsite-specific nor contextualized locally in time and space.

Environ. Sci. Technol. 2010, 44, 4416–4425

4416 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 12, 2010 10.1021/es100039n 2010 American Chemical SocietyPublished on Web 05/19/2010

Here we present an improved exposure methodology,expand its application to a wider range of disease agents,and present a Pareto analysis of the critical exposure pathwaysfor onward control. The revised method addresses some ofthe practical shortcomings inherent to eliciting expertknowledge in exposure workshops, especially given that forpolicy analysis, quantitative exposure data is frequently notavailable (risks are not spatially nor temporally contextualisedat this level). Carcass disposal is a multistage process chain,with elements in common with other multistage processes(e.g., food manufacture) where opportunities for interventionmust be risk-informed. Clear differences also exist, notablywith respect to the open, heterogeneous nature of the naturalenvironment, compared to the closed, batch, or continuousmanufacturing environment of a food processing plant.Nevertheless, the desire to identify risk-critical exposurepathways is a shared objective so to inform preventativecontrols.

MethodologyMethodological Developments. A generalized exposureassessment has been described for highly pathogenic avianinfluenza virus. It adopts an heuristic assessment of therelative availability of exposure pathways for each significanthazard using a nominal 1-4 scale. Availability “scores” aresurrogates for the relative likelihood of a significant hazardousagent accessing that specific exposure pathway. So a scoreof 1 indicates a plausible pathway, but one of negligible tovery low availability; 2 represents a plausible pathway of lowavailability; 3 represents a possible pathway for which thereis accepted evidence and medium availability; and 4 rep-resents a probable, direct pathway of high recognizedavailability (4). Though admittedly constrained by thisnominal classification of relative exposure probabilities, theapproach supports decision-makers and presents exposureinformation in a simple, defensible format, reflecting thecomplexity of the problem and richness of the supporting(elicited) evidence. It has been successfully used to direct

detailed quantified risk assessments toward particular ex-posure scenarios on the ground in the heat of an outbreak.

The revised methodology was developed using an expertworkshop held at Cranfield University, in January, 2008. Theworkshop considered a fuller range of notifiable exotic animaldiseases (the original approach was limited to avian influ-enza), including foot and mouth disease (FMD), classicalswine fever (CSF), and Newcastle disease, and incorporatedsome methodological improvements. Technical domainspecialists (e.g., veterinary specialists, public health experts,environmental exposure assessors, waste technologists,hydrogeologist, air quality experts) were assembled fromDefra, the Animal Health Agency, the Environment Agency,and the Scottish Environment Protection Agency and theHealth Protection Agency, alongside the research team. Theresearch agenda within the workshop was concerned with(i) improving the record of decision on the hazard screeningexercise for various diseases, (ii) distinguishing betweenimpacts for animal health, public health, and the environ-ment, (iii) collating waste technologies into categories thatallowed greater consistency of treatment in the exposureassessment (Figure 2), which is a extensive combinatorialexercise, and (iv) reviewing the metrics of the exposureassessment with a view to improving the utility of the datafor contingency planning.

The workshop structure uses a sequence of expertelicitations. Expert knowledge on the likelihood of exposurefor the significant agents along individual process chainswas secured and presented within the workshop for a criticalcomparison of each process chain (Figure 2). Quality checkingof all pathway scores was undertaken for the common stagesof each pathway (Figure 2) and then for the total pathwayscore, first by the experts within the workshop and thenindependently off line to check consistency within andbetween process chain. Using this approach, experts con-tribute individual domain expertise and also to a collectiveview for each disposal pathway, with an independent “armslength” quality audit being completed shortly after the

FIGURE 2. Suite of processes considered for carcass disposal. Disposal is envisaged as a multistage process from on-premisescollection to the elimination of residuals (column headings). Individual disposal options (boxed numbers) adopt individualconfigurations comprising these stages. Options 1 (do nothing), 2a (on-farm storage), and 2b (off-farm storage) were excluded fromthe figure because they do not adopt the five step disposal chain. Key: (3) on-farm burial, (4a) on-farm pyres, (6a) windrowcomposting on premises, (8a) on-farm in-vessel composting, (4b) off-farm mass pyres, (5) mass burial/inert controlled landfill, (6b)windrow composting off-premises, (9) air-curtain/mobile incinerators, (7) nonhazardous/hazardous controlled landfill, (8b) off-farm,biogas/anaerobic digestor/in-vessel composting, (10) pressurized or atmospheric rendering or alkaline hydrolysis, and (11)gasification and pyrolysis, licensed controlled incinerators.

VOL. 44, NO. 12, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 4417

workshop. A full hazard screen for significant hazards (Table1) and initial exposure assessment was completed on 14-15January, 2008. Subsequently, a subgroup met to finalize theexposure assessment and recommend improvements to thepresentation of the exposure metrics, which were held in an

Excel spreadsheet. The quality audit was undertaken on themetrics (February 2008) by members independent to theresearch team. Recommendations from this audit wereimplemented and a final exposure assessment prepared that(1) modified the hazardous agents list to comprise a wider

TABLE 1. Significant Hazards Passing All Three Filters and Posing Potential Impacts to Human Health (H), Animal Health (A), orEnvironmental (E) Receptorsa

a Shading indicates hazard not relevant for this disease.

4418 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 12, 2010

set of etiological agents, including avian influenza (controlagainst initial methodology), Newcastle disease (poultry),foot and mouth disease (cattle, pigs, sheep), bovine spongi-form encephalopathy (cattle), scrapie (sheep), swine vesiculardisease and African and classical swine fever (pigs); (2)specified to which receptor (animal health, public health,the environment) the hazard was likely to present significantconcern, redefining animal health impacts as solely fordomestic and livestock animals (Table 1); (3) specified forwhich receptor exposure was most likely to be of concern;and (4) formally captured qualifying comments from theexpert group (Table 1).

Having identified the significant hazards of carcassdisposal, the methodology does not attempt to distinguishfurther between the toxicological/pathogenic potencies ofthose hazards deemed by experts to be significant (Table 1),though the relevance of the potential harm posed to animalhealth, public health or the environment was considered byreference to the availability of the pathway to these receptors.

Exposure Assessment and Pareto Analysis. The process-ing of large numbers of elicited pathway availabilities formultiple agents and associated hazards may introduceinconsistencies in expert views as they apply their domainknowledge between similar unit processes for each disease;in short, the task of processing multiple combinations ofhazardous agents and exposure pathways for human, animaland environmental receptors becomes highly complex forparticipants because of the large number of combinationsinvolved. Improvements were made to the exposure assess-ment methodology for ease of data processing and visual-

ization of the key drivers of exposure within the spreadsheet:(i) the introduction of a benchmarked “base case” for eachprocess chain against which each process could be evaluated;(ii) restructuring the model to collate each of the disposaloptions and present them as a single worksheet, allowingpathways (Table 2) to be directly compared, securing a greaterlevel of consistency; (iii) presentation of individual pathwayscores within the model to allow a direct comparison of theexposure metrics for each disposal route.

Effective risk management infers the capacity to interveneat critical control points, thus using resources wisely fortargeted optimal risk reduction. Here, we evaluated pathwayscontributing most to the overall exposure using a Paretoanalysis (8). The approach has been described in themanagement literature by Frakes and Fox (9) and Omachonuand Ross (10), and exemplified for multibarrier systemscomprising multiple unit processes within the food industryby Arvanitoyannis and colleagues (11). By convention Paretocharts consist of histograms, the lengths of which areproportional to the arithmetic mean exposure (y-axis),organized from left to right from highest to smallestcontribution (x-axis). The right-hand y axis represents thecumulative contribution to overall exposure. Consistentwith the Pareto principle, a nominal threshold value wasset at 80% (black vertical line), where the group of pathwaysto the left of the line is responsible, in aggregate, for 80%of the exposure pathway availability, as ranked by workshopexperts.

TABLE 2. Full Suite of Exposure Pathways Considered

medium pathways abbreviation

above ground pathways associated withairborne and vectortransmission (wildanimal). Includes amenityimpacts

deposition onto crops and consumption AG dep crop consdirect contact AG dirc contingestion AG inginhalation AG inhnoise (nuisance) noiseodor (nuisance) odorsmoke (nuisance) smokewild animal consumption AG wild conswild animals (spread) AG wild sprd

surface water pathways associated withthe contamination ofsurface waters by runoff(leakage)

consumption of animals SW anim conscrop irrigation and consumption SW crop irrdirect contact SW dirc contfish or shellfish consumption SW Shellfishingestion of water SW inginhalation SW inhinhalation of irrigation water SW inh wtrleaching from SW to GW SW to GW

groundwater pathways associated withcontamination orgroundwater by leakege,infiltration and transportthrough soil;differentiation betweenpathways linked toprivated water supplies(private) and public watersupplies

consumption of animals (private) GW anim cons (p)crop irrigation and consumption (private) GW crop irr (p)direct contact (private) GW dirc cont (p)fish or shellfish consumption (private) GW shellfish (p)ingestion of water (private) GW ing (p)inhalation (private) GW inh (p)inhalation of irrigation water (private) GW inh wtr (p)consumption of animals GW anim conscrop irrigation and consumption GW crop irrdirect contact GW dirc contfish or shellfish consumption GW shellfishingestion of water GW inginhalation GW inhinhalation of irrigation water GW inh wtr

ground to surface water pathways associated withsurface watercontaminated by surfacingof contaminatedgroundwater

consumption of animals GS anim conscrop irrigation and consumption GS crop irrdirect contact GS dirc contfish or shellfish consumption GS Shellfishingestion of water GS inginhalation GS inhinhalation of irrigation water GS inh wtr


Results and DiscussionPareto Analysis. Table 1 presents the list of significant hazardsidentified by the expert workshop. Figure 2 provides a mapof the unit processes, many of which are common for therange of disposal options considered. Table 2 presents thefull set of exposure pathways considered in this analysis acrossall disposal options and unit processes. Using exposurepathway “availabilities” elicited in the expert workshop, aset of a Pareto charts were constructed for each of theprocesses represented in Figure 2 (numbered) by referenceto the distribution of exposures at each of the 5 stages (Figure3), and then by the individual exposure pathways associatedwith each of the 5 stages of carcass disposal (Figures 4 and5). This allows screening and prioritization to be applied attwo different levels: (i) at the process chain level, prioritizingprocess chains according to the potential risk of exposure,and identifying critical control points for managing exposuresalong the process chain, and (ii) of key exposure routes withineach process step, allowing pathways to be consideredirrespective of disposal route.

Consider Figure 3 summarizing those processing stagescontributing most to exposure across the full set of carcassdisposal options. The (waste) processing of carcasses usingon-farm pyres (option 4b) presents the greatest potential forexposure, followed by the collection of carcasses on farm foronward burial or pyre disposal (Figure 3). This approachrepresents a conceptual move toward a modular, “object-oriented” approach to the exposure assessment. Here, 80%of the total availability of onward exposures to human, animal,or environmental receptors is represented by the 16 processesincluding and above process 7 in Figure 3; being dominatedby on-farm collection and carcass processing, clearly il-lustrating the prime necessity of effective controls duringthe early stages of culling and waste processing.

Next consider Figure 4a-e, which in concert, allows avisual comparison of the importance of all those exposuresrepresented in Table 2 across each stage of disposal in turn(Figure 2) for the disposal option of mass pyres (option 4b).Ground- and surface-water exposure pathways dominate onpremises, with noise and odor nuisance prevalent for mass

pyres, consistent with the experience that contaminationcan have origins in multiple sources in the early stages ofdisposal (12-15). Quite obviously, these pathways arecandidates for control and reflect our experiences duringoutbreak management. By illustration, among the specificenvironmental impacts reported part-way through the 2001foot and mouth disease outbreak requiring effective man-agement (13, 16) were (i) a large number (∼200) of reportedwater pollution incidents from the surface runoff of bloodand carcass fluids early on in the crisis, when culling ratesoutstripped disposal capacity; though few of these resultedin significant water pollution (only 3 high category pollutionincidents from slurry spill and disinfectant runoff), (ii) thegeneration of large quantities of ash from constructed animalpyres (typically 15 tonne ash per 300 t pyre) required onwardcontainment and disposal, and (iii) localized, short-termderogations in local air quality during pyre burning (17, 18).

Contrast this analysis with the Pareto charts prepared forfixed plant thermal treatment options (evaluated in concertas gasification, pyrolysis, and licensed controlled incinerators(Figure 5a-e). The exposures generated while carcasses aregathered on premises remain with a reduction in exposurein the later stages of disposal compared to the mass pyreoption; for example, reception and processing of carcassesand elimination of residuals, can be observed. In the U.K.,waste incineration installations must operate within stan-dards required by the waste incinerator directive (19). TheWID sets benchmark criteria for reception and processing ofwaste and through the landfill directive the elimination ofresiduals. The tougher controls imply that “incinerationplants” present a higher level of technical sophistication andprotection (20). Despite this, aboveground exposure path-ways, smoke and noise are worthy of consideration.

Communicating Risk and Designing Interventions withPolicy Experts. This work represents the next stage in thechronological development of methods used to communicatethe extent and significance of exposures during carcassdisposal. The visualization of conceptual models of exposureand of summary data for policy level decisions are criticalif compelling arguments for policy improvements are to be

FIGURE 3. Screening and prioritizing carcass disposal steps, independently of process category. Key: (3) on-farm burial, (4a) on-farmpyres, (6a) windrow composting on premises, (8a) on-farm in-vessel composting, (4b) off-farm mass pyres, (5) mass burial/inertcontrolled landfill, (6b) windrow composting off-premises, (9) air-curtain/mobile incinerators, (7) nonhazardous/hazardous controlledlandfill, (8b) off-farm, biogas/anaerobic digestor/in-vessel composting, (10) pressurized or atmospheric rendering or alkalinehydrolysis, and (11) gasification and pyrolysis, licensed controlled incinerators.


justified. Early communications adopted qualitative reason-ing and summary schematics to compare the waste tech-nologies used for carcass disposal (18). Rapid qualitative riskassessment, backed up by selective quantitative backgroundstudies on air pollution, for example, was used to comparethe number of exposure pathways associated with specific

disposal options (18); (Figure 6); the number of pathwaysacting as a surrogate for the extent of public and environ-mental risks. Since Pollard et al. (4), we have move beyonda discussion of waste technology in isolation and haveconsidered the full process chain of disposal, ranking therelative availability of exposure pathways, and allowing a

FIGURE 4. Pareto charts (for illustration) of the mass pyre disposal chain. Five charts are displayed, one for each step: (a) onpremises, (b) transport, (c) reception of carcasses, (d) processing of carcasses, and (e) elimination of residuals. The horizontal axispresents the abbreviated reference for each of the pathways (see Table 2).


qualitative comparison of the distribution of exposure forhazards associated with avian influenza from farm throughto the disposal of residuals (4). Here, we now offer an object-oriented approach that seeks to present, semiquantitatively,the distribution and contribution of exposures both acrossthe process chain and by environmental pathway irrespectiveof the unit process, so allowing environmental regulators to

focus on specific environmental receptors during an out-break. We also extend the analysis to a broader suite of exoticdiseases.

The use of Pareto charts assists in the synthesis of thesegeneralized, policy-level exposure assessments and acts asan important communication tool for policy makers, allowingexposure comparisons between discrete waste processing

FIGURE 5. Pareto charts (for illustration) of the controlled incineration disposal chain. Five charts are displayed, one for each step:(a) on premises, (b) transport, (c) reception of carcasses, (d) processing of carcasses, and (e) elimination of residuals. The horizontalaxis presents the abbreviated reference for each of the pathways (see Table 2).


options, but importantly, also along the length of disposalchains. Qualitative comparisons can, for example, be ob-served between exposures associated with collection onpremises/transport and active carcass processing/residualsmanagement. By making the distribution of exposure avail-abilities visible, we bring an improved level of understandingon exposure across the process chain to the debate, allowingpolicy officials to direct risk management efforts accordinglyand communicate a rationale for intervention priorities. Ofcourse, decision-makers must integrate the analytical insightsthat exercises such as this one generate within the contextof U.K., European Union, and International standards andnorms. The relationship between analysts and decision-makers and the associated challenges of risk-informed policydesign and delivery are well characterized (21, 22, 22). Thissaid, this analysis offers a firm, evidence-based assessmentof the disposal options required in a modern policy andregulatory environment. The principal policy benefit fromthis work has been in operationalizing Defra’s contingencyplans during outbreaks. While these are published andregularly reviewed, the majority of stakeholders and inter-ested parties read them only in the event of an outbreakwhen it impacts on their own operational activities. Defrahas used this research to support its stated position that therisks to human health, animal health, and the environmentfrom our chosen disposal routes are minimal. This researchhas provided an evidence base to support a single disposalhierarchy in Defra’s published plans. Without this, it wouldhave been difficult to state, with any authority, whetherexisting controls were adequate. Defra is now consideringextending the principles and practice communicated in this

work to the development of a framework for assessing therisks of carcass disposal for new and emerging diseases. Defrabelieves the research also has relevance to the evaluation ofother putrescible wastes that may need to be disposed; suchas, for instance, during a terrorist attack or accidentaldischarge in a chemical, biological, radiological, or nuclear(CBRN) incident. The research clearly highlights the largenumber of high risk pathways during the initial phases ofcarcass disposal and during the collection and loading ofcarcasss on farm. This has resulted in a review of operationalguidance and modifications to the guidance on selection ofculling sites, rapid removal of carcasss, containment ofwashwater, and the loading of carcasss into leak-tested trucks.Defra has also used this research to develop a “landfillprotocol” for the disposal of carcasss in permitted landfillsites.

These implementation benefits are encouraging. The toolsdeveloped in this research embody well recognized limita-tions and are but one way of communicating multiplecontributions to exposure in a complex decision environ-ment. Our selection has been made by reference to variousalternative tools, some of which have been discarded on thebasis that either they were not appropriate for the studycontext or because the data produced during the elicitationworkshop was not of the relevant format for correct ap-plication of the tool. Other options, for reference, includethe consideration of so-called Ishikawa (fishbone) diagrams,which emerged from the discipline of production manage-ment, adopted as quality control tools within the manufac-turing sector (23, 24). Here, a fishbone graphic is used toexplore and display all potentially faulty processes, included

FIGURE 6. Contribution of multiple sources of hazards, pathways, and receptors to an exposure pool for onward Pareto analysis.


in manufacture that may lead to a defective product. Ishikawadiagrams can also be used alongside fault tree analyses orfailure mode effects analysis tools, where multiple eventscontribute to a release or a faulty product, a contaminatedfood product, for example. However, they are perhaps bestsuited to closed engineered systems where defined productlines, process flowsheets, and pathways are characterizedby materials flow in formalized product lines or via contained(e.g., piped) infrastructure. Here, rather, we were concernedwith the release of hazardous chemical and biological agentsat process points within an open heterogeneous system (theenvironment) and in assessing multiple exposures to a rangeof receptors along a process chain interacting with the widerenvironment. In this study, the distribution of opportunitiesfor exposures (of differing probability; Figure 6) is criticalhence the selection of the Pareto chart. In closed systems bycontrast, as Arvanitoyannis (11) displays for food manufactureand the contamination events that might spoil food products,there is the potential to adopt these tools in a single study.With concerns about the methodological biases inherent tothe use of risk tools designed for closed systems being appliedto open environmental systems, we discarded application offault and event tress in this particular case, although othershave adopted these, say for BSE prion exposures viaenvironmental pathways (25). In our study, 38 exposurepathways (Table 2) had to be assessed for each of 35 disposalsteps defined (Figure 2); this requiring the assessment ofmore than 1300 pathways of exposure. Even with a nominalranking approach, workshop facilitators needed to work toensure that expert fatigue from processing the sheer numberof combinations did not influence negatively the quality ofthe results. With these choices made and as configured, theformat of the output does not yet allow analysts to (1) inferthe importance of one individual hazardous agent aboveanother, (2) identify the source of the hazardous agent, or(iii) evaluate the harm to human receptors, animals, or thesurrounding environment. These are requirements of site-specific risk analysis in the operational setting that can beinformed by high level analysis but not determined at thismacro-level.

Notwithstanding this critique, the limitations and theavailability of other tools that might also be applied tocommunicate these messages, we believe this to be the firstapplication of Pareto analysis to generalized policy-levelexposure assessments. We reassert the view that risks toreceptors during carcass disposal activities are not confinedto the waste processing of animal carcasses in isolation, butrather present from farm until ultimate disposal of residuals.Pareto analyses can be used to prioritize exposure pathwaysaccording to their contribution to the overall exposurepotential across the disposal chain (Figure 6) and can assistin communicating the complexity of a multiattribute problemto decision-makers. Adopting an object-oriented approachallows efficiencies to be secured in both the analysis itselfand in the design of risk management measures that havecommon features irrespective of the waste processing option.Measures that address priority pathways during disposal arelikely to be most effective when a range of disposal strategiesis applied from a hierarchy of options, which is a necessityin Great Britain given the structure of the waste managementsector, the legislative constraints on certain options and theintense and close public and media scrutiny that ac-companies disease outbreaks. These are important consid-erations for national policy makers and international orga-nizations seeking to develop guidance on the processing ofanimal carcasses (26).

AcknowledgmentsWe thank Defra and the Animal Health Agency for fundingthis work and the expert participants from the Health

Protection Agency, Environment Agency, Scottish Environ-ment Protection Agency, Welsh Assembly, Scottish Executiveand the Department of Health. Peter Tangney and Jens Evans(Environment Agency) conducted an independent audit ofthe exposure assessment and Sophie Trombini (MSc, Cran-field University 2007/8) assisted with formatting the Paretoanalysis. JD is funded by Cranfield University. The Risk Centreis funded by Defra, EPSRC, NERC and ESRC under grantEP/G022682/1.

Literature Cited(1) Scudamore, J.; Trevelyan, G.; Tas, M.; Varley, E.; Hickman, G.

Carcass disposal: Lessons from Great Britain following the footand mouth disease outbreaks of 2001. Rev. Sci. Technol. Off.Epizoot. 2002, 21 (3), 775–784.

(2) Salminen, E.; Rintala, J. Anaerobic digestion of organic solidpoultry slaughterhouse waste-a review. Bioresour. Technol. 2002,83 (1), 13–26.

(3) Arvanitoyannis, I. S.; Ladas, D. Meat waste treatment methodsand potential uses. Int. J. Food Sci. Technol. 2008, 43 (3), 543–559.

(4) Pollard, S. J. T.; Hickman, G. A. W.; Irving, P.; Hough, R. L.;Gauntlett, D. M.; Howson, S. F.; Hart, A.; Gayford, P.; Gent, N.Exposure assessment of carcass disposal options in the eventof a notifiable exotic animal disease: Application to avianinfluenza virus. Environ. Sci. Technol. 2008, 42 (9), 3145–3154.

(5) Andrews, C., Apul, D. Linkov, I. Comparative risk assessment:past experience, current trends and future directions. InComparative Risk Assessment and Environmental DecisionMaking, Linkov, I. Bakr Ramadan, A., eds.; Kluwer AcademicPublishers: Leiden, Netherlands, 2004; pp 3-14.

(6) Cabinet Office. The National Security Strategy of the UnitedKingdom. Security in an Independent World; Command Paper7291; The Stationery Office: Norwich, U.K., 2008.

(7) Anderson, I. Foot and Mouth Disease 2001: Lessons to be LearnedInquiry; Report HC888; Stationery Office: London, 2002; 187pp.

(8) Craft, R. C.; Leake, C. The Pareto principle in organizationaldecision making. Manage. Decis. 2002, 40 (8), 729–733.

(9) Frakes, W. B.; Fox, C. J. Quality improvement using a softwarereuse failure modes model. IEEE Trans. Software Eng. 1996, 22(4), 274–279.

(10) Omachonu, V. K. Ross, J. E. Principles of Total Quality; CRCPress: Boca Raton, FL, 2004.

(11) Arvanitoyannis, I. S.; Varzakas, T. H. Application of failure modeand effect analysis (FMEA), cause and effect analysis and Paretodiagram in conjunction with HACCP to a potato chips manu-facturing plant. Int. J. Food Sci. Technol. 2007, 42 (12), 1424–1442.

(12) Ritter, W. F.; Chirnside, A. E. M. Impact of dead bird disposalpits on groundwater quality on the Delmarva Peninsula.Bioresour. Technol. 1995, 53 (2), 105–111.

(13) Environment Agency. Disposal of Culled Stock by Burial:Guidance and Reference Data for the Protection of ControlledWaters; Draft R&D Technical Report, version 8; EnvironmentAgency: Bristol, U.K., 2001.

(14) Cumby, T. R.; Adkin, A.; Burfoot, D.; Burton, C. H.; Keel, P.;Lyne, A.; Munday, D.; Parkin, S.; Turner, C. Williams, S.Prevention and Control of Animal Diseases: Review of Decon-tamination Techniques, SE4001; Silsoe Research Institute: Silsoe,Bedford, U.K., 2005.

(15) Marsland, P. A., Smith, J. W. N. Young, C. P. Foot and MouthDisease Epidemic. Disposal of Culled Stock by Burial: Guidanceand Reference Data for the Protection of Controlled Waters;National Groundwater and Contaminated Land Centre reportNC/02/04/01; Environment Agency and Water Research Centre(WRc): Bristol, U.K., 2003.

(16) Environment Agency. The Environmental Impact of the Footand Mouth Disease Outbreak: An Interim Assessment; Environ-ment Agency: Bristol, U.K., 2001.

(17) Lowles, I.; Hill, R.; Auld, V.; Stewart, H.; Colhoun, C. Monitoringthe pollution from a pyre used to destroy animal carcasses duringthe outbreak of Foot and Mouth Disease in Cumbria, UnitedKingdom. Atmos. Environ. 2002, 36 (17), 2901–2905.

(18) Department of Health. A Rapid Qualitative Assessment of PossibleRisks to Public Health from Current Foot & Mouth DisposalOptions; Main Report DH_4014754; Department of Health:London, U.K., 2001.

(19) Department for Environment Food and Rural Affairs. TheDirective in the Incineration of Waste: Environment Permitting


Guidance; Version 2.0; Defra: London, 2009; Available at http://www.defra.gov.uk/environment/policy/permits/documents/wid-guidance.pdf.

(20) Department for Environmental Food and Rural Affairs. Con-tingency Plan for Exotic Animal Diseases; Framework ResponsePlan, Version 3; Defra: London, 2008.

(21) Sawicki, D. S.; Craig, W. J. The democratization of data: Bridgingthe gap for community groups. J. Am. Plann. Assoc. 1996, 62 (4),512–523.

(22) Andrews, C. J. Humble Analysis: The Practice of Joint Fact-Finding; Praeger Publishers: Westport, CT, 2002.

(23) Cantu-Ortiz, F. J. Expert systems in manufacturing: An experi-ence in Mexico. Expert Syst. Appl. 1991, 3 (4), 445–455.

(24) Schonberger, R. J. Japanese production management: AnevolutionsWith mixed success. J. Oper. Manage. 2007, 25 (2),403–419.

(25) Comer, P. J.; Spouge, J. R.; Stearn, S. Risk assessment of BSEinfectivity in the environment from rendering of over thirtymonth scheme cattle. J. Risk Res. 1998, 1 (4), 281–293.

(26) International Organization for Standards. Soil qualitysGuidancefor burial of animal carcasses to prevent epidemics. Notificationof guideline under development at http://www.iso.org/iso/iso_catalogue (Accessed 7 January 2010).

ES100039N


Intervention Strategies for Carcass Disposal: Pareto Analysis of Exposures for Exotic Disease...

Documents

Transcript of Intervention Strategies for Carcass Disposal: Pareto Analysis of Exposures for Exotic Disease...