DEWORMING - Copenhagen Consensus Center€¦ · DEWORMING Andrew Hall & Sue Horton BEST PRACTICE PAPER: NEW ADVICE FROM CC08. COPENHAGEN CONSENSUS CENTER COPENHAGEN BUSINESS SCHOOL

DEWORMING

Andrew Hall & Sue Horton

BEST PRACTICE PAPER: NEW ADVICE FROM CC08

C O P E N H A G E N C O N S E N S U S C E N T E R C O P E N H A G E N B U S I N E S S S C H O O L • S O L B J E R G P L A D S 3 • 2 0 0 0 F R E D E R I K S B E R G • D E N M A R K

+ 4 5 3 8 1 5 2 2 5 5 • I N F O . C C C @ C B S . D K • W W W . C O P E N H A G E N C O N S E N S U S . C O M

copenhagen consensus center best practice paper: new advice from cc08

BestPracticePaper

Deworming

AndrewHall*andSueHorton†

TheauthorswouldliketothankNilanthideSilva,DirkEngelsandHaroldAldermanforhelpfulcommentsona

previousdraft.

Coverimagecredit:AndrewHall,WestminsterUniversity

*CentreforPublicHealthNutrition,UniversityofWestminster,115NewCavendishStreet,LondonW1W6UW,U.K.ah2108@googlemail.com†VisitingScientist,MicronutrientInitiative,andProfessorofEconomics,WilfridLaurierUniversity,WaterlooOnN2L3C5,Canada.shorton@wlu.ca

3

PREFACE

FortwoyearsbeforeCopenhagenConsensus2008,ateamofexpertswrotepapersidentifyingthebestwaystosolvetheworld’sbiggestproblems.Thosepapersshowedthatwehavetheknowledgetodotremendousamountsofgoodineachoftheseareas.Thatresearchwasutilizedbyapaneloftopeconomists,includingNobellaureates,whowerecommissionedbytheCopenhagenConsensusCentertoidentifythemosteffectiveinvestments.TheprioritizedlistproducedbyCopenhagenConsensus2008providesgovernments,donorsandphilanthropistswithaguidetotheareaswhererelativelysmallamountsofmoneycanproveextremelypowerful.Theresearchthatprovidedthebuildingblockstothisprocess–andafulldescriptionoftheoutcome–formthebook,‘GlobalIssues,GlobalSolutions,VolumeTwo’,publishedbyCambridgeUniversityPressin2009.Thisisanexcellentoverviewofglobalproblemsandthemostpromisingsolutions.GiventhelevelofinterestinCopenhagenConsensus2008,theCopenhagenConsensusCenterdecidedtocommissionaspecificsetofpapersthatdealwiththespendingoptionsgivenhighestprioritybytheexpertpanel.ThegoaloftheseBestPracticePapersistoprovideclearandfocusedempiricalrecommendationsonthecostsandbenefitsofimplementingthesolutions,andadviceonhowtodoso.TheproblemsdealtwithbyCopenhagenConsensus2008arevast.Thepracticalapproachesidentifiedhereproveincrediblypowerfulreading.TheCopenhagenConsensusCenterhopesthattheyshallproveaninvaluableresource,andfurtheradvancethegoalofpromotingthemostsoundinvestmentstohelphumanity.BjornLomborgCopenhagen,2009

4

EXECUTIVESUMMARY

Theprevalenceofinfectionwithatleastonespeciesofintestinalnematodeworm,whichincluderoundworm,hookwormandwhipworm,isestimatedas48%indevelopingcountries.Sub‐SaharanAfrica,SouthandSoutheastAsia(includingthePacificislands),andpartsofLatinAmericaaremostheavilyaffected.Transmissiondependsonsanitation,agriculturalpractices,whetherornotshoesareworn,andenvironmentalconditions.Althoughinfectioniswidespread,thenumberofindividualswithamoderatewormload(definedhereas10ormoreworms)isnotsubstantialuntiltheprevalenceexceeds50%.Masstreatmentissafeandinexpensive.Albendazoleandmebendazolecanbegivenassingle‐dosetreatmentstochildrenagedtwoyearsandolder,andtherearepreparationsavailableforchildrenaged1–2years.Thecostofdeliveringoneroundoftreatmentisapproximately$0.15perchildforchildrenwhenadministeredinschool,and$0.25perchildforpreschoolchildrenwhencombinedwithanotherinterventioninprogramssuchasChildHealthDaysorinprimaryhealthcarefacilities.Treatmentisrecommendedforchildrenaged2through14yearsiftheprevalenceinthecountryorregionexceedsaspecifiedthreshold.In2006theWorldHealthOrganization(WHO)reviseditsguidelinesregardingsuchthresholdsandnowrecommendstreatmentonceayearwhereinfectionrateswithanyintestinalwormexceed20%andtwiceayearwheretheyexceed50%oruptothreetimesayearifresourcespermit.Estimatesoftheprevalenceofinfectionswithintestinalnematodewormsindicatesthatthiswouldgiveanannualglobaltreatmentcostof$276millionforthedevelopingworldbutexcludingtheformermembersoftheSovietUnion,forwhichprevalencedataarenotavailable.Ananalysisoftherelationshipbetweentheprevalenceofinfectionandtheestimatedproportionofindividualswith10ormoreworms,whichmightbeconsideredasamoderatetoheavyinfection,indicatethatathresholdprevalenceof20%formasstreatmentisnotcost‐effectivewhenestimatedintermsofcostperdiseasedpersontreated.Weproposealternativethresholds:treatmentonceperyearwheretheprevalenceis40%ormore,twiceperyearwhereitis60%ormore,andthreetimesperyearwhereitis80%ormore.ThiswouldhavethebenefitoflowercostthancurrentWHOrecommendations($224millionannually)anddevotesahigherproportionofexpenditurestotreatingindividualswhoareinfected(74%asopposedto61%),andahigherproportiontotreatingindividualswith10ormoreworms(31%asopposedto21%).Treatmentisverycost‐effective.Inschoolchildren,thereductioninanemiainparticularduetotreatinghookwormisassociatedwithincreasedschoolparticipationandhenceschoolachievement,andultimatelywithincreasedproductivityofworking‐ageadults.Incountrieswithahighprevalenceofinfection,thebenefit:costratiomaybeashighas60:1(althoughthiswouldbereducedsomewhatifcostsofhiringadditionalteachersweretakenintoaccount).Forpreschoolchildren,therearefewerstudies.Estimatessuggesthoweverthatthebenefitsofreducedanemiacouldgivebenefit:costratiosof6:1;thisdoesnotincludeadditionalgainsduetobetterweightandheightgains.

5

Thepaperalsomakessomerecommendationsforthefuture.First,itwouldbegoodpracticetoalternatebetweendifferentclassesofdewormingdrugstohelpavoiddevelopingdrugresistance:mebendazole/albendazolecouldbealternatedwithpyrantel/oxantelforexample,althoughthiswouldrequirechangestotheWHOEssentialDrugslist.Second,itisgoodpracticetodosamplesurveysofprevalenceinschoolchildrenatleastonceeverytwoyears.Simplenon‐quantitativetestscostabout0.1%ofthecostoftreatment(theestimateislessthan$0.25millionworldwide),andcansavemoneybyeliminatingunnecessarytreatments.Third,giventhattreatmentthresholdsarebasedonthecombinedprevalenceofinfectionwithanyspecies,moreworkisneededonestimatingtheprevalenceofmixedinfectionsascurrentglobaldataarepublishedbyspecies.IntheCopenhagenConsensus2008,apanelofrenownedeconomistsrankedfivenutritionalinterventionsamongthetoptenof30proposalstoanswerthequestion,whatwouldbethebestwaysofadvancingglobalwelfare?Dewormingandothernutritionprogrammesinschoolwererankedasthesixthbestinterventionoverall,consideringtheirpotentialbenefit:costratio,anticipatedfeasibilityandthesustainabilityoftheintervention.InSection1wediscussbrieflythechallengesinvolvedincontrollingdiseasecausedbyintestinalnematodeworms,particularlyroundworm(Ascarislumbricoides),whipworm(Trichuristrichiura)andthetwomainspeciesofhookworm(AncylostomaduodenaleandNecatoramericanus).InSection2wediscusshowtotreatdiseaseandstarttoreducetransmissionbydewormingpeopleusinginexpensive,single‐doseanthelminticdrugswiththefocusparticularlyonschool‐agechildren,butwithsomediscussionoftreatingpreschoolchildrenandpregnantwomen.Section3providessomecost‐effectivenessandeconomicanalysis.Section4concludes.

6

1 THECHALLENGE

Ithasbeenestimatedthat1.22billionpeopleinlow,lower‐middleandupper‐middleincomeeconomies,or26%oftheirpopulation,areinfectedwithroundworms,0.80billion(17%)withwhipwormsand0.74billion(15%)withhookworms(deSilvaetal.,2003a).Fromtheseprevalencesitcanbeestimatedthatsome2.3billionpeople,orabout48%ofpeoplelivingintheworld’spoorestcountriesarelikelytobeinfectedwithatleastoneofthesetypesofwormsandthatnearly10%areinfectedwithtwoormoretypes,ifweassumetoafirstapproximationthatinfectionsareindependent(seeHalletal.,2008forfurtherdiscussion).Figure1,fromdeSilvaetal.(2003a),showsthedistributionofthesethreecommontypesofintestinalworms.Thecountriesofsub‐SaharanAfrica,SouthandSoutheastAsiaandpartsofLatinAmericaareworst‐affected,whichreflectsenvironmentalconditionsthatsuitthesurvivalofinfectiousstages,poorsanitation,poorpersonalhygieneandperhapsparticularbehavioursthatincreasetheriskofinfection,suchasusingfreshhumanfaecesasafertilizerornotwearingshoestoprotectfrominfectionwithhookworm.Theremayalsobeconsiderablevariationintheprevalenceofinfectionbyregion,withincountries,andevenwithindifferentregionsofthesamecountry,largelydependingonlocalenvironmentalconditions.Forexamplethedistributionofinfectionsvarieswithaltitude,probablyasaresultoftheeffectsoftemperature,humidityandultra‐violetlightonthesurvivalofwormeggsandlarvae,whileincrowdedurbanslumstheremaybeintensetransmissionofroundwormsandwhipworms.Infectionwithintestinalwormsisnotnecessarilythesameasdisease.Thisisbecausetheriskofbeingdiseaseddependsprincipallyonthenumberofwormsinthegut,commonlycalledthewormload,butalsoonthespeciesofworm,themixtureofspecies,thedurationofinfection,andthesize,ageandhealthoftheinfectedperson(Halletal.,2008).Wormloadstendtoaccumulateslowlybecauseeachwormwithinahumanhostistheresultofexposuretoasinglewormeggorlarva;becausewormscanlivefor1–3years,dependingonthespecies;andbecausethereseemstobenofullyprotectiveimmunity,soinfectionisacontinuousprocessanddiseasecandevelopslowly.Infectionsareusuallydiagnosedusingamicroscopetoseethecharacteristicwormeggsinafaecalsample.Theproportionofpeoplewitheggsintheirfaecesprovidesanestimateoftheprevalenceofinfection,butthisisapoorindicatoroftheriskofdiseaseandoftheimpactoftreatmentbecauseonlytwoworms–amaleandafemale–arenecessarytoproduceeggs,andtwowormsareveryunlikelytocausedisease.Theconcentrationofeggsinfaecescanbeusedasaproxyforthewormburden,butthereisevidencethateggcountsperfemalewormvaryfromcountrytocountry,forroundwormsatleast(Hall&Holland,2000).

7

Figure1ThedistributionofAscarislumbricoides,Trichuristrichiuraandthehookworms.

Source:deSilvaetal.(2003a)

Theriskthatafewwormscancausediseaseisverylow;diseaseismostlyassociatedwithmoderatetoheavyinfections.Althoughthereisnoacceptednumberofwormsthatdefinesmoderatetoheavyinfectionsforeachspecies,itistypicaltofind>80%ofwormsin<20%ofhosts(Anderson&May,1991).Theconsequencesare:thatnotallinfectedpeoplewillbenefitfromtreatmenttothesamedegree;thatanymeasuredeffectsoftreatmentonthe

8

moderately‐toheavily‐infectedminoritymaybedilutedintheaverage;andthatitmaynotalwaysbecost‐effectivetotreatthewholepopulationwhenonlyafewindividualsarediseased.Thewormburdenisabetterindicatoroftheriskofdiseasethantheprevalence,butestimatingtheburdeneitherrequiresexpellingandcountingworms,whichisdifficulttodo,orrequiresanestimateoftheeggcount,whichishighlyvariable.Unfortunatelytheprevalencecannotreadilybeusedtoestimatethemeanwormburdenbecausetherelationshipbetweenthetwoishighlynon‐linear.Figure2showsthatwhentheprevalenceofroundwormsisbelowabout50%,themeanwormburdenis2orless,butaboveaprevalenceof50%themeanwormburdenrisesexponentially.Thismeansthattheproportionofpeoplewithmoderatetoheavyinfectionswhoareatriskofdiseaserisessimilarly(Guyatt&Bundy,1991).

Figure2Therelationshipbetweentheprevalenceofinfectionwithanintestinalnematode

worm,inthiscaseAscarislumbricoides,andthemeanwormburdenestimatedbyapplyingthenegativebinomialdistributionusingaclumpingparameter(k)thatvarieslinearlywiththemeanwormburden(seeGuyattetal.,1990).

Themechanismbywhichintestinalnematodewormscausediseasediffersbetweenspecies:roundwormsfeedongutcontents,butthisprobablyhasonlyaminorimpactonhumannutritionastheneedsofwormsarerelativelysmall;roundwormsandhookwormscanbothcausemaldigestionormalabsorptionofnutrients;hookwormsandwhipwormsfeedonbloodandtissuefluids;allspeciescanelicitinflammatoryresponsesthataffectappetiteandmetabolicrate,andtheyallcancausethediversionofperhapsscarcenutrientstomountresponsestoinfection(reviewedbyHalletal.,2008).Bythesevariousmechanismswormsmayhaveeffectsonhaemoglobinconcentrationandthusonanaemia,onphysicalfitnessandworkproductivity,andonappetiteandgrowth,bothintermsofbodyweightandheight(Aldermanetal.,2006;Bhuttaetal.,2008;Brookeretal.,2008;deSilvaetal,2003b;Gulanietal.,2007;

9

Guyattetal.,2001;Halletal.,2008;Stoltzfusetal.,2004).Theremayalsobeindirectbenefitsthrougheffectsonparticipationinschoolandintellectualdevelopment(MiguelandKremer,2004;Bobonisetal.2006).Itistemptingtoconcludethatsimplytreatingwormswillrectifythediseaseordeficitsthattheyhavecaused,butthislogicisflawed.Forexample,ifwormscausemalnutritionandimpairchildren’sintellectualdevelopment,whywouldsimplyremovingthewormsreverseanydeficits?Toreplacelosthaemoglobinorachievecatch‐upgrowthforexample,childrenwillneedenergy,proteinandmicronutrients(Hall,2007;Halletal.2008).Andifchildrenhavemissedschoolbecauseofillnessduetoworms,theymayneedremedialeducationaswell.Aschildrenwithwormstendtoliveinpoorenvironmentsinwhichundernutritionandintellectualdeprivationalsooccur,removingwormsisonlyafirststepinimprovingchildgrowth,healthanddevelopment,andadditionalremedialtherapymaybeneeded.Thissuggeststhatnewdewormingprogrammesshouldnotexpectimmediateimprovementsinnutritionalstatusandgrowthunlesschildrenareknowntoconsumeadietthatisadequatebothinquantityandquality.Incountrieswherechildrenalreadyconsumeapoordiet,dewormedchildrenmayrequiretherapeuticmicronutrients,supplementaryfoodorevenremedialeducationiftheyaretorecoverquicklythelossescausedbyworminfections.Withoutthemthebenefitofdewormingmaynotbereadilyapparent.Theeconomicanalysisofbenefitsofdewormingisbestappliedtochildrenwhocanbekeptfreeofmoderateorheavywormburdensthroughouttheirchildhoodbydewormingoftenenoughtopreventmoderatetoheavyinfectionsfromaccumulating,ideallysupportedbysanitationandhealtheducation.Asreinfectionwithintestinalwormscanoccurimmediatelyaftertreatment,theaimofdewormingisfirst,toreduceinitialwormloadsby>80%andtherebyeliminatedisease,andthentorepeattreatmentoftenenoughtopreventmoderatetoheavyloadsfrombeingreaccumulated.Periodicdewormingalsohelpstoreducetransmissionbyremovingworms,becauseafemaleroundwormmayproduceupto200,000eggsadayandhasalife‐spanofupto18months.Dewormingschoolchildrenalonehasbeenshowntoreducetransmissiontountreatedmembersofthecommunity(Bundyetal.,1992b),sotreatmenthasexternalities.Howeverthelongtermsolutionistopromotechangesinbehaviourthroughpublichealtheducationsothatpeopleuselatrines(Nocketal.,2006)andtoinstalleffectivesanitationtokeeppeopleandhumanfaecesapart.School‐agechildrenaremostatriskofinfectionanddiseasecausedbyintestinalroundwormsandwhipwormswhilehookwormstendtobemorecommoninadolescentsandadults(Bundyetal.,1992a).In2001the54thWorldHealthAssemblyadoptedaresolutiontodewormatleast75%ofallschool‐agechildrenatleastonceayear(WHO,2001).Theglobalprogressmadetowardsmeetingthistargetisreviewedregularly.Themostrecentdataarefor2006,andofthe130endemiccountries,datawereavailablefortreatmentofschoolagechildrenin64,andforpreschoolchildrenin51countries(WHO,2008).Coverageofpreschoolchildrenwas55%andofschoolagechildren22%,butonlyinthosecountriesreportingdata.

2 SOLUTIONS

10

Thescientificbasisfortheshorttomediumtermsolutiontotheproblemofintestinalworms–massdeworming–liesinamixtureofepidemiology,pharmacologyandeconomics.First,whentheprevalenceofinfectionisgreaterthan50%thepurposeofdiagnosiswouldbetoidentifyuninfectedindividualstoexcludefromtreatmentratherthantoidentifythosetotreat.AsFigure2shows,themeanwormburdenincreasesexponentiallyatprevalencesabove50%andsothereforedoestheriskofdisease(Guyatt&Bundy,1991).Becausethecostoftreatmentismuchlessthanthecostofdiagnosis;becausethedrugstotreatwormsareverysafe,meaningthatthereisnoknownharmforanuninfectedpersontobetreated;andbecausethedrugsareeffectiveagainstallspeciesofintestinalnematode,givingmasstreatment(ratherthandiagnosingandtreatingpeopleindividually)wasidentifiedasthesimplestandmosteffectivemeansoftreatingmoderatetoheavyinfectionswithintestinalworms(WHO,1996).Thestrategyofmasstreatmentwasmadepossiblebytheavailabilityofsingledosedrugsthattreatallmainspeciesofintestinalnematodeworm(WHO,1995),althoughwithvaryingdegreesofefficacy:

• Piperazineisanolddrugusedoccasionallyatasingledoseof75mg/kgbodyweight,butisonlyeffectiveagainstroundworms.

• Levamisolegivenasasingledoseof2.5mg/kgiseffectiveagainstroundwormsandtosomedegreeagainsthookworms.

• Pyrantelpamoategivenasadoseof10mg/kgiseffectiveasasingledoseagainstroundwormsandhookworms,butnotagainstwhipwormsunlessgivenwithoxantel(Gustafssonetal.,1987).

• Amixtureofpyrantelandoxantelgivenatadoseof10mg/kgisequallyeffectiveasmebendazoleagainstroundwormsandhookworms,butismoreeffectiveagainstwhipworms(Albonicoetal.,2002).OxantelisnotcurrentlyontheWHOessentialdrugslist,however.

• Anewdrugcallednitazoxanideiseffectiveagainstroundwormsandwhipwormsbuthastobegiventwiceadayforthreedays.

Themaindisadvantageofallthesetreatmentsisthatbodyweightneedstobeknowntodeterminethedose,whichrequiresaweighingscale.Howeveritispossiblethatheightcouldbeusedtoestimatetheweightofchildrenandfromthat,thedoseofdrug,sothatthenumberoftabletscanbemarkedona“tabletpole”.Thishasbeendoneforivermectintotreatonchocerciasisandforpraziquanteltotreatschistosomiasis(Halletal.,1999).Themostwidelyusedtreatmentsforintestinalnematodewormsaretwobenzimidazolederivatives:

• Albendazole,givenassingledoseof400mgtoanyoneolderthan1year,ishighlyeffectiveagainstroundworms(88%curerate),effectiveagainsthookworm(72%curerate),butlesseffectiveasasingledoseagainstwhipworm(28%curerate)(KeiserandUtzinger,2008).

• Mebendazole,givenasasingledoseof500mgtoanyoneolderthan1year,ishighlyeffectiveagainstroundworms(95%curerate)butlesseffectiveagainsthookworm(15%curerate)andwhipworm(36%curerate)(KeiserandUtzinger,2008).

11

Anadvantageofalbendazoleandmebendazoleisthattheyareadministeredasasingle,standarddosetoanyoneolderthan2years,whichissimpleandensurescompliance.Bothareavailableasflavoured,chewabletablets,whichmakesthemeasytotake.Thereisalsosomeevidenceofgeographicaldifferencesintheefficacyoftreatmentswithbenzimidazoles:theyseemtobemoreeffectiveinAfricathanAsia(Bennett&Guyatt,2000).Studieshaveshownthatalbendazolecanbegivenatthesametimeaspraziquantel(Oldsetal.,1999),whichcommonlyoccurswithintestinalwormsinthesamechildren,mainlyinsub‐SaharanAfrica.Co‐administrationhelpstoreducethecostsofdeliveryandtreatment(PCD,1999).Ithasbeenassumedthat,becausemebendazoleisverysimilartoalbendaz‐ole,praziquantelcanalsobeco‐adminsteredwithmebendazole.Theadvantageofthesesingledosetreatmentsisthattheyensurecompliancewithtakingthetreatmentanditmeansthatoneperson,suchasaschoolteacher,couldtreathundredsofchildreninaday.Treatingchildreninschoolisparticularlyattractive,forseveralreasons.First,becausetheprevalenceofinfectionandwormloadstendtobeheaviestinchildrenaged5to15years;becausesuchchildrencontributemosttotransmissionwhentheydonotusealatrine;andbecauseschoolsprovideanexistinginfrastructurethroughwhichtodeliverandadministertreatments.Inaddition,parentsaregenerallyverysupportiveoftreatingworminfectionsandappreciatethepotentialbenefits(Brookeretal.,2001).Studieshaveshownthatthefinancialcostsofdewormingschoolchildrenaremodest(discussedinthenextsection).Genericversionsofbothmebendazoleandalbendazolearewidelyavailablebutthereisconcernforfakemedicinesinsomepartsoftheworld,sotreatmentsshouldbetestedforactiveingredients,oratrialofdrugefficacyshouldbedoneifthesourceisnotassured.Althoughtherearecoststodistributetreatmentstoschools,totrainteachersandforgeneraladministration,itseemsgenerallyassumedthatteachersdonotrequireanypaymentinadditiontotheirsalary,sothatthecostofactuallyadministeringtreatmentispaidfor.Whetherteachersarewillingtotakeonwhatcouldbeviewedasamedicalresponsibility,isanotherissue,andneedstobedealtwithbycarefulandsensitivetraining,whichwillhavecosts.Incountriesinwhichenrolmentratesarelow,thereisconcernthatnon‐enrolledchildrenwillmisstreatmentsdeliveredthroughschools.Thefewstudiesthathavebeendonecomparingintestinalworminfectionsinenrolledandnon‐enrolledchildrenhavenotshownsignificantdifferences(Beasleyetal.2000;Fentimanetal.,2001).Non‐enrolledchildrenmayactuallycomefromhouseholdscontainingenrolledchildren(Montresoretal.,2001)sotheremaybenojustificationforassumingthattherewouldbedifferences.MassdewormingisnowbeingaddedtoprogrammessupportedbyUNICEFthatdelivercapsulesofvitaminAtochildrenaged1–5years,suchasChildHealthDays(UNICEF/WHO,2004).AlargerandomizedcontrolledclustertrialinUgandaofalbendazolegiveneverysixmonthstochildren1‐7yearsatChildHealthdaysfoundsignificantdifferencesinweightgainbetweentreatedchildrenanduntreatedcontrols,andthechildrenwhogotthemostdosesoverthetwoyearstudyshowedthegreatestweightgain(Aldermanetal.,2006)

12

Theuseofalbendazoleandmebendazoletotreatchildrenlessthan24monthsoldwasrecommendedbyaWHOinformalconsultation(WHO,2003).Therationaleisthesameasforolderchildren:totreatorpreventdisease,mainlybecausethereisconcernthattheeffectsofwormsmaybeproportionatelygreaterinsmallchildren.Childrenagedlessthan24monthswhocannotchewshouldbegivenacrushedtabletorsyrup,whichistypicallymoreexpensivethanatablet,bothtopurchaseandtodistribute.Thisisbecausethereisariskthatachildwhoisunwillingtochewatabletmaychoketodeathifforcedtotakeone.ThishappenedinEthiopiain2007tofourchildrenagedlessthan36monthsold(WHO,2007).Therearenocostdataforthedistributionofdewormingmedicationasasyruptochildrenlessthan2yearsold.Ideallyallwomenofreproductiveageshouldbedewormedperiodicallytopreventthebuild‐upofmoderatetoheavyinfections,especiallyofhookworm,whichcancontributetobloodlossandanaemia(Pawlowskietal.,1991).Ithasrecentlybeenestimatedthat38millionwomenofreproductiveageinsub‐SaharanAfricaareinfectedwithhookworms(Brookeretal.,2008).TheWHOrecommendsthatalbendazoleandmebendazoleshouldnotbetakenbyanywomanduringthefirsttrimesterofpregnancy(WHO,2005).AstudyinSierraLeoneoftreatingwomenwithalbendazoleafterthefisttrimesterfoundthatithelpedtopreventadeclineinhaemoglobinconcentrationduringthesecondandthirdtrimesters(TorlesseandHodges,2000).Moreworkonthecost‐effectivenessofprogrammestotreatpreschoolchildrenandpregnantwomenisrequired,whichwilldependonthespeciesofworms,theprevalenceofinfection,thefrequencyoftreatmentperperson,andmostimportant,thecostsofdistributingandgivingtreatments.Asmanywomeninmanydevelopingcountriesdonotseekhealthcareuntiltheirthirdtrimesterofpregnancy,achievinghighcoverageofwomenfromthestartoftheirsecondtrimesterwillbeanissue.Inareaswheremalariaandhookwormco‐occur,treatmentsforwormscouldbegiventopregnantwomenintheirsecondorthirdtrimesterwith,orsoonafter,presumptiveintermittenttreatmentformalaria.

School‐basedtreatmentprogrammes

Thebasisfordecidingwhenandhowoftentogivemasstreatmentistheprevalenceofinfectionwithanyspeciesofintestinalnematodeworm(roundworm,whipwormandhookwormcombined)inasurveyofschoolchildreninasampleofschools.Anychildwithanyspeciesofintestinalnematodewormiscountedasinfected;tapewormsandschistosomesareexcludedasthedrugsusedtotreatintestinalnematodesdonotaffecttheseworms.Itisnotusuallypracticabletoassesstheprevalenceofinfectionineveryschool,soitisestimatedinarepresentativesampleandadecisionismadewhethertogivemasstreatmenttoallchildreninallschools.Thenumberofschoolstobeselectedandthemethodbywhichtheyaresampledwilldependonthescaleoftheintendedprogrammeofmasstreatment.Iftheprogrammeistobeimplementedinanenvironmentallyhomogenousarea,suchasanumberofsimilardistrictsoraregion,theneither10%ofprimaryschools(minimum=3)uptoamaximumof10shouldbeselectedforstudy.Therearetwomainwaystoselectthoseschools:eitherwithaprobability

13

thatisproportionatetothenumbersofchildrenenrolled(PPSmethod),whichwouldtendtoselectmorelarge,urbanorperi‐urbanschoolsthansmallruralschools;orbyalotqualityassurancesampling(LQAS)method,whichsamplesschoolsevenlyacrossageographicareaandisbestforhealthproblemsthatarenotevenlydistributed.ThePPSmethodiscommonlyusedbyUNICEFandotheragenciestoundertake30/30clustersurveys,sowillnotbedescribedhere(Bennettetal.,1991).ToselectschoolsusingtheLQASmethodtheareashouldbedividedonamapintoquadrantsofalmostequalsize,thenumberdependingonhowmanyschoolsarerequired,andtheschoolclosesttothecentreofeachquadrantshouldbeselectedforstudy.ThismethodhasbeenusedtomapthedistributionofinfectionswithSchistosomaspp(Brookeretal.,2007).Iftheareainwhichaprogrammeistobeimplementedisnotenvironmentallyhomogenous,thenitshouldbedividedintoecologicalzonesandthesamesamplingprocedureshouldbeappliedtoeachzone.Thefindingscanthenbeappliedtotheadministrativeunitswithineachzone.IfdataarewantedbyDistrictorRegion,thenthe10‐schoolsurveycanbedone,butthisisunnecessaryandwastefulofresourcesifthedistrictsareenvironmentallysimilar.Ineachschoolasampleoffreshfaecesshouldbecollectedfrom40children,20eachfromtwoclassescontainingchildrenwhoaremostlyaged10–14years.Thisagegroupisoldenoughtounderstandinstructionsaboutcollectingtheirfaecesandmaybemoreamenablethanclassesofolderadolescents.Thestoolsamplesshouldbeexaminedunderamicroscopefortheeggsofintestinalworms,whicharecharacteristicexceptthattheeggsofthetwospeciesofhookwormscannotbetoldapart(WHO,1994).Thereareanumberofmethodstoexaminefaeces,inorderofdecreasingdiagnosticsensitivity:theformalin‐etherconcentrationmethod;theKato‐Katzsmearmethod;asimpleKatosmear;orasimpledirectsmearoffaecesinsaline(WHO,1991).Saltflotationmethodsarenotrecommendedastheytendtodamagehookwormeggs.TheKatomethodsdonotallowthediagnosisofinfectionswithintestinalprotozoaatthesametime;theothermethodsdo.Thefirsttwomethodsaresemi‐quantitativeandallowanestimateoftheconcentrationofwormeggsinfaecestobemade;theothermethodsarequalitative.Thesurveytoestablishtheprevalenceofwormsrequiresonlyaqualitativeexaminationoffaeces.Toallowmonitoringfordrugresistanceaquantitativeexaminationoffaecesisneeded,butthisneedonlybedoneinacohortofchildrenin10–15schools(400–600children)inthewholecountryorregioninwhichaprogrammeisimplemented.Themeanconcentrationofeggsinfaeces,alsocalledaneggcountandexpressedaseggspergram(epg)offaeces,isaroughindicatoroftheintensityofinfectionwitheachspecies,althoughtherearemanyreasonswhyitcanbehighlyvariable(Hall,1982;Engelsetal.,1997;Hall&Holland,2000).Thepercentageofchildrenwhosefaecescontaintheeggsofroundworms,whipwormsorhookwormsshouldeachbedetermined,aswellasthepercentageofchildrenwhohaveanyoftheseinfections.Thespeciesofwormsmayinfluencethedrugused,asalbendazoleseemsbetterthanmebendazoleattreatinghookworm(Albonicoetal.,1994;Bennett&Guyatt,2000;

14

Keiser&Utzinger,2008)whileahighprevalenceandintensityofwhipwormsmayneedthreedailydosestoachievesubstantialcurerates(Hall&Nahar,1994;Bennett&Guyatt,2000)althoughthisisnotoftendone.Thecombinedprevalenceisusedtodeterminewhethertogivemasstreatmentandhowoftentotreat.Thecostofsuchsurveysdependslargelyontransportandlabour.Avehicleisneededtovisitschoolstocollectfaecalsamplesandreturnthemtoalaboratorythathasamicroscope,unlessaportablefieldmicroscopeisavailable.Atrainedtechniciancanexamine30–50samplesaday,dependingonthemethodused,whichmeansthatthesamplesfromoneschoolcantypicallybeprocessedinoneday.Theonlymaterialsneededforadirectsmearmethodarenon‐sterile,reusablebottlestocollectandtransportfaeces,andsomereusableglassmicroscopeslidesandcoverslips.Weestimatethatthecostperschoolin2008isapproximately$87(40samples),i.e.$900‐$1300percountry(Table2).Ifnationalsurveysweredoneannually,orsurveysweredoneineachIndianstateandChineseprovinceatthesamefrequencyasforacountry,annualtestingcostswouldnotexceed$0.25m,whichis0.1%ofourestimatedannualtreatmentcosts(Halletal2008).In1996aWHOinformalconsultationendorsedaWHOrecommendationthatacombinedprevalenceofintestinalnematodewormsof≥50%wassufficienttowarrantmasstreatmentofschool‐agechildren(WHO,1996).Theaimwastotreatdiseasebyeliminatingallmoderatetoheavywormburdensandtoreducetransmission.TheWHOalsoproposedacomplicatedstrategytodecidehowoftentogivemasstreatmentbasedonthreecategoriesofprevalenceandtheproportionofmoderatetoheavyinfectionsbasedoneggcountsforeachseparatespecies(WHO,2002).Thisrequiredanestimateoftheconcentrationofeggsinfaeceswhichisnoteasytodoinasmallruralhospitallaboratoryunlessthenecessarymaterialsaresupplied.Thebiologicalbasisandthepracticalityofthisrecommendationwasweak.

In2006anotherinformalconsultationendorsedareductioninthethresholdformasstreatmentonceayearto20%inastrategycalled“preventivechemotherapy”andsimplifiedthebasisforrepeatingtreatmentsasfollows:if≥50%ofchildrenareinfectedthenmasstreatmentshouldbegiventoallschool‐agechildrentwiceayear;if≥20%and<50%ofchildrenareinfectedthenmasstreatmentshouldbegiventoallschool‐agechildrenonceayear(WHO,2006).Provisionwasalsomadefortreatmentthreetimesayeariftheprevalenceofinfectionwas≥50%andresourceswereavailable,butnodifferentorhigherthresholdwasspecified.Theeffectofreducingthethresholdformasstreatmentfrom50%to20%meansthattreatmentsmaybegiventoupto80%ofchildrenwhodonotneedthem,whichincreasesthecostperinfectedpersontreatedbyuptofourfold,andincreasesthecostperdiseasedpersontreatedbyanexponentiallygreatermultiple.Thisisexaminedinthenextsection.

15

Table2.Estimatedcostsofprocessingstoolsamplesforoneschool(40children)

Cost Item Cost/school/day and basis

for cost calculation

Cost/school/day in 20083

Microscopy lab tests $28 (2001)1 $34

Driver $18 (2000)2 $22.50

Operating, car $14 (2000)2 (assuming 200

Km/day)

$16.80

Depreciation, car $10( 2000)2 (assuming life 8

years, 240 days/year)

$12.50

TOTAL $85.80

1Source:Mulliganetal,2003(revised2005).Costofstoolmicroscopytest$0.67to$0.70inSouthAsiaandsub‐SaharanAfrica,assume40/daytested

2Source:WHO‐CHOICE(2008),costsforSouthAsiaandsub‐SaharanAfrica(basedonthethreeregionsSEAROD,AFRDandAFRE)

3Source:updatedto2008usingUSConsumerPriceIndex Therearesomespecificgroupswhoshouldnotbetreated.First,asdrugsshouldnotbegiventowomenintheirfirsttrimesterofpregnancy(WHO,2005)itisrecommendedthatnogirlaged13yearsoroldershouldbetreated,incasesheispregnant.Itcannotbeassumedthatschool‐agegirlsarenotsexuallyactive.Astudyof9,000schoolchildreningrades4to6inprimaryschoolinTanzaniafoundthat20%ofgirlsreportedhavinghadsex,butonly39%of114girlswithbiologicalmarkersofsexualactivitysuchasaninfection,acknowledgedhavinghadsex,indicatingthatsuchactivitywasgreatlyunder‐reported(Toddetal.,2004).Pregnancyisalsoacommonreasonforyounggirlsdroppingoutofschool,particularlyinsub‐SaharanAfrica.Second,childrenwhoaresickforanyreasonwhentreatmentsarescheduledtobegivenshouldnotbetreated.Thisisnotbecausetreatmentisdangerous,butbecauseifasickchildsubsequentlydiesthetreatmentcouldbemistakenlyassociatedwiththedeathwhichcouldendangertheprogramme.Whenverylargenumbersofchildrenarebeingtreatedonthesamedaytheprobabilitythatachildwilldieisnolongerverysmall.Itisgoodpracticetomonitortheefficacyoftreatmentswithanthelminticdrugssothatthedevelopmentofdrugresistancecanbedetected(WHO,1999).Thisisdonebyexaminingfaecalsamplescollectedfromasampleofthesameindividualsbeforeandaftertreatmentandthencalculatingtheproportionofpreviouslyinfectedpeoplewhosefaecesnolongercontainwormeggs.Thisiscalledthecurerateandcanbecalculatedforeachspecies.Amoresensitivemeasureofefficacyistheeggreductionrate(Bundyetal.,1992a).Thisistheaveragedifferenceintheconcentrationofeggsofeachspeciesbeforeandaftertreatmentexpressedasapercentageoftheinitialaverageeggcount.However,ifthefecundityofwormsisinfluencedbythenumberofwormspresentinthegut(HallandHolland,2000)thenthefecundityofwormsleftinthegutafterincompletetreatmentmayincreaseproportionallyaftertreatmentsothattheefficacyoftreatmentmeasuredasaneggreductionratemaybeunderestimated(KotzeandKopp,2008).

16

Theriskofdevelopingdrugresistancecanbeminimisedbyensuringthataveryhighcurerateisachieved,butthisisdifficulttodoforwhipwormswithoutgivingmultipledosesofalbendazoleormebendazole(deSilvaetal,1997;Bennett&Guyatt,2000).Repeatedtreatmentofanyresidualwormloadwiththesamedrugcouldleadtotheselectionofwormsthatareresistanttothetreatment.Whenrepeatedtreatmentsaregivenasapartofdiseasecontrolprogrammesitisbestiftwodrugswithdifferentmechanismsofactionarealternatedorcoadministered,astheWHOnowrecommendstotreatmalaria,forexample.Thisisbasedonthelowprobabilitythatresistancewilldevelopinthesamespeciestotwodrugsatthesametimeandthat,ifthereisresistancetoeitherdrug,atleastonewillbeaneffectivetreatment.Currentprogrammestotreatintestinalwormsrelyheavilyonalbendazoleandmebendazole,whicharecloselyrelatedtreatmentswiththesamemodeofaction,sothereisariskthatresistancewilldevelopifthesedrugsareusedwidelyandrepeatedly.Resistancetothesedrugsiscommonamongtheintestinalwormsoffarmanimalsasthedrugshavebeenwidelyandheavilyusedandbecausesheepandcowsoftencontinuetofeedonpasturesthatmaybeheavilycontaminatedwiththeeggsofresistantstrainsofwormsfromthemselvesorotheranimals.Theauthors’recommendationisthatoperationsresearchshouldbeundertakentoexaminetheeffectonresistanceinintestinalnematodewormsofalternatingalbendazoleormebendazoletreatmentforseveralyearswithacombinationofpyrantelandoxantel.ThiscombinationisavailableinIndonesia,Malaysia,Singapore,thePhilippinesandVenezuela.Ifthisprovedeffective,pyrantel/oxantelwouldneedtobeaddedtotheWHOEssentialMedicineslist,anditislikelythatincreaseddemandwouldleadtoinexpensivesuppliesofpyrantel/oxantelbeingsourcedinternationally.

3 ECONOMICANALYSIS

Thecostofdewormingcanbelookedatinthreeways:asthecostperpersontreated,asthecostperinfectedpersontreated,andasthecostperdiseasedpersontreated.Whentheprevalenceofinfectionapproaches100%,asitcaninplaceswheretransmissionisveryintense,thenthecostperinfectedpersontreatedapproachesthecostperpersontreated.Howeverwhentheprevalenceis<100%someuninfectedpeoplewillbetreatedunnecessarily,sothecostperinfectedpersonrises.ThisisshowninTable1(fromHalletal.,2008)forprevalencesofinfectionbetween90%and20%;thelowervalueisthethresholdatwhichmasstreatmentisrecommendedbytheWHO(2006).Whentheprevalenceisaslowas20%,thecostofdrugsperinfectedpersontreatedis5timesthecostperpersontreated.Thecostperdiseasedpersontreateddependsontheprevalenceofinfection(p),themeanwormburden(M)andtheclumpingparameterkofthenegativebinomialdistributioninwhichp=1–(1+M/k)‐k.Valuesofkforintestinalnematodewormsaretypically<1,whichmeansthatwormsarehighlyaggregatedsuchthatasmallproportionofindividualshavemoderatetoheavyinfectionsandarelikelytobediseased.Thenegativebinomialdistributioncanbeusedtoestimatevaluesfortheprevalenceofinfectionandkforanymeanwormburden,assuming

17

thatkvarieslinearlywithM(Guyattetal.,1990),andtoestimatetheproportionofindividualsthathavemorethananygiventhresholdofworms.Table1showsthatwhentheprevalenceis20%thenonly0.01%ofindividualswillhave≥10worms,arelativelylowthresholdatwhichdiseasemightoccursothat,ifmasstreatmentwasgiven,itwouldcostUSD486perdiseasedpersontreated.(Forathresholdof≥15wormsthecostisaboutUSD27,000perdiseasedpersontreatedandforathresholdof≥20worms,USD1.4millionperdiseasedpersontreated.)TheanalysispresentedinTable1isbasedondataforasinglespeciesofworm,Ascarislumbricoides.Becausethedistributionofdiseaseforanyspeciesisdrivenlargelybytheextenttowhichwormsareaggregatedinafewhosts,theclumpingparameterkisofgreatimportance.Asvaluesofkrangefrom0.03–0.6forhookwormsandfrom0.2–0.4forwhipworms(Anderson&May,1991),theyindicateagreaterdegreeofaggregationofthesespeciesthanforroundworms,forwhichvaluesofkrangefrom0.3–0.9(Anderson&May,2001).SimilarlythecoefficientsintheequationlinkingkandM(thewormburden)inTable1havebeenestimatedforroundworm,butarenotknownsowellfortheotherspecies.Hencethenumericalvaluesinthetablearenotnecessarilyaperfectguidetocasesofinfectionwithwhipwormorhookworm,orindeedmixedinfectionswithmorethanonekindofintestinalworm.Howeverthemainpoint,thatdiseaseishighlyaggregated,especiallyatlowprevalencesofinfection,islikelytoholdtrue.TheanalysispresentedinTable1isalsodependentonthethresholdusedtoclassifyanyindividualasdiseasedornot,somethingthatisnotknown,mainlybecausetheeffectsofwormsarehardtoquantifyinthefirstplaceandvarybetweenspecies.Forexample,assuminganequalsexratio,10roundwormsweighabout25g,15wormsweigh35gand20wormsweighabout45g,whichisabout0.3%ofthebodyweightofanunderweight6‐yearoldschoolgirlof15kg(WHO,1983).ForthehookwormspeciesAncylostomaduodenale,10wormsareestimatedtocauseabloodlossof2ml/day,15worms3ml/dayand20worms4ml/day;thesamefiguresfortheotherhookwormspecies,Necatoramericanus,are0.4,0.6and0.8ml/dayofbloodaday,asthetwospeciesdifferinthevolumeofbloodlossthattheycause(RocheandLayrisse,1966).Bothspeciesnowco‐occurwidelythroughouttheworld.Thelossofbloodcausedbythesamenumberofwhipwormsisestimatedtobe0.05,0.07and0.1ml/dayofblood(Layrisseetal.,1967)althoughtheinflammatoryresponsetotheheadofthewormsembeddedintissuesmaybemoreimportant.Thesedifferenceshighlightthatwedonotknowhowtoquantifydiseaseduetoasinglewormspecies,letaloneconcurrentinfectionsoftwoofmorespecies,sothethresholdnumberofwormsusedtodefinediseaseinpreschoolchildrenorschoolchildren,whethermalnourishedornot,isarbitrary.Butwhateverthethreshold,theriskofhavingmorethansay,10worms,becomesextremelysmallwhentheprevalenceislowerthan50%,sothatfewindividualswillbenefit.Table1suggeststhatwhentheWHOthresholdof20%prevalenceisappliedtogivemasstreatment,thenonly1inevery10,000peopletreatedwillhave10ormoreworms.Wethink

18

thatthisprevalencethresholdistoolowtobeacost‐effectivewayoftreatingdiseasecausebyworms.Weproposethefollowingthresholdsandfrequenciesoftreatmentbasedonthecombinedprevalenceofinfectionswithroundworm,hookwormandwhipworminschoolchildren(Halletal..,2008):

• ifprevalenceis<40%,treatonlymoderatelyorseverelyunderweight,wastedoranaemicchildrenorchildrenwithdiagnosedinfections;

• ifprevalenceis40–<60%,givemasstreatmentonceayear,checkingevery2ytoseeifprevalenceis<40%;

• ifprevalenceis60–<80%,givemasstreatmenttwiceayear,checkingafter2ythatprevalenceis40‐<60%thengiveannualmasstreatments;

• ifprevalenceis80–100%,givemasstreatmentthreetimesayear,checkingafter2y,sothatifprevalenceis60–<80%treattwiceayearandif40‐<60%thentreatonceayear.

Inadditionschoolsshouldbesupportedtoinstallandmaintainsanitarylatrines,separatelyforboysandgirls;providecleanwaterandsoaptowashhands;andtopromotehealthybehavioursthroughhealtheducation.Thecostsandcost‐effectivenessoftreatingchildrenaged2to14basedonthetwosetsofthresholdshavebeenestimatedfor103countriesplusIndiaandChinaatthesubnationallevelsofstatesandprovincesrespectively,becauseoftheirlargesize(Halletal.,2008).(ThecountriesinAsiawhichwereformerlymembersoftheSovietUnionwereexcluded,asprevalencedatawerenotavailable).TheanalysisfoundthattheWHOthresholdsimplyhigherannualtreatmentcosts($276millionannuallycomparedto$224millionfortheproposednewthresholds),andimplyasmallerproportionofexpendituresgoingtochildrenwhoareinfected(61%ascomparedto73%),andasmallerproportionofexpendituresgoingtochildrenwith10ormoreworms(21%ascomparedto31%)(Halletal.,2008).Therehavebeenseveralanalysesofthecostsandcost‐effectivenessofdewormingdeliveredthroughschoolsandtopreschoolchildrenduringChildHealthDays.MiguelandKremer(2004)examinedaprogrammeinwhichKenyanschoolchildrenweretreatedtwiceayearwithalbendazoleinschoolsinwhichtheprevalencewithintestinalnematodewormswas>50%andannuallywithpraziquantelinschoolsinwhichtheprevalenceofschistosomiasiswas>30%.ThecostsoftheprogrammewerebasedonapriorestimateofUSD0.49peryearperchild(PCD,1999).Itwasestimatedthatdewormingdecreasedabsenteeismby25%andwasclaimedtobeamorecost‐effectivewayofincreasingschoolparticipationthananyalternative,includingschoolsubsidies(Miguel&Kremer,2004).Itcanbeestimatedfromthesefigures(Hortonetal.2008)thatthepresentvalueofincreasedwagesassociatedwithincreasedschoolachievementgivesabenefit:costratioof60:1,whichwouldfallifitwasnecessarytohireadditionalteachersbecauseofincreasedattendance,butisstillover3:1.Therewerealsoexternaleffectsonuntreatedchildreninparticipatingschools,probablyduetoreducedtransmissionofintestinalnematodeworms,andtherewereexternaleffectsonchildreninnon‐participatingschools,

19

probablythrougheffectsonthetransmissionofschistosomiasis,althoughthelatterreliedonnon‐experimentalmethodsofestimation(Miguel&Kremer,2004).Table1.EstimatesofthedrugcostsperpersontreatedonceforinfectionswithAscarislumbricoidesusingadrugcostingUSD0.03perdose.

As there is a linear relationship between the mean worm burden (M) and the clumpingparameterk forA.lumbricoides (Guyatt andBundy, 1991), this relationshipa has been used toestimatevaluesofkformeanwormburdensinstepsofroughlyM/2,startingatM=30,withtheaim of giving proportions infected (p) from 20 – 95% using the equation for the negativebinomialdistribution,inwhichp=1‐(1+M/k)‐k.Thenegativebinomialfunction(pnbinom)oftheRprogramme (RDevelopmentCoreTeam,2007)was thenused toestimate fromvaluesofkandMtheproportionofpeoplehavingmorethan10,15or20worms,arbitrarythresholdsthatcouldbeusedtoclassifyawormburdenasmoderateorgreater.Thecostswerethenestimatedper person, per infected person based on p, and per diseased person treated based on theproportionsinfectedwith≥10,≥15or≥20A.lumbricoides.Somevalueswere<0.0001.

ak=a+bMinwhicha=0.334andb=0.0172(GuyattandBundy1991)Source:Hall,etal(2008)InaschoolbaseddewormingprogrammeinUgandaBrookeretal.(2008)estimatedthatthecostofdrugswasUSD0.22perchildtreatedperyearforthreeyearsofwhichalbendazolecost10%,praziquantelcost82%andtheremainderwasthecostofshippingtoKampala.AsgenericdrugswereusedthecostwaslessthanathirdofsimilarprogrammesinGhanaandTanzania,whichusedproprietarydrugs(PCD,1999).ThetotalcostperchildtreatedwasUSD0.54peryear,sothatthedrugswere40%oftheoverallcost,whichconfirmsthatschool‐basedprogrammesarearelativelyefficientwaytoadministersometypesofhealthinterventions.HoweverthecostsperchildvariedbetweendistrictsfromUSD0.41toUSD0.91andwerelowestindistrictswiththehighestpopulationdensity,mainlybecausefixedcostssuchasallowancesfordistrictofficialswerespreadoverlargernumbersofchildrentreated.Therewas

Costs in USD per Proportion infected

With Proportion

Infected

(p)

Mean

Burden

(M)

Clumping

parameter

(k)a ≥10 worms

≥15 worms

≥20 worms

Person Infected

Person

Diseased

person

≥10 worms

Diseased

person

≥15 worms

Diseased

person

≥20 worms

0.95 30.0 0.850 0.6860 0.5817 0.4956 0.03 0.03 0.04 0.05 0.06

0.90 20.0 0.678 0.5430 0.4322 0.3483 0.03 0.04 0.06 0.07 0.09

0.85 14.6 0.585 0.4400 0.3316 0.2545 0.03 0.04 0.07 0.09 0.12

0.80 11.0 0.523 0.3563 0.2539 0.1852 0.03 0.04 0.08 0.12 0.16

0.70 6.4 0.444 0.2222 0.1381 0.0886 0.03 0.04 0.14 0.22 0.34

0.60 3.6 0.396 0.1159 0.0580 0.0302 0.03 0.05 0.26 0.52 0.99

0.50 2.0 0.368 0.0460 0.0162 0.0060 0.03 0.06 0.65 1.85 5.01

0.40 1.2 0.354 0.0149 0.0033 0.0008 0.03 0.07 2.01 9.09 38.88

0.30 0.6 0.345 0.0016 0.0001 0.0000 0.03 0.10 18.80 230.81 2,658.33

0.20 0.3 0.339 0.0001 0.0000 0.0000 0.03 0.15 485.68 27,335.62 1,435,480.19

20

a52.5%reductionintheproportionofanaemiacasesinacohortofchildrenover3yearsofmonitoring,from35.2%to18.5%,buttherewasnocontrolgroup.Thusthetrueeffectoftreatmentcouldnotbeseparatedfromnaturalincreasesinhaemoglobinconcentrationwithageamongboys,fromtheeffectsofthedietonironandvitaminintake,andperhapschangesinmalariatransmissionoverthe3yearperiodasaresultofcontrolprogrammes.Thecost‐effectivenesspercaseofanaemiapreventedwasestimatedtobeUSD3.19,whichalsovariedbydistrict,rangingfromUSD1.70toUSD9.51,butcostswereprobablylargerbecausethechangeintheprevalenceofanaemiawasover‐estimated.Theeffectivenessoftheprogrammealsodependedonepidemiologicaldifferencesasthemostcost‐effectivedistricthadthehighestinitialprevalenceofanaemia,butalsovariedwithotherunexplainedfactors.Thecost‐effectivenessoftheschool‐basedprogrammeinUgandacompareswithUSD7.43inmainlandTanzania(Guyattetal.,2001)andUSD3.57inZanzibar,Tanzania(Stoltzfusetal.,1998).Otherfactorsalsocontributedtothedifferencesincost,suchasthefrequencyoftreatmentandtheperiodoverwhichoutcomemeasurementsweremade.Ifwesummarizethedataforschoolchildren,thecostperchildperroundoftreatmentisapproximately$0.15,assumingthatdewormingcostsaresharedwithanotherprogram,suchasschistosomiasistreatment.Twostudiesprovideestimatesofcosts,butnotcost‐effectiveness,ofdewormingpreschoolchildren.Aldermanetal.(2006)foundastatisticallysignificanteffectof6‐monthlytreatmentwithalbendazolegivenatChildHealthDaysinUgandaonweightgainofchildren<7yage.AsingledoseofproprietaryalbendazolewasestimatedtoaddUSD0.21perchildtothecostofachildhealthdaybecauseaproprietarybrandofalbendazolewasusedforscientificreasons,butallthecostsofstaffwereattributedtogivingvitaminAandotheractivities.Theannualcostforbi‐annualHealthDayswasUSD0.42.Substitutinggenericdrugscouldreducethecostconsiderably,howeveritisnotreasonabletoattributeallthestaffcoststotheotheractivities.FiedlerandChuko(2008)estimatedthecostofdistributingalbendazoleinEthiopiatwiceperyearduringChildHealthDaysatUSD0.56perchildperroundoftreatment,andproratedstaffcostsamongthevariousactivities:deworming,givingvitaminA,andnutritionscreening.Thusthecostperchildtodewormpreschoolchildrenisabout$0.25perchildperround,aslongasitcanbecombinedwithanexistinghealthinterventionsuchasChildHealthDaysorcanbeincorporatedwithintheprimaryhealthcaresystem,ifcoverageisgood.Bobonisetal.(2006)analyzedthecostofprovidingdewormingandironsupplementsinIndianpreschools.Childrenaged2‐6yearsinpreschoolsinpoorcommunitiesineasternDelhiweregivenalbendazoleandironsupplementsproviding33.3mgofelementalironwithfolicacidforthreeconsecutivedays,bothadministeredthreetimesperyearata“healthcamp”.Additionally,allchildrenreceived200,000IUvitaminAonceayearasdidallchildrenintwootherwiseuntreatedcontrolgroups.Only30%ofchildrenwereinfectedwithworms.Thereweresignificantdifferencesbetweengroupsinthegaininz‐scoresofweight‐for‐ageandweight‐for‐height,butthetreatedgrouphadsignificantlylowermeanvaluesatbaseline(Bobonisetal.,2006).Therewasnosignificantchangeinhaemoglobinconcentrationbut

21

participationratesinpreschoolincreasedbyalmost6%therebyreducingabsenteeismby20%.ThecostwasestimatedatUSD1.70perchildperyear.Althoughnoformalcost‐effectivenessanalysiswasundertakenastheeffectsofironanddewormingcannotbeseparated,itconfirmsgenerallythefindingsofMiguelandKremer(2004).Therearenocost‐effectivenessstudiesinpreschoolchildren,toourknowledge.Wecanmakeestimatesoftheorderofmagnitudeoftheeffectofdeworminginpreschoolers,asfollows.Themostsignificantoutcomeofdewormingfromaneconomicperspectiveisthereductionorpreventionofanaemia,assumingthatthedietofanaemictreatedpeoplecontainssufficientnutrientstoenablehaemopoiesis;theeffectongrowthisimportantforhealthbutnotlargeenoughtohaveasignificanteconomicimpact.AccordingtoBhuttaetal.(2008)thesizeofthereductioninanaemiaduetodewormingchildrenrangesfrom4.4%to21%,withamedianvalueof13%,whichtranslatestoanincreaseinhaemoglobinof1.71(0.70‐2.73)g/L.HortonandRoss(2003,corrected2006)estimatedforarangeofdevelopingcountriesthatthepresentvalueofthemediancognitivelossattributabletoanaemiaisUSD15‐25perchild.Hencereducinganaemiaby13%couldleadtobenefitsofapproximatelyUSD2.60perpreschoolchildcomparedwithacostofdewormingofUSD0.40‐50perchild.Thus,thebenefit:costratiois5:1to6:1inthisage‐group,butwouldbereducedto2.4:1ifa5%discountratewereused.Theseestimatesareconservativeanddonottakeintoaccountbenefitsdueimprovementsinweightandheight.Thebenefit:costratioislikelyalsotovarywiththecoverageoftreatmentgiventothetargetgroup.Butifthepotentialimprovementinhaemoglobinconcentrationafterdewormingisconstrainedbyadeficiencyofanynutrientorbyanothercauseofanaemia,suchasmalaria,thenthemagnitudeofthepotentialimpactofdewormingwillbeunderestimated.AstudyinTanzaniaofgivingironandvitaminAtoschoolchildrenaftertreatinghookwormsandschistosomiasiswithalbendazoleandpraziquantelshowedanabsoluteincreaseinhaemoglobinconcentrationof22.1g/Landrelativeincreaseof18.5g/L(95%CI14.8–22.2)comparedwithanincreaseincontrols,whowereonlydewormed,of3.5g/L(Mwanrietal.,2000).ThissmallincreaseinthehaemoglobinconcentrationofcontrolsissimilartotheeffectestimatedbyBhuttaetal.(2008).ForthisreasonchildrenshouldbegivenacourseofmicronutrientssupplementsthatprovidevitaminA,folate,vitaminB12andiron,afterdeworming.TeachersinMaliandthePhilippineshavebeenshowntobeabletogiveweeklyirontabletstotheirpupilswithstatisticallysignificanteffectsonhaemoglobinconcentration(Halletal.,2002;Roschniketal.,2004).ThecostshavenotbeenestimatedbutamultivitaminsupplementforadultscostsUSD0.01pertabletifboughtinlargequantitiesinacountrysuchasBangladesh.ItisreasonabletouseacostofaboutUSD0.50perpreschoolchildtreatedperyear(tworoundsoftreatment),andaboutUSD0.30perschool‐agechildtreated(tworoundsoftreatment),forupto80%ofthetargetpopulation,butthisassumesthatgenericnotproprietarydrugsareusedandthatdewormingiscombinedwithanotherinterventionforpreschoolchildrenorisgiveninschools,thusreducingthedistributionanddeliverycosts.DistributingtreatmentsinChildHealthDaysislikelytobemoreefficientinlargecommunitiesthansmallones,butislesslikelytoreachremoteareas,althoughsuchareasarealsolikelytobelessdenselypopulated.Theimplicationisthatthecostsofincreasingcoveragearelikelytoincreaseascoveragealso

22

increases.Thecoverageofschool‐agechildrenwilldependonenrolmentrates,whichareimprovingasaresultofeffortstoachieveaMillenniumDevelopmentGoal.Butthereisnoreasonwhycoverageofchildrenwhoareenrolledandattendingschoolshouldnotapproach100%,especiallyifchildrenwhoareabsentfromschoolaretreatedlater.Wehaveundertakenelsewheredetailedcalculationsoftheannualtreatmentcostsofdeworming,includingbothdruganddeliverycosts(Halletal..,2008).For100%coverageofallchildrenin105developingcountries,thecostsare$224millionperyearusingtheproposed(morecost‐effective)treatmentthresholds,or$276millionperyearusingthecurrentWHOthresholds.Althoughthisisalargeannualexpenditure,thepotentialbenefitsintermsofimprovedchildhealthandeducationareverylarge,usingthebenefit:costratioof6:1(preschoolers)andevenhigherforschool‐agechildren.

4 Conclusions

Thispaperhasdiscussedanumberofpolicyrecommendations.Wehavefocussedonlyonintestinalnematodesandexcludedschistosomiasisandotherworms.Massdewormingofschoolchildrenisinexpensiveandsafe,withtheonlycaveatthatcareisneededwhentreatingadolescentgirlsovertheageof12ytobesuretheyarenotinthefirsttrimesterofpregnancy.Massdewormingofpreschoolersagedtwoandaboveisalsoinexpensiveandsafe.Operationalcostdataarelackingfordewormingpreschoolersundertheageoftwo,althoughdewormingisalsosafeforthisgroupusingsyrups.Operationalcostdataarealsolackingforpregnantwomen.Studiesandmodelssuggestthatmassdewormingiscost‐effectivewhentheprevalenceofinfectionwithintestinalnematodesexceeds50%,asfirstrecommendedbytheWHO(1996).TheWHO(2006)recommendationalsoallowfortreatmentthreetimesperyearwhereinfectionissevere.Cost‐effectivenessdatahoweverdonotsupportthe“preventivechemotherapy”approachoftreatingonceperyearforprevalenceratesaslowas20%,withoutfurtherevidence.Ourrecommendationisthatthethresholds40%,60%and80%offerasimplebasisonwhichtogivemasstreatment1x,2xor3xayear,andcovertheincreasingriskofdiseasewhileminimizingcosts.Albendazoleandmebendazolearebotheffectiveandlow‐costandhavetheadvantagethattheycanbeco‐administeredwithpraziquantelinplaceswhereindividualsareinfectedbothwithintestinalwormsandschistosomiasis.However,givenconcernsaboutthedevelopmentofdrugresistance,itwouldbebesttoalternatethemwithadrugthathasadifferentmechanismofeffect.Itisrecommendedthatoperationsresearchbeconductedalternatingalbendazole/mebendazolewithacombinationofpyrantelandoxantel;thedoseofthelattercouldbeestimatedbasedonheight,toavoidtheneedforweighingscalesorastandarddoseforagecouldbedetermined.Thiscouldlowertheriskofdevelopingdrugresistancetoalbendazoleandmebendazole,neitherofwhichisfullyefficaciousagainsthookwormsorwhipworms.

23

Asurveyoftheprevalenceofwormsinasampleofschoolsisrecommendedtodetermineiftreatmentiswarrantedandhowoftenitshouldbegiven.Faecaleggcountsshouldbeestimatedinchildreninasampleofschoolstoestimatethecurerateandtheeggreductionratetomonitordrugeffectivenessanddetectdrugresistance.Althoughsuchsurveysincreasecosts(byapproximately0.1%worldwide),theycanincreasecost‐effectivenessbyavoidingunwarrantedtreatments.Thecostsoftreatmentarenotinsubstantial;althoughthecostperschool‐agechildisonlyaboutUSD0.30peryearfortworoundsoftreatment(includingdistributioncosts)(and$0.50peryearperpreschoolchild),theannualglobalcostswouldbe$224million(proposednewthresholds)or$276million(WHOthresholds).However,thebenefitsintermsofimprovedhealth,nutritionandeducationarealsoverysubstantial,giventheveryhighbenefit:costratioofintervention.

24

REFERENCES Albonico,M.,Bickle,Q.,Haji,H.J.,Ramsan,M.,Khatib,K.J.,Montresor,A.,Savioli,L.,Taylor,M.(2002).Evaluationoftheefficacyofpyrantel‐oxantelforthetreatmentofsoil‐transmittednematodeinfectionsTransactionsoftheRoyalSocietyofTropicalMedicineandHygiene96,685‐690.Albonico,M.,Smith,P.G.,Hall,A.,Chwaya,H.M.,Alawi,K.S.,Savioli,L.(1994).ArandomizedcontrolledtrialcomparingmebendazoleandalbendazoleagainstAscaris,Trichurisandhookworminfections.TransactionsoftheRoyalSocietyofTropicalMedicineandHygiene88,585‐589.Alderman,H.,Konde‐Lule,J.,Sebuliba,I,Bundy,D.,Hall,A.(2006).EffectonweightgainofroutinelygivingalbendazoletopreschoolchildrenduringchildhealthdaysinUganda:clusterrandomizedcontrolledtrial.BritishMedicalJournal333,122‐126Anderson,R.M.,May,R.M.(1991).InfectiousDiseasesofHumans.DynamicsandControl.Oxford:OxfordSciencePublications.Beasley,N.M.,Hall,A.,Tomkins,A.M.,Donnelly,C.,Ntimbwa,P.,Kivuga,J.,Kihamia,C.M.,Lorri,W.,Bundy,D.A.(2000).ThehealthofenrolledandnonenrolledchildrenofschoolageinTanga,Tanzania.ActaTrop76,223‐229.Bennett,A.Guyatt,H.(2000).Reducingintestinalnematodeinfection:efficacyofalbendazoleandmebendazoleParasitologyToday16,71‐74.Bennett,S.,Woods,T.,Liyanage,W.M.,Smith,D.L.(1991).Asimplifiedgeneralmethodforclustersamplingindevelopingcountries.WorldHealthStatisticsQuarterly,44,98–106.BhuttaZ.A.,T.Ahmed,R.E.Black,S.Cousens,K.Dewey,E.Giugliani,G.A.Haider,B.Kirkwood,S.S.Morris,H.P.S.Sachdev,M.Shekar,(2008).Whatworks?Interventionsformaternalandchildundernutritionandsurvival.Lancet371,417‐440Bobonis,G.,E.Miguel,andC.Puri‐Sharma(2006).Anaemiaandschoolparticipation.JournalofHumanResources41(4),692‐721Brooker,S.,Marriot,H.,Hall,A.,Adjei,S.,Allan,E.,Maier,C.,Bundy,D.A.,Drake,L.J.,Coombes,M.D.,Azene,G.,Lansdown,R.G.,Wen,S.T.,Dzodozmenyo,M.,Cobbinah,J.,Obro,N.,Kihamia,C.M.,Issae,W.,Mwanri,L.,Mweta,M.R.,Mwaikemwa,A.,Salimu,M.,Ntimbwa,P.,Kiwelu,V.M.,Turuka,A.,Nkungu,D.R.,Magingo,J.(2001).Communityperceptionofschool‐baseddeliveryofanthelminticsinGhanaandTanzania.TropicalMedicineandInternationalHealth6,1075‐1083.Brooker,S.,Hotez,D,Bundy,D.(2008).Hookworm‐relatedanaemiaamongpregnantwomen:asystematicreview.PLosNeglectedTropicalDiseases2(9):e291.doi:10.1371/journal.pntd.0000291BrookerS,KabatereineNB,MyattM,StothardJR&FenwickA.(2005).RapidassessmentofSchistosomamansoni:thevalidity,applicabilityandcost‐effectivenessoftheLotQualityAssuranceSamplingmethodinUganda.TropicalMedicineandInternationalHealth200510:647‐658.Brooker,S.,N.B.Kabatereine,F.FlemingandN.Devlin,(2008).Costandcost‐effectivenessofnationwideschool‐basedhelminthcontrolinUganda:intra‐countryvariationandeffectsofscaling‐up.HealthPolicyandPlanning23,24‐35

25

Bundy,D.A.,Hall,A.,Medley,G.F.,Savioli,L.(1992a).Evaluatingmeasurestocontrolintestinalparasiticinfections.WorldHealthStatisticsQuarterly45,168‐179.Bundy,D.A.,Horton,J.,Wong,M.S.,Lewis,L.L.(1992b).ControlofgeohelminthsbydeliveryoftargetedchemotherapythroughschoolsTransactionsoftheRoyalSocietyofTropicalMedicineandHygiene84,115‐120.deSilva,N.,Guyatt,H.,Bundy,D.(1997).Anthelmintics.AcomparativereviewoftheirclinicalpharmacologyDrugs53,769‐788.deSilva,N.R.,S.Brooker,P.J.Hotez,A.Montresor,D.EngelsandL.Savioli(2003a).Soil‐transmittedhelminthinfections:updatingtheglobalpicture.TrendsinParasitology19,547‐551deSilva,N.R.,S.Brooker,P.J.Hotez,A.Montresor,D.EngelsandL.Savioli(2003b).Soil‐transmittedhelminthicinfections:updatingtheglobalpicture.DiseaseControlPrioritiesProjectWorkingPaperno.12.Availableathttp://www.dcp2.org/file/27/wp12.pdf,accessedSeptember2008.Engels,D.,Nahimana,S.,DeVlas,S.J.,Gryseels,B.(1997).VariationinweightofstoolsamplespreparedbytheKato‐Katzmethodanditsimplications.TropicalMedicineandInternationalHealth2,265‐271.Fentiman,A.,Hall,A.,Bundy,D.(2001).Healthandculturalfactorsassociatedwithenrolmentinbasiceducation:astudyinruralGhana.SocialScienceandMedicine52,429‐439.Fiedler,J.L.,Chuko,T.(2008).Thecostofchildhealthdays:acasestudyofEthiopia’sEnhancedOutreachStrategy(EOS).HealthPolicyandPlanning23:222‐233.Gulani,Al,NagpalJ.,Osmond,C.,Sachdev,H.P.S.(2007)Effectofadministrationofintestinalanthelminticdrugsonhaemoglobin:systematicreviewofrandomizedcontrolledtrials.BritishMedicalJournal334,1095‐1097.Gustafsson,L.L.,Beerman,B.,Abdi,Y.A.(1987).Handbookofdrugsfortropicalparasiticinfections.London,TaylorandFrancis.Guyatt,H.L.,Bundy.D.A.(1991).Estimatingprevalenceofcommunitymorbidityduetointestinalhelminthes:prevalenceofinfectionasanindicatoroftheprevalenceofdisease.TransactionsoftheRoyalSocietyofTropicalMedicineandHygiene85,778–782.Guyatt,H.L.,Bundy,D.A.,Medley,G.F.,Grenfell,B.T.(1990).TherelationshipbetweenthefrequencydistributionofAscarislumbricoidesandtheprevalenceandintensityofinfectioninhumancommunities.Parasitology101,139‐143.Guyatt,H.L.,Brooker,SA.,Kihamia,C.M.,Hall,A.andBundy,D.A.(2001).Evaluationofefficacyofschool‐basedanthelmintictreatmentsagainstanaemiainchildrenintheUnitedRepublicofTanzania.BulletinoftheWorldHealthOrganization,79,695‐703Hall,A.(1982).Intestinalhelminthsofman:theinterpretationofeggcountsParasitology85,605‐613.Hall,A.(2007).Micronutrientsupplementsforchildrenafterdeworming.LancetInfectiousDiseases7,297‐302.Hall,A.,Hewitt,G.,Tuffrey,V.anddeSilva,N.(2008).Areviewandmeta‐analysisoftheimpactofintestinalwormsonchildgrowthandnutrition.MaternalandChildNutrition4,118‐236

26

Hall,A.andHolland,C.(2000)GeographicalvariationinAscarislumbricoidesfecundityanditsimplicationsforhelminthcontrol.ParasitologyToday16,540‐544.Hall,A.,Horton,S.anddeSilva,N.(2008).Thecostsandcost‐effectivenessofmasstreatmentforintestinalnematodeworminfectionsusingdifferenttreatmentthresholds(mimeo.).Hall,A.andNahar,Q.(1994).AlbendazoleandinfectionswithAscarislumbricoidesandTrichuristrichiurainchildreninBangladesh.TransactionsoftheRoyalSocietyofTropicalMedicineandHygiene88,110‐112.Hall,A.,Nokes,C.,Wen,S.T.,Adjei,S.,Kihamia,C.,Mwanri,L.,Bobrow,E.,deGraft‐Johnson,J.,Bundy,D.(1999).Alternativestobodyweightforestimatingthedoseofpraziquantelneededtotreatschistosomiasis.TransactionsoftheRoyalSocietyofTropicalMedicineandHygiene93,653‐658.Hall,A.,Roschnik,N.,Ouattara,F.,Toure,I.,Maiga,F.,Sacko,M.,Moestue,H.,Bendech,M.A.(2002).ArandomisedtrialinMalioftheeffectivenessofweeklyironsupplementsgivenbyteachersonthehaemoglobinconcentrationsofschoolchildren.PublicHealthNutrition5,413‐418.Horton,S.,Alderman,H.andRivera,J.(2008).CopenhagenConsensus2008ChallengePaper:HungerandMalnutrition.Availablehttp://www.copenhagenconsensus.comHorton,S.andRoss,J.(2003).Theeconomicsofirondeficiency.FoodPolicy28(1),51‐75.SeealsoHorton,S.andJ.Ross,2006.Corrigendum.FoodPolicy32,141‐143.Keiser,J.andUtzinger,J.(2008).Efficacyofcurrentdrugsagainstsoil‐transmittedhelminthinfections.JAMA299:1937–1948.KotzeA.C.andKoppS.R.(2008)Thepotentialimpactofdensitydependentfecundityontheuseofthefaecaleggcountreductiontestfordetectingdrugresistanceinhumanhookworms.PLoSNeglectedTropicalDiseases2:e297.Layrisse,M.,Roche,M.,Aparcedo,L.,Martínez‐Torres,C.(1967).BloodlossduetoinfectionwithTrichuristrichiura.AmericanJournalofTropicalMedicineandHygiene16,613‐619.Miguel,E.andM.Kremer(2004).Worms:IdentifyingImpactsonHealthandEducationinthePresenceofTreatmentExternalities,Econometrica72,159‐217Montresor,A.,Ramsan,M.,Chwaya,H.M.,Ameir,H.,Foum,A.,Albonico,M.,Gyorkos,T.W.,Savioli,L.(2001).SchoolenrollmentinZanzibarlinkedtochildren'sageandhelminthinfectionsTropicalMedicineandInternationalHealth6,227‐231.Mulligan,J‐A,Fox‐RushbyJ.A.,Adam,TandMillsA.(2003,revised2005)Unitcostsofhealthcareinputsinlowandmiddleincomecountries.DCPPWorkingPaperno.9Availableathttp://www.dsp2.org/file/24/wp9.pdf,accessedAugust2008Mwanri,L.,Worsley,A.,Ryan,P.,Masika,J.(2000).SupplementalvitaminAimprovesanemiaandgrowthinanemicschoolchildreninTanzania.JournalofNutrition130,2691‐2696.Nock,I.H.,Aken’Ova,T.andGaladima,M.(2006).Deworming:addingpublichealtheducationtotheequation.TrendsinParasitology23,7–8.Olds,G.R.,King,C.,Hewlett,J.,Olveda,R.,Wu,G.,Ouma,J.,Peters,P.,McGarvey,S.,Odhiambo,O.,Koech,D.,Liu,C.Y.,Aligui,G.,Gachihi,G.,Kombe,Y.,Parraga,I.,Ramirez,B.,Whalen,C.,Horton,R.J.,Reeve,P.(1999).Double‐blindplacebo‐controlledstudyofconcurrentadministrationofalbendazoleandpraziquantelinschoolchildrenwithschistosomiasisandgeohelminthsJournalofInfectiousDiseases179,996‐1003.

27

PCD[PartnershipforChildDevelopment](1999).Thecostoflarge‐scaleschoolhealthprogrammeswhichdeliveranthelminticsinGhanaandTanzania.ActaTropica,73,183‐204Pawlowski,Z.S.,Schad,G.A.,Stot,G.J.(1991).Hookworminfectionandanaemia.Approachestopreventionandcontrol.Geneva,WorldHealthOrganization.RDevelopmentCoreTeam(2007).R:Alanguageandenvironmentforstatisticalcomputing.RFoundationforStatisticalComputing,Vienna,Austria.ISBN3‐900051‐07‐0,URLhttp://www.R‐project.org.Roche,M.,Layrisse,M.(1966).Thenatureandcausesof"hookwormanemia".AmericanJournalofTropicalMedicineandHygiene15,1029‐1102.Roschnik,N.,Parawan,A.,Baylon,M.A.,Chua,T.,Hall,A.(2004).WeeklyironsupplementsgivenbyteacherssustainthehaemoglobinconcentrationofschoolchildreninthePhilippinesTropicalMedicineandInternationalHealth9,904‐909.Steketee,R.W.(2003).Pregnancy,nutritionandparasiticdiseases.JournalofNutrition133,1661S‐1667SStoltzfusR.J.,H.M.Chwaya,A.Montresor,J.M.Tielsch,J.K.Jape,M.AlbonicoandL.Savioli(2004).Lowdosedailyironsupplementationimprovesironstatusandappetitebutnotanaemia,whereasquarterlyanthelminthictreatmentimprovesgrowth,appetiteandanaemiainZanzibaripreschoolchildren.JournalofNutrition134,348‐356Stoltzfus,R.J.,Albonico,M.,Chwaya,H.M.,Tielsch,J.M.,Schulze,K.J.,Savioli,L.(1998).EffectsoftheZanzibarschool‐baseddewormingprogrammeonironstatusofchildren.AmericanJournalofClinicalNutrition68,179‐186.Todd,J.,F.Mosha,R.Balira,A.I.N.,Obasi,J.,Changalucha,D.A.,Ross,D.C.W.,Mabey,M.,Plummer,M.,Grosskurth,H.(2004).Thesexualhealthofpupilsinyears4to6ofprimaryschoolsinruralTanzania.SexuallyTransmittedInfections80,35–42.Torlesse,H.,Hodges,M.(2000).Anthelminthictreatmentandhaemoglobinconcentrationsduringpregnancy.Lancet356,1083.WHO(1983).Measuringchangeinnutritionalstatus.Geneva,WorldHealthOrganization.WHO(1991).BasicLaboratoryMethodsinMedicalParasitology.Geneva,WorldHealthOrganization.WHO(1994).Benchaidsforthediagnosisofintestinalparasites.Geneva,WorldHealthOrganization.WHO(1995).WHOModelPrescribingInformation.DrugsUsedinParasiticDiseases.2ndedn.Geneva,WorldHealthOrganization.WHO(1996).ReportoftheWHOInformalConsultationontheuseofchemotherapyforthecontrolofmorbidityduetosoil‐transmittednematodesinhumans.Geneva,WorldHealthOrganization.WHO(1999).ReportoftheWHOInformalConsultationonmonitoringofdrugefficacyinthecontrolofschistosomiasisandintestinalnematodes.Geneva,WorldHealthOrganization.WorldHealthOrganization(2001).Schistosomiasisandsoil‐transmittedhelminthinfections.Fifty‐fourthWorldHealthAssemblyResolutionWHA54.19.Geneva.Availableathttp://www.who.int/wormcontrol/documents/wha/en/(lastaccessed29/8/2008)

28

WHO(2002).Preventionandcontrolofschistosomiasisandsoil‐transmittedhelminthiasis.TechnicalReportSeries912.Geneva,WorldHealthOrganization:63pp.WHO(2003).ReportoftheWHOinformalconsultationontheuseofpraziquantelduringpregnancy/lactationandalbendazole/mebendazoleinchildrenunder24months.Geneva,WorldHealthOrganization.WHO/CDS/CPE/PVC/2002.4.WHO(2005).WorldHealthOrganizationModelFormulary.Geneva,WorldHealthOrganization.WHO(2006).Preventivechemotherapyinhumanhelminthiasis.Geneva,WorldHealthOrganization.WHO(2008).Soil‐transmittedhelminthiasis:progressreportonnumberofchildrentreatedwithanthelminthicdrugs.Weeklyepidemiologicalrecord.83:Nos.27/28,237‐252.WHO‐CHOICE(2008).(CHoosingInterventionsthatareCost‐Effective).Priceoflocal(non‐traded)goods.Availableathttp://www.who.int/choice/costs/prog_costs/en/index.html,accessedAugust2008WHO/UNICEF(2004).HowtoadddewormingtovitaminAdistribution.Geneva,WorldHealthOrgan

Copenhagen Consensus 2008Each day, decisions are made about global political priorities. Unfortunately, political decisions seldom take into account a comprehensive view of the costs and benefits of implementing one solution and not the other or solv-ing one problem instead of another.

In the Copenhagen Consensus 2008, an expert panel of eight outstand-ing economists, including 5 Nobel Laureates, delivered a ranked list of the most promising solutions to the ten of the most pressing challenges facing the world today: Global Warming, Terrorism, Hunger and Malnutrition, Water and Sanitation, Air Pollution, Women and Development, Conflicts, Subsidies and Trade Barriers, Education and Diseases. Around 55 of the world’s lead-ing economists and specialists in the ten challenges compiled up-to-date analysis of the best solutions to ensure the best possible foundation for the prioritization of the solutions.

An overview of how to implement the top solutions from the Copenhagen Consensus 2008Five best practice papers have now been written on the top-solutions on the economists’ list from the Copenhagen Consensus 2008. These papers provide clear and focused empirical recommendations on the best practice and the costs and benefits of the solutions helpful for governments, inter-national organizations, NGO’s and philanthropists in both developed and developing countries.

The Copenhagen Consensus Center analyzes the world’s greatest chal-lenges and identifies cost efficient solutions to meeting these challenges. The Center works with multilateral organizations, governments and other entities concerned with mitigating the consequences of the challenges which the world is facing. With the process of prioritization, the center aims to establish a framework in which solutions to problems are prioritized according to efficiency based upon economic and scientific analysis of distinct subjects.

The Copenhagen Consensus Center is headed by Bjørn Lomborg

Copenhagen Consensus Center Copenhagen Business School • Solbjerg Plads 3 • DK-2000 Frederiksberg • Denmark

Phone +45 3815 2254 • info.ccc@cbs.dk • www.copenhagenconsensus.com