View
229
Download
0
Category
Preview:
Citation preview
7/27/2019 Banerjee Thesis
1/38
ChildHealthandImmunizationStatus
inanUnregisteredMumbaiSlum
AThesisSubmittedtotheFacultyoftheDepartmentofGlobalHealthand
PopulationHarvardSchoolofPublicHealthinPartialFulfillmentofthe
RequirementsfortheDegreeofMastersofScience.
Boston,Massachusetts
JoyaBanerjee
14May,2010
7/27/2019 Banerjee Thesis
2/38
2
Acknowledgements
Thisthesisdrawsuponfindingsfromathree-yearurbanhealthstudy(2009-2011)bytheHarvardSchool
ofPublicHealthandtheIndianresearchcollectivePUKAR(PartnersinKnowledge,ActionandResearch),
supportedbytheRockefellerFoundation.Thestudy,ExploringthePhysicalandSocialDeterminantsof
Urban Health is currently underway in the unregistered urban slum settlement, Kaula Bandar inMumbai, India. My sub-study on urban child health and immunization status was conducted in
collaboration with my research partnerTejal Shitole, from June-August 2009 andDecember-January
2009/10.
IwouldliketothankAnitaPatil-Deshmukh,ExecutiveDirectorofPUKAR,forbeingmysecondthesis
reader,abrilliantmentorandsupervisor,andaboveall,ateacherofformativelessonsthatcouldnever
belearnedfromabookorinaclassroom.IwouldliketoexpressmydeepgratitudetoTejalShitole,
without whom this project would not have been possible, aswell as to Kiran Sawant and Shrutika
Shitole,fortheirstrongcommitmenttotheKaulaBandarcommunity,theirpatience,deepknowledgeof
thecommunity,andcreativenessindevisingsolutionstoentrenchedproblemsinIndiasresource-poor
populations.
DavidBloom,myfirstthesisreader,hasbeenaconstantsourceofinspirationandinsight-thankyoufor
including me in this formidable experience. Adriane Lesser and Heather Lanthorn, HSPH students
workingwithmeduringthefirstpartofthestudyinMumbai,wereincrediblealliesintheirabilityto
applytheirknowledgeandskillstotheproblemsathand,andtohelpmeovercomemanychallenges
andconfusionsduringthe surveydesignandresearchprocess. Iowemysinceregratitudeto Jennifer
OBrien andMarija Ozolins for their diligence, research assistance, logistical support, and continued
feedback.SiddharthAgarwalwaskindenoughtoprovidehiswealthofexpertiseonchildimmunization
inIndiaandgraciouslymadehimselfavailabletomeformymanyquestions.KirtiGheiandSimaShah,
whohelpedturnadauntingideaintoanactionableresearchproject,basedontheirprolificexperiences
studyinghealthsystemsinIndia,wereinvaluablesupporters.
Thisthesishasbeenreadandapprovedby:
__________________________________________________
ProfessorDavidE.Bloom
__________________________________________________
&Dr.AnitaPatil-Deshmukh
7/27/2019 Banerjee Thesis
3/38
3
TableofContents
Acknowledgements.............................................................................................................................2
Abstract...............................................................................................................................................4
Introduction.........................................................................................................................................5
I.Urbanizationandthegrowthofslums..............................................................................................6
India:GrowingCitiesandUrbanSlums......................................................................................7
II.ChildHealth&ImmunizationCoverageinIndia..............................................................................10
II.ShortcomingsofUrbanHealthDeliverySystems............................................................................12
BarrierstoAccess....................................................................................................................12
AcceptabilityofHealthInterventionsandHealthEducation....................................................13
IV.TheCaseofKaulaBandar..............................................................................................................14
Methods.................................................................................................................................14
Results:KaulaBandarEthnographyandDemographics...........................................................15
KBDemographics............................................................................................................17
ChildHealthandMothersHealthSeekingBehavior........................................................18
MothersHealthKnowledgeaboutImmunization............................................................20
TheHealthImpactofTenure...........................................................................................22
QuantitativeFindings..............................................................................................................23
Methods..........................................................................................................................23
1)Outcome:ImmunizationStatus...................................................................................24
2)Outcome:Mortality.....................................................................................................25
3)Outcome:MothersHealthKnowledgeaboutImmunization........................................25
V.TheWayForward...........................................................................................................................26
TheEconomicBenefitsofImmunization..................................................................................26
StructuringanEffectiveIntervention.......................................................................................27
Incentives:TheLastMile......................................................................................................29
V.Conclusion.....................................................................................................................................30
WorksCited.......................................................................................................................................31
Exhibit1:ChildHealthSurveyData....................................................................................................35
Exhibit2:TheNotificationProcessandtheImpactofTenureonHealth.............................................37
7/27/2019 Banerjee Thesis
4/38
4
Abstract
Asof2007,morethanhalfoftheworldspopulationwaslivinginurbanareas,withaboutone-third
(fully1billionpeople) livinginurbanslums.Indiaalone ishometoroughly170millionslumdwellers,
one-thirdofwhomarechildrenunderfiveyearsofage. Thesqualid,congestedlivingenvironmentof
such slums includes limited access tobasic services such aspipedwaterand improved sanitationwhoseabsenceincreasestheriskofinfectiousdiseaseamongchildren.Thisthesisusesthecaseofan
unregistered urban slum, Kaula Bandar (KB), inMumbai, India, toexaminethe determinants of child
mortalityandimmunizationcoverageusingprimaryquantitativeandqualitativedatafromahousehold
survey(n=226households)andfocusgroups.Resultsindicatethatalthoughimmunizationservicesare
widelyavailableinurbancenters,a knowledge-actiongapkeepsimmunization rates lowandchild
mortalityhighinslumcommunities.Inparticular,lackofknowledgeabouttheprotectivebenefitsof
immunization,lackoftrustedchannelsofinfluence,andsystemicbarrierstoaccessinghealthservices
(suchascost,providerdiscrimination,anddistrustofthesystembynon-tenuredslumdwellersatriskof
losing their homes) leave the majority of urban slum-dwelling children unprotected from vaccine-
preventable diseases. A quantitative analysis of the determinants of childmortality, immunization
status,andmothershealthknowledgeofimmunizationrevealsthatunvaccinatedchildreninKBare3.2timesmorelikelytodiebeforetheageof5thanvaccinatedchildren.Dataonhealthdeterminantsand
outcomes fromchildren ingovernment-registered slums arecomparedwith those from unregistered
communities, revealing profound disparities not just between urban non-slum and urban slum
populations, butalsowithin theslum gradient. Forexample, roughly29%of children inKBare fully
immunized,comparedwith69%ofchildreninregisteredslumsreceivingoutsideservices.Similarly,the
infantmortalityrateinKB(58per1,000livebirths)ismorethandoublethatofregisteredslums(25per
1,000livebirths).Lastly,toaccountfortheuniqueconstraintsfacedbyurban-slumdwellers,thethesis
proposescontextuallyappropriateinterventionsthatcouldincreasevaccinationratesandputIndiaon
thepathtowarduniversalimmunizationcoverage.
7/27/2019 Banerjee Thesis
5/38
5
Introduction
Asof2007,morethanhalfoftheworldspopulationwaslivinginurbanareas,andoneinthreecity-
dwellerswas living in anurban slum (UN-HABITAT, 2008/9). Thereare approximately 1 billion slum
dwellersintheworld(UN-HABITAT2006/7),andthisnumberisprojectedtogrowto1.4billionbythe
year 2020 (UN-HABITAT, 2008/9). The growthof cities has had positiveimplications formanyurbancitizens:onaverage,urbanareaspresentgreateremploymentopportunities,increasedaccesstohealth
services, higher-quality educational opportunities, and modern amenities and technologies. These
featureshavecontributedtogreatereconomicandhealthoutcomesamongurbandwellers,including
lowerinfantandchildmortalityratescomparedtoruralregions(NFHS-3,2005/6).
Amidtheprosperityofthegreaterurbanlandscape,however,existsaplethoraofslumsthathavenot
necessarilyreapedthebenefitsofgreatercontemporarycityliving.Thestarklivingconditionsinslums
characterized by extreme population density, poor sanitation, and a lack of access to basic health
servicesencourageahostofhealthchallenges.AccordingtoUN-HABITAT,[s]lumdwellersdieearlier,
experiencemorehunger,havelesseducation,havefewerchancesofemploymentintheformalsector
andsuffermorefromill-healththantherestoftheinhabitantsofcities(UN-HABITAT,2008/9).This
environmentishighlyconducivetothedevelopmentandspreadofinfectiousdiseases,especiallyamong
immunonave populations such as children. Research demonstrates that children living in slums
shoulderadisproportionateburdenofdiseasecomparedtotheiradultcounterparts(Cornia,1987),and
the long-term consequences of these childhood illness can be devastating including permanent
stunting,physicaldisability,andlife-longimpairedcognition(Cornia,1987).
Althoughchildimmunizationhasavertedmorethan2.1milliondeathsannuallyandcountlessepisodes
ofillnessanddisability,20%ofchildrenworldwide,especiallythoselivingonlessthan$1aday,remain
unvaccinated (UNICEF2009).Childhood immunizationprovides an opportunity tomitigate childhood
morbidity and mortality in urban slums through the prevention of a host of infectious diseases.
Immunizationhasresultedinlarge-scalereductioninchildmorbidityandmortality,encouragingmany
tocategorizeitasoneofthegreatestachievementsinthehistoryofpublichealth(CDC,1999).Still,despitethepromiseofchildhoodvaccination,coverageamongurbanIndianchildrenremainslow,with
only69%ofchildrenaged12-
23 months considered as
fully vaccinated in Mumbai
(NFHS-3, 2005/6). Data
suggest that coverage rates
among urban poor children
largely living in slums are
approximately 40%, even
lowerthantheurbanaverage
of 58%. (Ghei et al, 2010).Figure 1 shows large
differences in immunization
status between urban poor
and non-poor, and also that
urban and rural rates are
verysimilar.
Figure1:IndiasChildImmunizationStatus
(UrbanHealthResourceCenter,2007)
7/27/2019 Banerjee Thesis
6/38
6
InIndiaalone,thereare169millionslumdwellers(IIPS,2009)andthecountryishometo17%ofthe
worldsslumdwellers(UN-HABITAT2008/9).AsthenumberandsizeofIndiasslumscontinuestogrow,
the number of children living in conditions conducive to infectious disease will also increase. This
upward trajectory encourages concentrated efforts to increase coverage of childhood immunization
acrosstheurbanlandscape.
Thisthesiswillusethecaseofanunregisteredurbanslum,KaulaBandar(KB)inMumbaitoexaminethe
determinantsofchildhealthandimmunizationstatus.Itwillexaminetheimpactof:
Barrierstoaccessinghealthinformationandservices Relevanceandacceptabilityofexistinghealthinterventions Landtenure Householdfactors(income,levelofeducation,religion,age,etc.) Healthknowledgeandhealth-seekingbehavior.
DrawingonprimarydatacollectedfromahouseholdsurveyinSummer2009,thisthesiswillassesschild
healthstatusinKBusingthefollowingindicators:childmortalityrates,immunizationstatusofchildren
underfiveyearsold,andmothershealthknowledgeandhealthseekingbehavior.ThecaseofKBwill
alsobeused toexamine theroleof tenurestatusindeterminingimmunizationcoverage;specifically,
dataonhealthoutcomesinKB,anunregisteredslum,willbecomparedtooutcomesfromregistered
slumsbasedoncensusdata,anddatafromtheNationalFamilyandHealthSurvey(NFHS-3,2005/6).
Lastly,throughananalysisofbarrierstoaccessingservices,thisthesiswilloffercontextuallyappropriate
interventionstoscale-upchildhoodimmunizationandimprovechildhealthinIndiasurbanslums.
I.Urbanizationandthegrowthofslums
TheUnitedNationsdefinesslums as communities characterized by insecure residentialstatus,poor
structuralqualityofhousing,overcrowding,andinadequateaccesstosafewater,sanitation,andother
infrastructure (UN-HABITAT,2003).From2000to 2030,theworldsurbanpopulationisprojectedto
grow at anannual growth rate of 1.8%,nearly double the rate of total population growth. In less-
developedareas,theurbangrowthratecurrentlyaverages2.3%(UNPopulationDivision,2005),andfor
slums,therateisdramaticallyhigher,atupto6%peryear(Chatterjee,2002).95%ofthisgrowthinthe
next 40yearswillbe indeveloping countrycities (UN-HABITAT,2008/9). The rapid growthof slums
represents a fundamental transformation of thephysical and social environmentof urban life and
humanhealth(Unger&Riley,2007).
Whereasonlytwocenturiesago,urbanareaswerehotbedsofinfectiousdiseaseandhighmortality,
todayscitiesareonaveragehealthierthanruralareas(UN-HABITAT,2006/7).Themodernizationof
citiesbringsincreasedaffluence,widespreadavailabilityofgoodsandservices,highqualityeducationandinfrastructure,andgreateremploymentopportunitiesforbothwomenandmen( Ibid).Inareport
onurbanhealth ineightIndiancities, theInternationalInstituteofPopulationSciencesnotesthat,It
was often assumed that the heavy concentration of health facilities and personnel in urban areas,
particularlyintheprivatesector,wouldautomaticallytakecareoftheincreasingurbanpopulationand
its health needs (IIPS, 2009). However, because the majority of urban growth has occurred in
developingcountriesthatarealreadystrugglingtomeetthebasicneedsoftheirpopulations,thereare
notenoughphysicians,drugsandfacilitiestomeetthedemandofthelargenumberofpatients.
7/27/2019 Banerjee Thesis
7/38
7
Becauseslumsareoftencentersof highmobilityandmigration,a continuous influxofnewpoolsof
infectious agents combined with high population density puts slum dwellers at high risk of
contracting communicable diseases (Agarwal, Sangar & Ghei, 2005). Numerous social determinants
cause poorhealthin slums aswell, including hiddenand hard-to-reach slumpockets, aweaksocial
fabric,socialexclusion,classdiscriminationtowardslumdwellers,alackofcoordinationamongvarious
stakeholdersandapaucityofpoliticalwillandconsciousnesstoaddressthehealthneedsofIndiasmost
marginalized(Agarwal,etal.2005).
UngerandRileypointoutthathealthservicesavailableinslumsareoftencomprisedofaninconsistent
patchwork ofpublic,private, andcharity-based providers. Inadequateor inappropriate care at these
places permits the progression of preventable diseases (Unger & Riley, 2007). Given barriers to
accessinghealthservicessuchascostandproximity,manyslumdwellersoftenwaituntilhealthcareis
absolutelynecessarybeforeseekinghelp.Riley,etal,addthat,theformalhealthsectorencounters
slumresidentsonlywhentheydeveloplate-stagecomplicationsofpreventablechronicdiseases(Riley
etal,2007).
Accordingto theInternational Instituteof PopulationSciences(IIPS),notallslumsarecreatedequal,and striking variability exists across these growing communities (IIPS, 2009). The five shelter
deprivationsUN-HABITATusesasindicatorsofslumstatussanitation(properdisposalofhousehold
andindustrialwaste,andsewagesystems),accesstowater,adequatelivingspace,durablehousingand
thesecurityoftenurediffergreatlywithinandbetweenslumcommunities(UN-HABITAT2006/7).
Oneofthegreatestdifficultiesinimprovingslumhealthistoprovideservicesinareaswhereresidents
are illegally occupying private land. According toUN-HABITAT,up to one-third of theworldsurban
population is constantly threatenedby forced eviction and insecurityof tenure.Many of the largest
slums intheworld,suchasKibera inKenya,KhayelitshainSouthAfrica,OrangiTowninPakistan,and
Dharavi in India, began as illegalsettlements (Peopleof Kibera,2010, Slum RehabilitationAuthority,
2007).Land tenureand landsecurity issues cancreate insurmountable bureaucratic obstacles to theprovision of basic rights,amenities andhealth services, since communities are frequently bulldozed.
Governmentsfearlegitimizing theillegaloccupationof privatelandbyprovidingbasic services.The
uncertainty of land rights significantly deters investments in housing, sanitation, infrastructure and
healthservicesinslumcommunities.
India:GrowingCitiesandUrbanSlums
Indias rapid industrialization, recurring famines,droughts, andcrop failures havespurred large-scale
migration to urban areas, where the burgeoning middle class and affluent elites provide prolificemployment opportunities in Indiasmassive serviceandmanufacturing industries. Figure2 charts a
steadyincreaseinthesizeofIndiasurbanpopulationoverthepastcentury.Infact,2009wasthefirst
year that Indiasmanufacturingoutputovertookagricultural output (Economist,2009).Areas ofhigh
population density are also those where jobs tend to cluster, drawing hundreds of thousands of
migrantseachyeartourbanhubs.Yeta2010studyonthedeterminantsofhealthinurbanslumsargues
that,[t]hecontinuedinfluxofindividualstoacommunityseverelylimitedbytenuredisputessuggests
thatmanymigrantsencouragedbythepromiseandprofitofcitiesareactuallyencounteringapotential
7/27/2019 Banerjee Thesis
8/38
8
povertytrap(PUKAR&HSPH,2010).
Upon arrival, many destitute rural migrants
findthattheycannotaffordtheastronomical
housingcostsofmegacitiessuchasMumbai,
and resort to taking up residence in
unregistered urban slums (Sadiq, 2008).
Thereisasevereshortageoflegal,affordable
housing for those at the lowest end of the
wealth spectrum in the largest cities (Ibid).
What little affordable housing is available
tends to be located outside of city centers,
suchthatthecostoflong,arduouscommutesisintroduced(upwardsofthreehoursdaily).
Thelackofurbanplanningleavesaffordable
areassuchasKBisolatedwithlimitedaccess
totransportation,schoolsandservices.Thus
in Mumbai, the second largest city in the
worldwithapopulationofalmost14million
residents, 56%ofMumbaikars1live in slums
(IIPS,2009).OfMumbaispoor(thoseinthe
lowest wealth quartile), almost 80% live in
slums,whichcoveronly6%ofthecitysland
(IIPS, 2009). Figure 3 outlines Indias slum
conditions.
The International Institute of Population
Sciences(IIPS),thebodythatconductsIndias
National Family and Health Survey (NFHS),
defines slums according to characteristics
1ResidentsofMumbai
Figure2:UrbanGrowth
(IIPS,2009)
Figure3.
IndiasUrbanSlumConditions:
APortraitinNumbers
Approximately40%ofIndias1.2billionpeople
liveinurbanareas(IIPS,2009)
UrbanpoorconstitutenearlyathirdofIndias
entireurbanpopulationwithnumbersthatare
growingatthreetimestherateofnationalpopulationgrowth(Agarwaletal.,2005)
Indiaishometo63%ofallslumdwellersin
SouthAsia,amountingto169millionpeople,
or17%oftheworldsslumdwellers(UN-
HABITAT,2006/7)
1in10childrenbornamongtheurbanpooris
notlikelytoreachhisorherfifthbirthday
(Agarwal,etal.2005)
UrbanareasinIndiahavelessthan4%of
governmentprimaryhealthcarefacilities( Ibid)
Thereis,onaverage,onlyonehealthfacility
forevery150,000peopleinIndiaslargest
7/27/2019 Banerjee Thesis
9/38
9
such as dilapidated and infirm housing structures, poor ventilation, acute overcrowding, faulty
alignment of streets, inadequate lighting, paucityof safe drinking water,water-logging during rains,
absenceoftoiletfacilities,andnon-availabilityofbasicphysicalandsocialservices(IIPS,2009).
TheofficialIndianCensusdefinesslumsasacompactareawithapopulationofatleast300orabout
60-70 households of poorly built, congested tenements, in an unhygienic environment usually with
inadequate infrastructure and lacking in proper sanitary and drinking water facilities (IIPS, 2009).
However,thelegaldefinitionofslumsdiffersfromstatetostate.Bothdefinitionsarevague,makingit
difficulttoenumerateslumpopulationsorcomparehealthoutcomesbetweenslumsandwiththeurban
non-slumpoor. The distinction between urban poor and urban slum dwellers (as defined below) is
important because there are large disparities in health and living conditions between the two
populations.
Itisoftenassumedthatslumdwellerslivebelowthepovertyline.However,whileslumsandpoverty
arecloselyrelatedandmutuallyreinforcing,therelationshipisnotalwaysdirectorsimple(UNHuman
SettlementsProgram,2003). IncitiessuchasDelhiandMumbai, someslumdwellerspresent income
levelsthatputthemwellabovethepovertyline(Singh,2001).Nevertheless,slumshavethehighest
concentrationsofpoorpeopleandtheworstlivingconditionsinurbanareas(IIPS,2009).
Aslumis consideredtobegovernment-recognized,legal,andpermanentwhenitisnotifiedtobea
registeredslumbyacompetentadministrativeauthority(IIPS,2009).Designatingparticularslumsas
notified allows for government investment in civic and municipal services such as water, trash
removal,andproperlypipedwater.However,unregisteredslumsinhabitedbymigrantworkers,new
ruralmigrants, or long-term slum residents who find it difficult to find formal housing within their
earningcapacity,arenotenumeratedwithincensus figuresandhaveextremely lowreachof services
andcivicfacilities(IIPS,2009).Forthepurposesofthisthesis,notifiedslumswillrefertoregistered,
permanent slums that receive government services such as pipedwater, sanitation and anganwadi
centers.2Unregistered slums will refer to non-tenured, informal, unauthorized colonies occupying
privateor otherwise non-residential land, resettlement colonies, squatter settlements, etc. Exhibit 2providesadiagrammaticillustrationofthenotificationprocessandtheimpactoftenureonhealth.
AforthcomingmanuscriptbyPUKARandHSPHonthedeterminantsofurbanhealthinKBfoundthat,
[i]nIndia,tenure(i.e.legallandownership)stronglyshapesthecontoursofthe slumgradient,which
oftenbeginswithsquattercommunitiesandendswithformallegalization(PUKAR,2010).Slumsthat
haveexisted formanyyears often become notified slums, (i.e. they become permanent, officially
registered by government authorities). However, a constant influx ofmigrants from rural areas has
increasedthenumberofillegal,unregisteredsettlements.Infact,NFHS-3foundthatapproximately
one-quarterofIndiancitydwellersdidnotfeelsecurefromeviction(NFHS-3,2005/6).
Given that residents of untenured slums are ineligible to receive municipal services from the
government, theymustinsteadseek outalternative sources. Toobtain these basic necessities, slum
dwellerspayastronomicalpricesandoftenreceivepoorquality,intermittentservices.Becauseof the
constantthreatofforcedeviction,householdinvestmentremainslow.Thisresultsinapoorstandard
of living, which can reinforce and ingrain the cycle of poverty (PUKAR & HSPH, 2010). If these
2Anganwadicentersarecommunity-basedfunctionariesofthe
IntegratedChildDevelopmentScheme(ICDS)a
GovernmentofIndiaprogramstartedin1975,toprovidechildeducationhealthandnutritionservices(Gheietal,
2010).
7/27/2019 Banerjee Thesis
10/38
10
communitieswereabletoobtainlandtenureanditsaccompanyingservices,theirdailycostsof living
and their disease burden would fall, and household investment would increase, accelerating their
escapefromthepovertycycle( Ibid).
HSPHandPUKARarguethat[p]olicyinterventionsthatmovesettlementstothelegalendoftheslum
spectrum can disrupt the poverty cycle by assuaging land security fears and ushering inmunicipal
servicessuchas pipedwaterandtoilets.Together,landsecurityandimprovementsinbasicamenities
canpromotehouseholdinvestment,aspeopleanticipateandplanformorepermanentstaysintheir
householdsandcommunities.(PUKAR&HSPH,2010).
InIndiasrecentpast,destructiveslum-clearingpolicieshavehastenedthepushforresidentsofillegal
settlementstoachievelegalstatusandreceiveservices.Forexample,initsefforts tomakeMumbaia
world class city theMaharashtra state governmentenacteda slum-clearing campaign in2004 and
2005thattoredownmorethan90,000shanties.(UN-HABITAT,2006/7).However,urbanslumgrowth
rates suggest that the number of unregistered settlements will only continue to grow,making this
approachuntenableandunsustainable.ThegovernmentofIndia(GOI)oftenresettlesslumdwellersto
tenements on the fringes of cities, destroying close-knit slum communities and dividing large, joint
family residences into nuclear units residing in separate flats. Unsurprisingly, it is very common forresettled slum dwellers to sell their new property and return to the slums (Mehta, 2004). The civil
societyorganizationAshainterviewedresettledslumdwellersin2006,whoreportedthatontheone
handtheywerehappytohavesecuretenureandaccesstobasicservicesbutontheotherhandthe
resettlementlocationofferedfeweremploymentopportunitiesandledtoincreasedtravelcosts.(Asha,
2006).
II.ChildHealth&ImmunizationCoverageinIndia
Highinfantandchildmortalityandmorbidityarerevealingindicatorsofthepervasive,negativehealthoutcomes ofovercrowding, poorenvironmentalconditions,sub-standardhousingconditions,and the
lackofaccesstoquality,affordablehealthcare.Childhealthandimmunizationstatusinparticular
areausefullensthroughwhichtoassessoverallslumpopulationhealth.
Childrenareatparticularriskofcontractingdebilitatingandoftenfatalcommunicablediseases.In
2008aloneinIndia,outof27millionlivebirths,therewereover1.5millionneonataldeaths(WHO,
2008).AccordingtoNFHS-3,Indiasurbanunder-fivemortalityrate(U5MR)was74deathsper1,000live
births,comparedtotheaverageof48amongIndiastotalpopulation(IIPS,2009).InMumbai,theinfant
mortalityrate(IMR)variesfrom30-43deathsper1,000,whichmeansthat3-4%ofchildrendiebefore
reachingtheirfirstbirthday(IIPS,2009).Theneonatalmortality rate (NMR),IMRandU5MRaremuch
higher among the urban poor than the urban average, and are equally as poor as their ruralcounterparts(Agarwaletal.,2005).
Table1:ChildandInfantMortalityinMumbaiSlums
(IIPS,2009)
(Deathsper1,000livebirths)
InfantMortality 24.9
Under5Mortality 32.7
7/27/2019 Banerjee Thesis
11/38
11
Measles inparticular isamajor killerof slum-dwelling Indianchildren dueto prolongedexposure to
infected siblings in very small living spaces with several family members dwelling in a single-room
household (Banerjee & Shitole, 2009/10). According to the NFHS-3, more than one in five poor
householdshasatleastsevenpersonssleepingperroom(IIPS,2009).
TheWHO/UNICEF Expanded Programme on Immunization (EPI, provided by the Indian government)
includes one dose of BCG vaccine to prevent tuberculosis, three doses of DPT vaccine to prevent
diphtheria,pertussis,andtetanus, threedoses oforal polio vaccine (OPV),andonedose ofmeasles
vaccine(WHOandUNICEF,2005).Achildshouldbefullyimmunized(i.e.shouldhavereceivedalltheEPI
vaccines)byageone.
The Indian government provides free immunization services to all children in government health
facilities.However, ratesamongsturbanpoorchildren remainlowerthan theurbanaverage:roughly
69%ofurbanchildreninIndiaages12-23monthsarefullyimmunized(IIPS,2009)comparedtoonly
44%ofurbanpoorchildren(Banerjee,A.&Duflo,2008).Thisnumberlikelyoverlooksthechildrenliving
inunregisteredslums,whoseimmunizationcoverageappearstobemuchlowerbasedondatafromKB.
Thefirstimmunization(BCG)isgivenatbirth,andbyinferenceitseemsmorelikelythatahospital-born
babywill be immunized forBCGthan a home-delivered child.Among theurbanpoor, almost three-
quartersofbabiesaredeliveredathome(Agarwaletal,2005),asignificantcauseoflowerimmunization
ratesinthispopulation,aswillbedemonstratedbyprimarydataonKBinsectionIV.
In2000,IndiasNationalPopulationPolicypledgeduniversalimmunizationcoveragebytheyear2010.
However,accordingtoSiddharthAgarwal,aleadingexpertonchildhealthinIndianslums,[a]lthough
coverage hasincreasedsubstantially inrecentyears,largenumbersofslum-dwellingchildren remain
incompletely immunized. Also, dropout rates remain problematic for immunization programs that
require multiple injections before the vaccine schedule can be completed. As of 1993, the Indian
MinistryofHealthandFamilyWelfareestimatedthatdropoutratesforDPTandOPVwereashighas70% of participating children (Pande & Yazbeck, 2003). This is important given that the failure to
complete vaccine regimens can leave childrenat risk and pave the way for resurgent epidemics of
infectiousdisease.
The impact of the failure to achieve universal coverage on childhood mortality and morbidity from
vaccine-preventableillness(VPI)isdifficulttoestimate.Since71.8%ofinfantandchilddeathsoccurat
home, causesofdeathcannotbe effectivelydetermined (Awasthietal.,2003).Additionally,in2003,
94.5%ofchildhospitaldeathsdidnothaveadeathcertificateandphysiciansoftentimesfailtodiagnose
their patients conditions (Ibid). Physicians frequently prescribe medications without providing a
diagnosis,oranyaccompanyinghealthinformationprevention,treatmentorcare(Banerjee&Shitole,
2009/10).
Based on this unreliable data, Indias Integrated Children Development Scheme (ICDS) system
underreporteddeathsin2003byone-third(Awasthietal.,2003).Itisalsodifficulttofindreliabledata
oninfantandneonatalmortalityrates, sincedataarecollectedstatewideandareonlydisaggregated
intourbanvs.ruralcategories,excludinginter-populationheterogeneityandsignificantdisparitiesbased
onsocioeconomicstatus,caste,ethnicbackgroundandplaceofresidence(Vaid,etal.2007).
7/27/2019 Banerjee Thesis
12/38
12
II.ShortcomingsofUrbanHealthDeliverySystems
Indiahasfallendismallyshortofits2001goalofuniversalimmunizationcoverageby2010.Thereforeit
is necessary to identify specific barriers to scaling up immunization coverage in urban slum
environmentsinordertodesignandimproveviablehealthinterventionsthatwillmakeuniversalaccess
a realisticpossibility.Challenges foruptakefall into twomaincategories:barriers toaccessinghealthservices by slum-dwellers, and a lack acceptability of immunization services due to the absence of
adequatehealtheducationefforts.
BarrierstoAccess
Starkdifferencesexistbetweenthemorbidity andmortality ratesofdisenfranchised Indiansliving in
registeredslums(suchasJogeshwariandDharavi)versusthoselivinginunregisteredslumswherethere
islittletonoprovisionofhealthservices,water,orsanitation.Manyoftheseunregisteredslumsdonot
fall on the radarofgovernment agenciesor NGOs (someonpurpose, toavoid the threatof forced
eviction).Thelackofadequatecityplanningandmappingleadstoaconcentrationofhealthservicesinasmall number of areas, and oftentimes duplication of services leads to repeated interventions in
registeredcommunities,whileunregisteredslumshaveatotallackofaccess(Agarwal,etal.2005).
Investments inhuman resources for public health have not increased in response tourbangrowth.
Giventheimmensevolumeofpatientsandthelimitedresourcesofgovernmenthospitals,healthcare
providersareoverwhelmed. InlargeIndiancities suchasMumbaiandDelhi,governmenthealthcare
workers (for example, staff atanganwadi centers, community-based institutions that provide basic
healthservices)facenumerouschallengesthatdecreasemoraleandharmrecruitmenteffortsfornew
workers.Payislowandopportunitiesforprofessionaldevelopmentorpromotionarescarce.Ageneral
lack of effective management in many institutions results in ineffective outreach efforts and poor
programoutcomes.LongandarduouscommutesonIndiancitiespoorpublictransportcantake3hoursormoreoutofahealthworkersday(Agarwal,etal.2005).
Beyondtheproblemoftoofewtrainedhealthworkersandfacilities,othersignificantbarriersprevent
parents (most oftenmothers) from bringing their children tohealth clinics for treatment of illness,
immunization,orroutinecheck-ups.Clinichoursoftencoincidewithworkinghours,havelengthywait
times,oraresimplytoofaraway andunaffordable. Inurban slums, thepreferredhealthprovideris
mostoftentheclosestone;parentsoftentakechildrentoNGOhealthproviders(Awasthi etal.,2003),
local pharmacists and chemists, bhagats (traditional healers) and local doctors who may not be
qualified practitioners (Banerjee & Shitole, 2009/10). For treatment or immunizations that require
multiplevisits,mothersfinditdifficulttomakerepeatvisits,ormakethiseffortonlyformalechildren
(Gaudin&Yazbeck,2006).3Injoint-familysituations,parentsandin-lawscanexhibitpowerfulresistance
when it comes to a young mother spending time and household income on doctor visits and
immunizations(Ibid).
3ThiswasnotthecaseinKB.Mothersshowedapreferenceforfemalechildrenandhealthoutcomesforboth
genderswereverysimilar.
7/27/2019 Banerjee Thesis
13/38
13
AcceptabilityofHealthInterventionsandHealthEducation
Oftentimes specific urban slum-dwellingcommunities hold strong traditional and/or religious beliefs
againstimmunization,ordonotunderstandthebenefitsofimmunization.Forexampleapproximately
halfof KBspopulation consists ofmigrantsfromBiharandUttarPradesh (U.P.) (Banerjee&Shitole,
2009/10).Inthesetwoareasthereis apervasiveurbanmyththattheoralpoliovaccine(OPV)causessterilityinmen.GivenIndiasnefarioushistoryofSanjayGandhisforcedsterilizationofpoorpeopleand
ethnic/religious minorities (predominately Muslims) in the 1970s, such fears are not historically
baseless and gave rise to a distrust of the medical establishment by Indias poor (Gwatkin, 1979).
However,therefusaltotakeOPVhasadangerouseffect:Indiaisoneofonlyfourcountrieswherepolio
hasnotbeeneradicatedintheworld(WHO,2006).
Whilepatientscanseeandappreciatethedirectbenefitofcurativehealthservicestothemselvesandto
their children, the long-term, preventive benefits of immunization to both the child and to the
communityarelesstangible.Mothersoftendonotunderstandwhytheyshouldputupwithacranky
childwithasorearmforaweekwhens/hewashealthytobeginwith(Banerjee&Shitole,2009/10).
Additionally,serviceproviderswhohaveanegativeanddiscriminatoryattitudetowardslumdwellers,or
whocauseinjuryorinfectionthroughpoorvaccinationprocedurescaninhibitwomenfromcompleting
theirchildrensseriesandbringingtheirsiblingsatall(Agarwal etal.,2005).
Agarwalalsopointsout that evidence suggests that poor uptakeof immunizationin urban areas is
associated withmothers unawareness about repeat visits to achieve complete immunization rather
thanoverallvaccineawareness(Agarwal,etal.2005).Asaresult,halfofurbanpoorchildrenwhobegin
theirimmunizationserieseventuallydropout(Ibid).
For womenwho had their childrenin the village before migration tourban areas, unreliable health
servicesareamajordeterrentforaccessingimmunizationservices.AbhijitBanerjeeandEstherDufloof
MITs Jameel Poverty Action Lab conducted a randomized controlled trial on child immunization to
evaluate theimpactofincentiveson increasinguptake.Throughweeklyvisitsof healthfacilities,theyfoundthat45%ofthestaffin chargeof immunizationswereabsentonanygivenworkday,andthere
wasnopredictablepatterntotheirabsence(Banerjee,A&Duflo,2008).Banerjeesays,[g]iventhata
fullimmunizationcourserequiresatleastfivevisitstoapublichealthfacility,theunreliabilityofthe
[healthserviceproviders]maydeterfamiliesfromtakingtheirchildren tothesub-center tocomplete
thefullimmunizationschedule.(Banerjee,A&Duflo,2008).
BecauseantibioticsareoftengivenasinjectionsinIndiaratherthanpills,itiscommonformarginalized
women to failto differentiatebetweeninjectionstotreat illness ratherthan vaccinationsto prevent
illness.Womenwhoparticipatedinfocusgroupsreportedthatdoctorsinthetwoclosestgovernment
hospitals openly discriminate against poor patientswhom they assume are uneducated and will be
non-compliant.Women said they were shouted at and blamed for childrens illnesses or infectedinjectionwounds.Theyweretoldtoboiltheirwater,washtheirhands,andtonoturinateordefecatein
commonoropenspaces.GivenKBandotherslumsconstraints,suchpublichealthmessagesarelargely
useless(Banerjee&Shitole,2009/10).
Othersreportedbeingturnedawayfromimmunizationservicesbecausetheydidnothavetheirchilds
birthregistrationdocument(whichisagainstgovernmenthospitalpolicy(Jaju,2009).Thediscriminatory
attitudeofhealthproviderswasidentifiedasamajordisincentivetoavailingofhealthservicesamongst
7/27/2019 Banerjee Thesis
14/38
14
women infocusgroupsonbarriers toaccessinghealthservicesandchallengestotheacceptabilityof
immunization(Banerjee&Shitole,2009/10).
IV.TheCaseofKaulaBandar
Toprovideacomprehensiveassessmentofthedeterminantsofchildhealthandimmunizationstatusin
bothunregisteredslums(usingtheexampleofKB)andregisteredslums,thisthesiswilldrawuponthree
datasources:
1) TheprimarydatasourceisahouseholdsurveyonchildhealthandimmunizationstatusinKB4,(asub-studyoftheHSPHandPUKARstudyonthedeterminantsofurbanhealth),
2) A2008/9householdsurvey 5todemonstratehowthelackoflandtenureandbasicutilitiesimpactsthehealthandwealthoftheKBcommunity,
3) The2001Censusdataonurban,notifiedslums(supplementedbyNFHS-32005/6findingson8Indiancities
6
)
Methods
TheHSPHandPUKARstudyassessedfivedomains:
Childhealth&immunization, Maternalhealth, Water&sanitation, Healthnetworks, Smallenterprise.
This sub-study on child health and immunization was designed and administered amongst 226
householdsinKB,interviewingthemothersof258childrenunderfiveyearsold.Thewomenchosento
participateinthehouseholdsurveyreflectthebroadspectrumofcommunitieslivinginKBthatdefine
themselvesaccordingtoreligion,language/ethnicgroupand/orincomelevel.Thesurveywasconducted
inHindi/ Urdu, Tamil and Marathi, and obtained informed consent with signatures or thumbprints.
Findingsweresupplementedwith17 filmedfocusgroupsandone-on-one interviews.Forchildhealth
histories,therecallperiodforquestionsaboutseriousillnessesandfrequentlyoccurringsymptomswas
oneyear,toaccountforthesmallsamplesize.
4Studydesigned,translatedandadministeredbyJoyaBanerjeeandTejalShitole.DataanalysisbyJoyaBanerjee5ByPUKAR,2008/9,supportedbytheFordFoundation
6NFHS-3teamsupervisorsdecidedtoincludebothnotifiedandunregisteredslums,basedontheCensusdefinition
ofslumsprovidedonpage7.However,theNFHS-3doesnotprovideinformationonwhatpercentageoftheslums
wereunregistered,andthereforeitisnotpossibletodrawaccurateconclusionsontheimpactoflandtenureon
healthstatusfromthesedata.NFHS-3usedverysimilarmethodologytotheCensus,choosingrepresentative
samplesof2,000from8Indiancities(1,000slumand1,000non-slumindividualseachrepresentingahousehold).
Stateweightscorrectedforoversampling,sothatindicatorsbasedonthesedatawererepresentativeatthecity
level,aswellasforslumandnon-slumareaswithincities.
7/27/2019 Banerjee Thesis
15/38
15
Surveyparticipantswereselectedusingrepresentativesamplingbasedonthegeographicallocationof
their household in the slum, which is divided into distinct ethnic communities, self-defined by
inhabitantsaccordingtolanguage,religion,homelandandothercharacteristicspeculiartothecultural
groupsinIndia.Forexample,MuslimfamiliesfromBiharoccupiedonemainsectionoftheslum,while
TamilboatrepairmenwhosefamilieshadbeeninKBfordecadesoccupiedanother.Theslumsitsupona
wharf and has one main road, with narrow, perpendicular lanes that denote separate communities
(Figure6).Participantswerechosenfromeveryotherlane,andbasedonwhichmotherswereavailable
forinterviewsondaysthesurveywasbeingconducted(includingnon-workweekdaysandtimes).
Thestudysobjectiveswere:
Toascertainthedeterminantsofchildmorbidityandmortality, Toassessthehealthknowledgeandhealth-seekingbehaviorofKBmothers, Tocollectdemographicinformationoneachhousehold(focusingprimarilyonthemother), Toobtainhistoriesofthehealthofchildrenunderfive.
Methodsforthequantitativecomponentofthestudyareonpage23.
Results:KaulaBandarEthnographyandDemographics
KBisan unregisteredslumsettlementontheeasternwaterfrontofMumbai,hometo18,000-25,000
residents (PUKAR,2008/9). TheKBcommunity ismadeup primarily of Tamil boat-repairing families,
someofwhomhavebeenthereforover40years,andmigrantlaborers(whoaremuchmorerecent
arrivals, mostly from Bihar andU.P.) (Banerjee& Shitole, 2009/10).Seventypercentof KB residents
reportthattheyinitiallycametotheareainsearchofemployment(PUKAR,2008/9).Thewharfthat
theyliveon(picturedinFigure6)isownedbytheBombayPortTrust(BPT).TheBPThasbeentryingto
forciblyremoveresidentsforfourdecades.
Figures4-6:SituatingKaulaBandar(GoogleMaps,2010)
7/27/2019 Banerjee Thesis
16/38
16
BecauseKBresidentsareoccupyingBPTslandillegally,theyhavenoaccesstocleanwater,sanitation,
electricityorbasichealthservicesguaranteedtoall Mumbaikarsthatownorrenttheirhomeslegally.
Thousandsoftinyone-roomhomessitperchedoneatopanotheramidstlanessonarrowtheyoften
requireinhabitantstowalksinglefile.KBhasanextremelyhighpopulationdensity (Table2),a factor
thatgreatlyincreasestheriskofcontractingcommunicabledisease.Consequently,tuberculosis,measles
andothercommunicablediseasesareverycommonamongstbothchildrenandadults.
Table2.Populationdensitycomparison
(PUKAR,2008,WorldGazetteer,2008,USCensusBureau,2008)
Population Landarea(km2) Populationdensity(peopleperkm
2)
KaulaBandar 18,000-25,000 0.08 230,769-320,512
Mumbai 13,830,884 603 22,937
NewYorkCity 8,363,710 789.4 10,452
TheuseofunwieldyandflammablematerialsinthebuildingofhomesputstheKBcommunityatgreat
riskoffiredamage.InFebruaryof2010,afiresweptthroughthecommunityanddestroyed251homes.
Thelackofpipedwaterandtheinaccessibilityofslumhomesmeantthatthefirebrigadetookhoursto
arriveandtodousetheflames.
To obtain water for bathing, drinking, cooking and cleaning, homes would tap into the Bombay
MunicipalCorporation(BMC,thelocalagencyresponsibleforprovidingbasicamenitiestoMumbaikars)
watersupplypipes.However,approximately20yearsago,theBMCbecameawareoftheillegalusageof
waterbythecommunity.Theysubsequentlydestroyedtheundergroundpipingbygoingtoeachhome
inKBanduprootingthepipingsystem(PUKAR&HSPH,2010).However,fivehouseholdsweremissed,
andcurrentlythesefivesupplywatertotheentireKBcommunitythroughanintricatesystemofrubbertubingat excessive costs. FortenuredMumbaikars,theBMCchargesamonthlyfeeofRs.30
7per
householdforaccesstoaproperwatertap.However,inKB,anaverageresidentpaysbetweenRs.150-
2998,a5-10folddifference(PUKAR&HSPH,2010).Thefirebrigadepipesareoldanddamaged,allowing
the surroundingseawater, garbage, fecalwasteand othertoxins to leak into the pipedwater.Only
59.6%ofhouseholdsusedanymethodofwaterpurificationsuchasboiling,waterpurificationtabletsor
drops,etc.
Only3%ofKBhouseholdshadaccesstoatoiletintheirhome,andthemajorityofsurveyrespondents
reported that there are less than five public toilets in the community, for between 18,000- 25,000
people(PUKAR,2009).Observeddatasuggestthatresidentsmakefrequentuseofthesurroundingsea
asa toiletandgarbage-dumpingsite.TherubbertubingsystemthatsupplieswatertoKBhomesrunsdirectlythroughthiscontaminatedseawater.Thelackofdrainageandsewagesystemsexacerbatethe
spreadofwater-borneillnessesandcommunicablediseases,especiallyduringmonsoonfloodingseason
(Banerjee&Shitole,2009/10).
7$0.66USD
8$3.31-$6.60USD
7/27/2019 Banerjee Thesis
17/38
17
Because the majority of KBs tightly packed
homes lackwindows orproper ventilation, and
Mumbai temperatures can reach over 40C
(104F)duringthecityshottestseasons,children
frequently swim in the surrounding,
contaminated seawater to cool off. They have
nowhere else to play except for a small,
abandoned boat yard full of rusty pipes.
Consequently, skin infections, parasites and
tetanus are common childhood causes of
morbidity.
KBDemographics9
ThemajorityoffamiliessurveyedwereMuslims,
an observation closely supported by the PUKAR
2008/9surveyaswell, whichfound that51% of
respondents were Muslim, 42.9% were Hindu,
3.8%Christian,and2.1%other(Figure7).
When asked to report which languages they
speak, some respondents distinguished Hindi
fromUrdu,butitwas laterdiscoveredthatsome
considered them to be the same language.10
Therefore Figure 8 masks the true number of
Urdu speakers who may have reported beingHindi speakers, since Hindi is one of the two
lingua francas of Mumbai (the other being
English).
Figure9showsthatthemajorityofrespondents
earned between Rs. 3,000- 5,000 per month11
andsubstantialpercentage,12%,earnaboveRs.
5,000 per month12. The per day equivalent
would be Rs. 100- 16713, signifying that the
average KB resident earns above the global
povertylineofRs 57 per day,or $1.25 per day(PPP). When women from these families were
9AllsubsequentfindingsinthissectionarefromthechildhealthstudybyBanerjee&Shitole,2009/10.
10UrduisconsideredtobeamorepureformofHindi,andisspokenpredominatelyinMuslimcommunities.The
twousedifferentalphabetsbutareessentiallythesamelanguage,barringsomedissimilarvocabulary.11$67.79-$112.86USDpermonth
12112.86USD
13$2.26-$3.76USD
Figure7
Figure9
Figure8
7/27/2019 Banerjee Thesis
18/38
18
askedwhytheydonotusetheirhigherearningstoseekoutimprovedlivingconditionsinotherlocales,
someparticipantsrespondedthattheydidnotwishtoleavetheirfamilyortight-knitcommunity.
TheaverageKBhomehas6.15residents(Table3) andisapproximately2.5mx2.5m(8x 8).Bathing
andcooking(overkerosenefires)areconductedinsidethehomeorimmediatelyoutside,inthelanes,
causingfrequentrespiratoryinfectionsinchildren.
Thegreaternumberof children perMuslimhouseholdcould beexplained by themuch higher total
fertilityrates(TFR)andlowerlevelsoffemaleeducation(whicharesignificantlycorrelatedwithfertility
rates)(Bloom&Canning,2006)inthestateswheremostMuslimfamiliesmigratedfrom.InBihar,the
TFRis4.3and74.6%ofurbanpoorwomenhavenoeducation.InU.P.,theTFRis 4.1birthsper1,000
womenand67.7%ofwomenhavenoeducation(UrbanHealthResourceCenter,2007).Overall,almost
halfofwomeninKBhavenoeducation(Figure10).
Onesurprisingfindingisthatonly10.9%ofwomenare
employed, even counting home-based production or
servicework. Therecentproliferation ofmedia about
another Mumbai slum, Dharavi14, has created the
commonimpressionthattherearethousandsofmicro-
industries booming in Indias slums. The surveyinstrument asked, Do you have a job for which you
earn money? and interviewers gave examples that
includedin-homeworksuchastailoring.Ofthe10.85%
who are employed, women reported working as
servants,tailors,cooks,andshopkeepers.
ChildHealthandMothersHealthSeekingBehavior
InKBonly29%ofchildrenarefullyvaccinated,comparedto69.8%ofchildreninMumbaiasawhole,and68.7%innotifiedslums(IIPS,2009).Freevaccinationisavailableatgovernmenthospitalsthatare
reachablebytrainortaxi.Thecostof transporttotheclosesthospitals(Rs.415foraone-way,second-
classtrainride,andapproximatelyRs.40-6016bytaxitoJJHospitalorCamaHospital,respectively)was
14WherethefilmSlumdogMillionairewasfilmed
15$0.09USD
16$0.89-$1.34USD
Table3:HouseholdComposition
Mean Standard
Deviation
#ofpeopleperhousehold 6.15 2.58
#ofchildrenperhousehold 2.91 1.53
#ofchildrenperHinduhousehold
2.86 1.37
#ofchildrenperMuslimhousehold 3.17 1.62
Figure10:
7/27/2019 Banerjee Thesis
19/38
19
consideredtobeadisincentive forsomemothersto seekingcareina healthcarefacility,certainly to
makingmultiplevisitstocompletetheentireimmunizationschedule.
Womenreported seeing local doctors inKB for most of their childrens healthcare needs, and only
taking them to hospitals (most often government hospitals) in serious conditions (such as excessive
vomiting and diarrhea, persistent fevers and coughs, etc.) Table 3 provides a breakdown of which
healthcarefacilitiesaremostcommonlyutilized.ItisunclearwhattrainingthedoctorsinKBhave,but
interviewswiththemrevealedthattheydonotofferimmunizations,butdotreatcommonchildhealth
issues.Womenreportedthatthemostcommonseriousillnessestheirchildrenhaveexperiencedare:
Pneumoniaandotherrespiratoryinfections, Tuberculosis, Malaria, Typhoid, Tetanus, Jaundice, Measles.
Itisimportanttoconsiderthatbothwomenandphysicians(inKBandatfourgovernmenthospitals)
reportedthatphysiciansdonotnormallydiscussdiagnoseswithwomen,becausetheyassumetheyare
uneducated and will not understand. They merely assess the childs symptoms and prescribe
medication, which women pick up from the medico or chemist (small pharmacies where
certificationstatusofpharmacistsisunknown).Themedicationcomesintheformofloosepillswrapped
in foil or plastic, with no identifying information. Side effects from interactions are consequently
common.
Womenjudgetheeffectivenessofaproviderbasedonthetimeittakesfortheirchildrentorecover.If
theyaredissatisfiedwiththelengthoftherecoveryperiod,theyseeanotherdoctorandgivetheirchild
adifferentmedication.Table4showsastrongpreferencefordoctorsinKB,andinfocusgroupswomen
cited the reason for the preference as low cost and proximity. Women also said that doctors in
governmenthospitalsseethemonlyforafewminutesanddonotconveyanypreventiveorcurative
healthinformation.Withregardtoimmunization,somephysiciansinstructedwomenthattheinjections
givenwereforthepreventionofillness,andthatitwasveryimportanttosavethechildsvaccination
card(butdidnotexplainwhyitwasimportant).Only29%ofwomenstillhavetheirchildsvaccination
Table4:MothersHealthSeekingBehavior
SourceofHealthcare %
DoctorinKB 44.2
DoctoroutsideKB 8.5
Privatehospital 7.8
Governmenthospital 16.7
Bhagat(traditionalhealer) 0.8
Treatathome 3.5
Notreatmentsought 17.4
Noresponse 1.2
7/27/2019 Banerjee Thesis
20/38
20
card(andthereforedataonimmunizationarelargelybasedonmothersrecallof
7/27/2019 Banerjee Thesis
21/38
21
Only29.1%ofchildrenwerefullyimmunized
in KB (Figure 12). Child immunization status
seemedtodependlargelyontheconvenience
factorformothers,meaningthatiftheservice
was provided, for free, without the mother
havingtotravelortaketimeoutofherday,
immunization rates would be high. For
example,88.8%ofchildrenweregivenBCG,a
vaccinegivenatbirth,atimewhenawoman
delivering in a health facility has access to
immunization services. Similarly, Oral Polio
Vaccine(OPV)ratesarethehighestoutofany
vaccination, at 95%. This is because the
government Pulse Polio door-to-door
campaign is highly effective and reaches
almost every household in KB, with workersgoing from lane-to-lane givingdrops toeach child, and
returningtothehomeifthechildwasnotpresentinitially.ComparetheseratestoMeasleswith31.78%
coverage,avaccinegivenbetween9-12monthsthatrequiresthemothertovisitahealthfacility.
SomeKBmothersreportedthatthelocalDockyardHospitalhasbeenorganizingimmunizationcampsin
KB.Therearemixedfeelingsaboutthecamps:manywomenreportedthattheinjectionsitesontheir
children became swollen and infected. Several developed cysts that required costly operations (at
approximately Rs. 1,00017, one-third of monthly income). This might suggest improper sanitizing
practicesorperhapsthere-useofneedles.However, infocusgroups,womenbyand largeseemedto
knowtheimportanceofusinganewneedleforeachpatient,andthatHIVinfectioncouldresultfrom
usingthesameneedleonmultiplepatients.Manywomenreportedwatchingtomakesurethedoctor
usedanewneedleduringimmunizationprocedures.
17$22.36USD
Figure12
7/27/2019 Banerjee Thesis
22/38
22
Table5.ComparisonofKBtoNotifiedSlumsandPoorestWealthQuartile18
Oneveryindicator inTable5exceptfordeliveriesassistedbyhealthpersonnel,KBfaredmuchworse
than both notified slums, and inmost casesworse than the poorest quartile aswell. The numberssuggestthatKBspopulationisunusuallydisenfranchisedcomparedtothoselivinginnotifiedslumsand
theurbanpoor,especiallywithregardtoeducation,tenure,infantmortalityrate,andtheincidenceof
diarrhea.
TheHealthImpactofTenure
Toassessthevalidityofthesefindings,andtocompareKBanunregisteredslumtonotifiedslums,
datafromKBarecomparedtoCensusdataonthesamehealthindicatorsinTable6.Itisimportantto
note thattheCensus iscarriedoutonlyonceeverydecade,and sothedata are from2001.Whereavailable,NFHS-3datafrom2005/6(usingsimilarmethodologyto theCensus) isusedtosupplement
Censusfindings,whichmaybeoutdated.Censusdataweredisaggregatedintocity,slum,non-slumand
18Someindicatorsarehighlightedbecausetherecallperiod(past1 year)maymakethesedataincomparableto
those fromtheCensus(2weeks).Thesurveyinstrumentaskedwomenaboutcommonillnesseswithinthepast
year;2weekswouldnotberepresentativeofKBgiventhesmallsamplesize.However,observationaldataandthe
qualityofKBsconditionssuggestthatthesedatamayverywellbeaccurate.
IndicatorKaulaBandar
(unregistered)
Census(notified
slums)
Census
(poorest
quartile)
NFHS-3(both)
%ofwomenemployed 10.9% 27.1% 27% 27.9%
%ofwomenwithnoeducation 45.4% 19.3% 46.2% 19.9%
%ofwomenwith
7/27/2019 Banerjee Thesis
23/38
23
poorest wealth quartile. For the purposes of accurate comparison, only slum and poorest wealth
quartileareincludedhere.
Table 6 revealsstriking discrepancies
in immunization coverage. These
couldbeduetothefactthatnotified
slums have health interventions
specificallyforchildrenandmothers,
such as the Integrated Child
Development Scheme (ICDS) which
operates through 40,000 anganwadi
centers nationwide, where
community health workers offer
immunization, child health services
andantenatalcare(IIPS,2009).Therearealsomanyurbanhealthcentersinorincloseproximityto
notifiedslums.Thereisaproliferationofslum-focusedNGOs,especiallyinDharavi,thatprovidehealth
informationandservices,andencourageimmunization.
TheoneinconsistentfindingisthatOPVincidenceinKBismuchhigherthaninnotifiedslums.Thiscould
beduetothegeometriclayoutofKB(Figures4-6)whichmakesiteasiertovisiteveryhomeinaslum,
comparedtootherslumssuchasDharaviwithserpentinelanesandanun-plannedlayoutforanever-
growingnumberofhomes.
QuantitativeFindings
Methods
A predictive logistic regression model was built from the retrospective cohort study data using
backwardselimination(p
7/27/2019 Banerjee Thesis
24/38
24
DurationofresidenceinKB(continuous), Birthplacecityhospital,villagehospital,andwith/withoutadai.21
1)Outcome:ImmunizationStatus
Table7:PredictionofImmunizationStatus
Parameter Estimate
LikelihoodRatio95%
ConfidenceLimits Pr>ChiSq
Intercept 0.6467 -1.5047 2.8183 0.5537
Language:Other -0.6222 -2.0043 0.8590 0.3842
Language:Urdu -1.9730 -3.5340 -0.4711 0.0102
Notemployed 2.2913 0.8493 3.7545 0.0016
Notbornincityhospital -2.2980 -3.6995 -1.0877 0.0005
#Peopleinthehouse 0.3341 0.0636 0.6526 0.0253
Table7showsthatsignificantpredictorsofimmunizationstatusarethecovariateslanguage,
employmentstatus,birthlocationandthenumberofpeopleinthehousehold .
Language:Urduspeakersare1.97timeslesslikelytogetimmunizedcomparedtoHindispeakers,holdingallothervariablesconstant.(Hindi=referencegroup)
22Thismaybebecause
mostUrduspeakershavemigratedfromresource-poorareassuchasU.P.andBiharwhichhave
notoriouslylowhealthandeducationoutcomesparticularlyforwomen,whoaremore
marginalizedintheseregionsthaninotherpartsofthecountry(UrbanHealthResourceCenter,
2005/6).23
Employment:Womenwhoarenotemployedare2.29timemorelikelytohaveimmunizedtheirchildrencomparedtothosewhoareemployed,holdingallothervariablesconstant.Infocus
groups,whenaskedtoexplainthisfinding,womenrespondedthatthosewhoareemployeddo
nothavetimetotaketheirchildrentothehospital,waitinlongqueues,andsacrificeadays
wages.
Birthlocation:Childrenborninavillagehospitalorathome(villageorurban)are2.30timeslesslikelytobeimmunizedthanchildrenborninacityhospital(governmentorprivate).Village
hospitalshaveinconsistentqualityofcare,unreliableprocurement,highratesofabsenteeism
amongpersonnel,andfrequentshortagesofsupplies.Basedonresponsesinfocusgroupswith
21Askilledbirthattendant.
22Languagegroupdatawasincludedtoexaminebehavioralpatternswithindistinctethnicgroups,whichare
demarcatedlargelybylanguageinIndia.However,thesurveyaskedparticipantswhichlanguagestheyspeak
(multipleoptionswerepossible),andnotwhattheirprimarylanguageis.ThusifaTamilresidentlearnedHindi
duringhisorhertimeinMumbai,itwouldappearasthoughtherearemoreHindispeakersthanthereactuallyare,
andthiscouldbiasthefindings.23SeediscussionoflanguagegroupsinprevioussectionKBDemographics
7/27/2019 Banerjee Thesis
25/38
25
KBmothers,womenwhodeliverathomearelesslikelytomakethevisittoahospitallaterto
immunizetheirchildren.
Numberofpeopleinthehouse:Perincreaseofonepersoninthehousehold,amotheris0.33timesmorelikelytogetherchildimmunized,holdingallothervariablesconstant.(6.15=
referencegroupwithastandarddeviationof2.58)Itisunclearwhythismightbe;perhapswith
eachsuccessivechildamotherismorelikelytolearnofimmunizationthroughdoctorvisitsorfromfamilyandneighbors.However,thiscontradictsthelaterfindingthatperincreaseofone
childinthehousehold,amotheris0.39timeslesslikelytohavecorrectknowledgeofthe
purposeandbenefitsofimmunization.
2)Outcome:Mortality
Table8:Predictionofchildmortalitystatus
Parameter Estimate
LikelihoodRatio
95%Confidence
Limits Pr>ChiSq
Intercept -1.7337 0.0473 3.4830 0.0447
Immunizationstatus
(dichotomous)
-3.2177 1.9683 4.5386 ChiSq
Intercept 0.0579 -1.4363 1.5249 0.9386
Mothersage 0.0583 -0.0053 0.1253 0.0786
Income:
7/27/2019 Banerjee Thesis
26/38
7/27/2019 Banerjee Thesis
27/38
27
StructuringanEffectiveIntervention
Given the unprecedented growth of Indias slums, it is imperative that the GOI devote increased
resources specifically toward the health of urban slum dwellers. Interventions should take into
consideration the specific challenges faced by urban slum dwellers without tenure, which may be
distinct from the more general category of urban poor. Unger and Riley argue that [a]ppropriateinterventionsandtreatmentsareonlyeffectiveonceprovidedinthecontextofaccessibleandutilized
healthcareservices(Unger&Riley,2007).
India is making progress by including slum dwellers and other marginalized urban dwellers in its
eleventhFiveYearPlan(2007-2012).TheplanincludestheintroductionoftheNationalUrbanHealth
Mission(NUHM)forinclusivegrowth,coveringallcitieswithapopulationofmorethan100,000(IIPS,
2009).Itmustnowassesstherelevanceandimpactofcurrenthealthprogramsaffectingslumdwellers,
andcraftnew interventions thatbettermeetthepopulationsuniqueneeds.Earlierplans since2001
shifted away from slum-clearing policies, instead focusing onslumsasanintegral part of the urban
landscape. Thepolicies recognized slums as contributors to city economies throughboth their labor
market contributions and informal production activities (Asha, 2006). Insteadof treating slums as aproblemtobesolved,newpolicieswerebasedonthepremisethatlocalbodiesshouldaccommodate
slumsandrecognizetheircontributionsthroughprovidingaffordablehousingandservices.
Indiahasalreadyshowedthatitiscapableofimplementingalarge-scalehealthinterventioninaslum
populationwithnear-universalcoverage.IndiasPulsePolioCampaignofferingdoor-to-doororalpolio
vaccination,reached94.96%ofchildreninKB.Suchsuccesspointstotherealpossibilitythatuniversal
accesscanbewithinIndiasreachifadequateresourcesaremobilizednow.It isclearthatnewhealth
carefacilities,anddrugandvaccinestockpilesareneededtomeetthegrowingdemandofurbanslum
populations.AsthefindingsfromKBdemonstrated,theexpansionofprimaryhealthcareinfrastructure
into areas now overlooked will improve the immunization status of slum-dwelling children and, by
corollary,long-termoverallpopulationhealthandproductivity.(Agarwaletal.,2005).
Inorderto createacceptabilityofanddemandforimmunizationservices,slumdwellersmustbe
providedwithaclearunderstandingofthebenefitsofvaccines,theimportanceoffollowingthrough
withallscheduleddoses, informationonpotentialsideeffects, theunderstanding thatothersin their
communityaretakingadvantageofthefreeservice,and incentivestoovercomebarrierstoaccessing
health services. Interventionsmust be cost effective,geared toward changing behavior in thetarget
population,andmustbemeasurableandscalable.
TherearefivenecessarystepstoworktowardsuniversalvaccinecoverageinIndia:
1) Identify all slums (notified and unregistered) usingmappingand vulnerability assessment as a
planning tool. By understanding which areas are endemically weak and assessing adequacy the ofinfrastructure,itwillbecomeclearwherenewfacilitiesareneeded,andprimarylevelfacilitiescanbe
re-definedtoreachleft-outareas(Agarwal etal.,2005).
2) Strengthened and regular immunization services, particularly for areas with poor access.
Immunization strategiesmust focusonproviding the complete series and making sure that women
knowwhyitis crucialforthemtoreturn,ratherthansimplyimmunizingasmanychildrenaspossible
withoutregardtofollow-up.
7/27/2019 Banerjee Thesis
28/38
28
3) Because of the unique role women play in safeguarding the health of their children, slum-
dwelling women should be trained as community health and outreach workers. Community
participationiswellknowntoincreasebehaviorchangeandtheacceptanceofmanydifferenttypesof
healthinterventionsthroughthecreationoflocalownership.Itcanincreasetrustandawarenessofthe
benefits of immunization programs, and this knowledge is likely to spread toother households and
communities given the close-knit nature of slum communities. Community health workers should
emphasize theimportanceofretainingvaccinecards,proofof vaccinationanddateswhen theywere
given(evidencethatiscrucialtoevaluatingtheefficacyofanintervention).
The KB child health study found highly correlated patterns of health-seeking behavior, health
knowledge, and incidence of communicable disease based on which lane houses are situated in
(Banerjee&Shitole, 2009/10).ThusKBslane-basedbehaviormaybea highlyeffectivedistribution
channel for immunization health information. 56% of KB women already know the purpose of
immunizationandthesewomencanbemobilizedas apowerful resource inefforts toincreasehealth
education.
4) Simply offering immunization series is not enough. A comprehensive intervention requiresproximate health facilities, affordable services, adequate supplies, vaccine efficacy, correct
administration,postadministrationcounseling,followupwithleft-outchildrenanddrop-outs,andsafe
disposalofvaccinesyringes;butalsobymeasuringtheinterventionsimpactonslum-dwellingchildrens
morbidity andmortality fromvaccine-preventablediseases, througha systemof rigorousmonitoring
and evaluation. This will require improved disease surveillance, especially in areas of low reach.
(Agarwal,etal.2005)
Agarwalsuggestsholdingregularimmunization/outreachcampsthatarepublicizedwell,andoccurina
convenient,fixedlocationonaregularschedule(Agarwal,etal.2005).Campsmustobtainthesupport
oflocalstakeholderssuchasslumcorporators,religiousleaders,teachers,maleheadofhouseholdsand
motherinlawsinordertohavewidelocalbuy-inandtoensurethatallpotentialbarriersareaddressedina culturallyappropriate, relevantmanner.Slumschoolsandmadrassahs,communityworshipareas
suchastemplesandmasjids,anganwadicenters,andprivatedoctorsclinicsareconvenientandtrusted
locations tohostsuch camps.Agarwalalsosuggests that healthcareworkers candistributepictorial
cardsorleafletsindicatingtheappropriateagesfordifferentvaccinesandreinforcingthesignificanceof
timelyimmunizationwhenchildrenvisithealthservicesforconsultation.Thesewillencouragefamilies
togettheirchildrenimmunizedevenwhentheymigratetootherareas.
5) Foranyofthistobeeffective,theremustbea convergenceofstakeholdersinordertobuild
political momentumand tomanage resources effectively. The wide range of stakeholders who can
influence the success of an immunization intervention should not be underestimated. Stakeholders
couldincludegovernmentofficials,privateproviders,communityoutreachworkers,NGOs,faith-based
organizations,donoragencies,religiousleaders,teachers,andtheslum-dwellersthemselves.Traditional
birthattendants,dais,couldbeparticularlyhelpfulinidentifyingthehomeswheretheyhaverecently-
deliverednewborns(deQuadros,2008).
Toreducetheduplicationofservicesandtoensurecoordinationbetweenmultiplestakeholdersatthe
citylevel,allstakeholdergroupsshouldmeetregularlytoidentifyeachotherscomplementaryrolesand
capacities.Theyshouldworktogether tocreatea time-boundplanfor scaling immunizationservices
withmeasurableoutcomesandarigorousmonitoringandevaluationcomponent.
7/27/2019 Banerjee Thesis
29/38
29
Incentives:TheLastMile
BanerjeeandDufloconductedarandomizedcontrolledtrialinIndia(n=1640)inwhichtheyexamined
theefficacyofnon-cashincentives(insteadofcashgivingonekiloof daalorlentilsandasetofbowls)
onimmunizationrates.Onegroupoutofthreehamletsinvolvedinthetrailreceivedwellpublicized,
regularlyoccurringimmunizationcamps.Thesecondreceivedthecampsaswellasthesmall,non-cashincentive of daal, worth Rs. 40.
24The third group was the control and did not receive camps or
incentives(BanerjeeA&Duflo,2008).
Duflounderscoredtheimportanceofincentivesinaninterviewinwhichshesaid,evenwhenyoufully
fixthesupplyproblemandimmunizationservicesareavailable,youstillneedtonudgepeopleabit.You
needtoprovideanincentivefortheparenttobringthechildandyouneedtoprovideanincentivefor
the service provider to be there aswell (Duflo & Banerjee, 2008). Banerjee and Duflo tested the
hypothesisthatimprovingreliabilityofservicesimprovesimmunizationrates,andsmall,non-financial
incentiveshavelargepositiveimpactsontheuptakeofimmunizationservicesinresource-poorareas
(Ibid).Banerjeenotesthat:
IninterventionA,evenwhenaccesswasgoodandasocialworkerconstantlyreminded
parentsofthebenefitsofimmunization,morethan80percentdidnotgettheirchildren
fullyimmunized.Morethan75percentobtainedthefirstshotwithouttheincentive,but
thenstoppedattendingthecampsonlyafter2or3shots.Thisshowsthattheparentsdo
nothave strongobjections to immunization, but that they arenotpersuaded enough
about itsbenefits to overcome thenatural tendency to delay a slightly costly activity
(Ibid).
Anumberofpreviousstudieshaveshownthatuptakeofpreventivebehaviorsisverysensitivetosmall
incentives or small costs, suggesting that incentives can play a role in promoting preventive health
services(Kremer,2007;Cohen,2007;Thornton,2005).However,otherresearchershavesuggestedthat
inresourcepoorsettings,ensuringareliablesupplyofhealthservicesandeducatingparentsaboutthebenefitsofpreventivecarearemoreimportantthanprovidingincentives.(Morris,etal.,2004)
BanerjeeandDuflosfindingssupportthecaseforprovidingincentives:Whilecontrolhamletshadafull
immunization rate of 6.2%, hamlets inwhicha reliablecampwas held showed ratesof16.6%, and
addingtheincentivepushedratesto38.3%,astatisticallysignificantincrease.Moreover,whileacamp
without incentives increased immunization ratesonlyin thehamletwhere it tookplace,campswith
incentivesalsoincreasedratesinneighboringvillages(BanerjeeA&Duflo,2008).
SimilarincentiveschemeshavebeenhighlysuccessfulinMexico(formerMinisterofHealthJulioFrenks
conditionalcashtransferprogram,Oportunidades,andNewYorkMayorBloombergsOpportunityNYC
program.
AparticipantinoneofKBsfocusgroupsmentionedthatinherhomevillageinU.P.,thereisanincentive
schemethatprovidesconditionalcashtransferstofamiliesthatbringtheirchildreninforimmunizations
andwhocompletethevaccineschedule.Shesaid,Theygiveyoumoney,sonoweveryonegoes.A
cash(ornon-cash)transfercouldbeapowerfulincentiveforKBwomentoimmunizetheirchildren.In
24$0.88USD
7/27/2019 Banerjee Thesis
30/38
30
fact,womenwerepartiallyincentivizedtoparticipateinthefocusgroupdiscussionsbyreceivingakilo
ofriceeach.
V.Conclusion
The link between immunization and child mortality is well known. However, prior studies on
immunizationinIndiadidnotestablishhowchildmortalityratesandimmunizationcoveragediffereven
within thesame social,medicalandhealth care systems.Access tocarewithin thesamemetropolis
variesgreatlydependingonmanyfactorsthatdeterminesocietalinequityincludingincome,migration
status,ethic/religiousgroup,geographiclocation,etc.Indiasdiseasesurveillancesystemmusttakeinto
accountthesedifferencesbyconsistentlydisaggregatingitsdatainfuturecensuses.
Withincreasedinvestigation into utilizationof immunizationservices,qualityof care, followupvisits
andimpactonchildmorbidityandmortality,itwillbefurtherestablishedthatunregisteredslumshave
farlessaccesstoservicesthannotifiedslumsreceivingoutsideinterventions.Equityinhealthaccess
requiresurgentattentiontotheimprovementofbothaccessandqualityofcare.
Focused,cost-effectiveandmeasurableimmunizationinterventionsthattakeintoaccountthecontext,
needsanddesiresofthetargetcommunitywilldramaticallyimprovetheimmunizationhealthstatusof
slum-dwelling children. In the words ofUnger and Riley, slums are complex, and our effortsmust
matchthiscomplexity(Unger&Riley,2007).Meaningfulinvolvementofcommunitymemberswillgive
communities local ownership and help make the program more relevant and likely to succeed.
Immunization confers enormous health benefits on the individual; but also, overall, long-term
populationhealthandproductivitywillimproveduetothedeclineincommunicabledisease.Vaccines
have existed for decades, and yet vaccine-preventable diseases still kill hundreds of thousands of
children in developing countries every year. With the rich and growing body of evidence around
successfulprograminterventions,thereisnoreasonwhyIndiacannotattainuniversalcoverageinthenearfuture.
7/27/2019 Banerjee Thesis
31/38
31
WorksCited
1. AgarwalS,BhanotA,GoindiG.(2005)Understandingandaddressingchildhoodimmunizationcoverageinurbanslums. IndianJournalofPediatrics,42(7):653-63.Accessed27May,2009.
http://www.indianpediatrics.net/july2005/july-653-663.htm
2. AgarwalS,Sagar&KGhei.(2005)TheneedfordedicatedfocusonurbanhealthwithintheNationalRuralHealthMission. IndianJournalofPublicHealth,49(3):48-55.
3. Asha(2006)AshaKiran:QuarterlyNewsletter.12(2):1-2.Accessed15April,2010.http://data.ashanet.org/files/Chapters/Seattle/Newsletters/Newsletter-2006Q1.pdf
4. AwasthiS.andAgarwalS.(2003)DeterminantsofchildhoodmortalityandmorbidityinurbanslumsinIndia.IndianPediatrics,40(12):1145-61.
5. BanerjeeA,DufloE,GlennersterR,KothariD.(2008)ImprovingimmunizationcoverageinruralIndia:aclusteredrandomizedcontrolledevaluationofimmunizationcampaignswithandwithout
incentives.DepartmentofEconomics,MassachusettsInstituteofTechnology. Accessed10April,
2010.http://catalogue.polytechnique.fr/site.php?id=335&fileid=2331
6. Banerjee,JandShitoleT.(2009/10)ExploringthePhysicalandSocialDeterminantsofUrbanHealth.(unpublishedresearchstudy)HarvardSchoolofPublicHealthandtheIndianresearch
collectivePUKAR.
7. BanthiaJ.(2001).FinalPopulationTotals,UrbanAgglomerationsandTowns.NewDelhi:CensusofIndia.
8. BrnighausenT;BloomDE;CanningD;O'BrienJ.(2008)AccountingforthefullbenefitsofchildhoodvaccinationinSouthAfrica.SouthAfricanMedicalJournal,Suid-AfrikaanseTydskrifVir
Geneeskunde,98(11):842,844-6.
9. BloomDE,CanningD,JamisonD.(March,2004)Health,wealthandwelfare. FinancialDevelopment.
10.Bloom,DE,CanningD.(2006)Booms,BustsandEchoes. FinanceandDevelopment;43;3:8-1711.ChatterjeeG.(2002)ConsensusversusConfrontation:LocalAuthoritiesandStateAgenciesform
PartnershipswiththeUrbanPoorinMumbai.UN-HABITAT,Nairobi.
12.CentersforDiseaseControlandSurveillance(CDC).(1999).TenGreatestPublicHealthAchievements,1900-1999.MMWR1999:48:241-3.
13.CentralIntelligenceAgency(2008).Distributionoffamilyincome-Giniindex.In TheWorldFactbook.
14.CohenJ,DupasP.(2007)Freedistributionorcost-sharing?Evidencefromarandomizedmalariapreventionexperiment.BrookingsInstitute. BrookingsGlobalEconomyandDevelopmentWorkingPaper,No.11.
15.Cornia,G,JollyR,StewartF.(1987)Introduction.In Adjustmentwithahumanface:ProtectingtheVulnerableandPromotingGrowth.ClarendonPress,Oxford.
16.deQuadros,CA.(2008)Thewholeisgreater:howpoliowaseradicatedfromthewesternhemisphere.InPracticeofInternationalHealth,editedbyDanielPerlmanandAnanyaRoy,Oxford
7/27/2019 Banerjee Thesis
32/38
32
UniversityPress.p.56-71.
17.Duflo,E.andBanerjeeA.(2008)Interview:MITEconomistsEstherDufloandAbhijitBanerjee.(InterviewedbyTimOgdenandLauraStarita).In PhilanthropyAction.
http://www.philanthropyaction.com/articles/interview_mit_economists_esther_duflo_and_abhijit_
banerjee/
18.Economist.(2009)AnImperfectStorm.Nov13thissue,In TheWorldin2010.Accessed2April,2010.http://www.economist.com/theworldin/displaystory.cfm?story_id=14742411
19.GaudinS,YazbeckAS.(2006)ImmunizationinIndia1993-1999:wealth,gender,andregionalinequalitiesrevisited. SocialScienceMedicine,62(3):694-706.DepartmentofEconomics,Oberlin,
OH.
20.GheiK,AgarwalS,SubramanyamMA,SubramanianSV.(2010)AssociationbetweenchildimmunizationandavailabilityofhealthinfrastructureinslumsinIndia.ArchivesofPediatric&
AdolescentMedicine.164(3):243-9.
21.GoogleMaps.Imagesacquired9April,2010.22.GovernmentofIndia,MinistryofHomeAffairs,OfficeoftheRegistrarGeneral.(2001)2001Censusdata.Accessed25May,2010. http://www.censusindia.net/23.Gwatkin,DR.(1979)PoliticalWillandFamilyPlanning:TheImplicationsofIndiasEmergency
Response.PopulationandDevelopmentReview,5(1):29-59.
24.Hotez,P.(2008)ForgottenPeople,ForgottenDiseases.GeorgeWashingtonUniversityandSabinVaccineInstitute.ASMPress,Washington,DC.
25. InternationalInstituteofPopulationSciencesandMacroInternational.(2008)Errata.NationalFamilyandHealthSurvey(NFHS-32005/6).Mumbai,India.Accessed8Dec,2009.
http://www.nfhsindia.org/NFHS-3%20Data/Errata_IndiaReport_sheet_short_21Oct08.pdf
26. InternationalInstituteofPopulationSciences(IIPS),MinistryofHealthandSocialWelfare-GovernmentofIndia.(2009)Healthandlivingconditionsineighturbancities.BasedonNFHS-32005/6.Deonar,Mumbai,India.
27. Jaju,NB.(17June,2009)KJSomaiyaHospital,Sion,Mumbai,India.(InterviewbyJoyaBanerjee)28.KremerME.(2007)TheIllusionofSustainability, QuarterlyJournalofEconomics;1007-1065.29.MahadeviaD,ShahP.(2009)TenuresecurityandUrbanPoverty,SocialProtectioninAsia.30.MehtaS.(2004)MaximumCity:BombayLostandFound.VintageBooks,NewYork.31.Mercer.(2009).WorldwideCostofLivingsurvey.Accessed4May,2010.
http://www.mercer.com/costoflivingpr
32.Montgomery,M.andNationalResearchCouncil.(2003) Citiestransformed:demographicchangeanditsimplicationsinthedevelopingworld.Washington,DC,NationalAcademiesPress.
33.NationalFamilyandHealthSurvey(NFHS-3).(2005/6).IndianInstituteofPopulationSciences,Mumbai.
34.MorrisSS,FloresF,OlintoP,MedinaJM.(2004)MonetaryincentivesinprimaryhealthcareandeffectsonuseandcoverageofpreventivehealthcareinterventionsinruralHonduras:cluster
randomizedtrial.Lancet.364:2030-2037.
7/27/2019 Banerjee Thesis
33/38
33
35.PandeRP,YazbeckAS.(2003)What'sinacountryaverage?Wealth,gender,andregionalinequalitiesinimmunizationinIndia.InternationalCenterforResearchonWomen,Washington,
DC.SocialScienceMedicine,57(11):2075-88.
36.PeopleofKibera(2010).ADesperateSituation.Accessed12May,2010.http://www.peopleofkibera.com/kibera/life
37.PUKAR(PartnersinUrbanKnowledge,ActionandResearch).(2008/9)HouseholdsurveyofKB38.PUKARandtheHarvardSchoolofPublicHealth(HSPH),(2010).UrbanSlumsandTenureStatus:
WhataretheImplicationsforHealth?(InPress).
39.RileyLW,KoAI,UngerA,ReisMG.(2007)Slumhealth:Diseasesofneglectedpopulations. BMCInternationalHealthandHumanRights,7:2.
40.Sadiq,K.(2008)VotersAcrossBorders.In PaperCitizens:HowIllegalImmigrantsAcquireCitizenshipinDevelopingCountries.OxfordUniversityPress:NewYork.
41.Singh,M.(2001)NumberCrunchersandSlumSums. BusinessStandard.December10issue,Mumbai.
42.SlumRehabilitationAuthority,GrihaNirmanBhavan,GovernmentofIndia.NoticeNo.SRA/ADTP/DRP/DCR/01/07.Accessed12May,2010.
http://www.sra.gov.in/htmlpages/pdf/DCR_English_Notification.pdf
43.ThorntonR.(2005)ThedemandforandimpactoflearningHIVstatus:Evidencefromafieldexperiment.JobMarketPaper,HarvardUniversity,AmericanEconomicReview.
44.UngerA,RileyLW.(2007)SlumHealth:FromUnderstandingtoAction. PLoSMed4(10):e295.Accessed11April,
2010.http://www.plosmedicine.org/article/info:doi/10.1371/journal.pmed.0040295
45.UNDP(2007/8).InequalityinincomeorExpenditure/Giniindex,HumanDevelopmentReport2007/08.Accessed3February,2008. http://hdrstats.undp.org/en/indicators/161.html
46.UNICEF.(2009).Immunization:TheBigPicture.Accessed9Dec,2009.http://www.unicef.org/immunization/index_bigpicture.html
47.UNFPA.(2007)StateoftheWorldPopulation2007:UnleashingthePotentialofUrbanGrowth.NY.48.UngerA,RileyLW(2007)SlumHealth:FromUnderstandingtoAction. PLoSMed4(10):295.49.UN-HABITAT.(2006/7)StateoftheWorldsCities:TheMillenniumDevelopmentGoalsandUrban
Sustainability:30YearsofShapingtheHabitatAgenda.Accessed6Dec,2009.
http://www.un.org/Pubs/chronicle/2006/issue2/0206p24.htm
50.UN-HABITAT.(2008/9)StateoftheWorldsCities:HarmoniousCities.Accessed3May,2010.51.UNHumanSettlementProgram(2003).TheChallengeofSlums:GlobalReportonHuman
Settlements.Accessed14Dec,2009. http://www.UN-
HABITAT.org/pmss/getPage.asp?page=bookView&book=1156
52.UNPopulationDivision(2005).WorldUrbanizationProspects:The2005Revision.NewYork.Accessed9Dec,2009.http://esa.un.org/unpp/
53.UrbanHealthResourceCenter.(2005/6)KeyIndicatorforUrbanPoorinBiharfromNFHS-3andNFHS-2).Accessed14April,2010. http://uhrc.in/downloads/Factsheet-Bihar.pdf
7/27/2019 Banerjee Thesis
34/38
34
54.UrbanHealthResourceCenter.(2005/6)KeyIndicatorforUrbanPoorinUttarPradeshfromNFHS-3andNFHS-2).Accessed14April,2010. http://uhrc.in/downloads/Factsheet-UP.pdf
55.UrbanHealthResourceCenter.(2007)StandardoflivingindexbasedreanalysisofNationalFamilyHealthSurvey(NFHS-3):Indiaandstatereports.InternationalInstituteforPopulationSciences
(IIPS)andORC-Macro,Mumbai,India. http://www.uhrc.in/name-CmodsDownload-index-req-getit-
lid-92.html
56.UnitedStatesCensusBureau,PopulationDivision.(2008)AnnualEstimatesoftheResidentPopulationforIncorporatedPlacesOver100,000.
57.Vintage2008PopulationEstimates:IncorporatedPlacesandMinorCivilDivisionsover100,000.58.VaidA,MammenB,Primrose,andKangG.(2007,May)InfantMortalityinanUrbanSlum.
DepartmentofGastrointestinalSciences,ChristianMedicalCollege,TamilNadu,India. Indian
JournalofPediatrics,74(5):449453.
59.WorldBank(2005).NewGlobalPovertyEstimatesWhatitmeansforIndia".60.WorldBank(2007).WorldDevelopmentIndicators. 61.WorldBank.(2009)WorldDevelopmentIndicators.Accessed11Dec,2009.
http://datafinder.worldbank.org/about-world-development-indicators?cid=GPD_WDI
62.WorldGazetteer.World:largestcitiesandtownsandstatisticsoftheirpopulation."Retrieved13May,2010.
63.WorldHealthOrganization(WHO)andUNICEF.(2005)Globalimmunizationvisionandstrategy.Geneva.Accessed17April,2008.
http://www.who.int/vaccines-documents/DocsPDF05/GIVS_Final_EN.pdf
64.WHOandUNICEF.(2008)GlobalImmunizationData.Geneva.Accessed17April,2008.http://www.who.int/immunization/newsroom/Global_Immunization_Data.pdf
65.WHO.(2006)GlobalPolioEradicationNowHingesonFourCountries;Polio-freeCountriesSeektoProtectThemselves.Geneva.
http://www.who.int/mediacentre/news/releases/2006/pr56/en/index.html
66.WHO.(2008)IndiaImmunizationProfile-India.Accessed14June,2009.http://www.who.int/vaccines/globalsummary/immunization/countryprofileresult.cfm
67.WHO.ImmunizationWebpage.Accessed14June,2009.http://www.who.int/topics/immunization/en/
7/27/2019 Banerjee Thesis
35/38
35
Exhibit1:ChildHealthSurveyData
Table10:ChildHealthHistories
(n=258childrenunder5yearsoldin226households)
Number %ChildsAge
7/27/2019 Banerjee Thesis
36/38
36
Table11:ImmunizationStatus
Immunizations # %
None 29 11.24%
Fullyimmunized 75 29.07%Donotremember 105 40.70%
Hasvaccinationcard(listingdatesofeach
vaccine) 75 29.07%
Nocard(databasedonrecall)* 59 22.87%
OPV 245 94.96%
NoOPV 13 5.03%
BCG 229 88.76%
DPT1(6weeks) 127 49.22%
DPT2(10weeks) 116 44.96%
DPT3(14weeks) 104 40.31%
Measles(9-12months) 82 31.78%
7/27/2019 Banerjee Thesis
37/38
37
Exhibit2:TheNotificationProcessandtheImpactofTenureonHealth.
(Takenfromtheforthcomingmanuscript,PUKAR&HSPH2010,asadaptedfromMahadevia&Shah,
2009)
Figure13.Theevolutionofslumliving
Figure13.Consequencesofinsecuretenureonhealthandwellbeing
7/27/2019 Banerjee Thesis
38/38
Figure14.Consequencesofinsecuretenureonhealthandwellbeing
Recommended