Dagu project baseline survey

Improving health worldwide www.lshtm.ac.uk

Dr.YemisrachB.Okwaraji

PresentedforDaguSupervisorsWorkshop

30-Jan-2017

Dagu Baseline Survey

Improving health worldwide www.lshtm.ac.uk

•  ThefirstsetofprimarydatacollectedjustbeforethefullimplementaJonofOHEP

•  Themainaimsinclude:(i)ToobtainbaselineinformaJononthecharacterisJcsofhealthproviders,households,womenaged13-49andtheirchildren;(ii)ToenableanexaminaJonofthedifferencesincharacterisJcsofthestudypopulaJonbetweentheeligibleindividualsintheintervenJonandcomparison,and(iii)ToprovideinformaJononoutcomesagainstwhichtheOHEPprojectimpactwillbeevaluated.

Aims

Study population •  Forthehouseholdsurvey

•  ArepresentaJvesampleofallhouseholdsinthestudyworeda•  withparJcularfocusonwomenaged13to49years,and•  Childrenundertheageof5years.

•  Forthehealthprovidersurvey•  HealthCentrestaffwhoprovidecareforneonatesandchildrenunder5yearsofageworkingatHealthCentreinselectedhouseholdclusters.

•  HealthExtensionWorkersintheselectedhouseholdclusters.

•  Women’sDevelopmentArmymemberswhoresideinselectedhouseholdclusters.

Designing the sample

•  Samplingframe•  WeusedexisJngsamplingframefromCSA•  thelistsofenumeraJonAreas(EA)basedonthe2007EthiopianHousingandPopulaJonCensuswereused.

•  200enumeraJonareaswereselectedwithprobabilityproporJonaltosize(PPS).

•  Stra6fica6on•  StraJfiedbyintervenJonandcomparisonarea•  Represen\aJvesampledrawnfromeachstratum

Designing the sample •  Samplesize

Survey Interven6ongroup Comparisongroup Total

Household 3000 3000 6000

Childrenaged2-59months 1747 1747 3494

HealthPosts 100 100 200

Sickchildrenobservedathealthposts

400 400 800

Selection of comparison sites •  ThecomparisonworedaswereidenJfiedbyOromia,Tigray,SNNPandAmharaRHBS

•  SupportfromtheirrespecJvelocaluniversiJes-MU,HU,JUandUoG.

•  Basedonallorsomeofthefollowingcriteria:1.  WoredaswhicharesimilartointervenJonworedasin

•  PopulaJonsizeandnumberofPHCUs•  performancedatafrom2007/2008•  LengthofWDAnetworkestablishment•  LengthofJmesincestarJngICCMandCBNCprogram•  SociodemographicstatusofthepopulaJon•  Priorexposuretoothersimilarprograms•  Burdenofdisease.Malariavs.non-Malariakebeles

2.  WoredaswithnopresenceofpartnersworkingondemandgeneraJon.

3.  Woreda’sproximitytotheintervenJonworedas

4.  WoredasadjacenttointervenJonWoredas.

Selection of comparison sites •  ForOromiaregion

•  SimilarpopulaJonsizeandsociodemographiccharacterisJcs(forexampleAgrarianvspastoralist)

•  DistancefromtheintervenJonsitetoavoidcontaminaJonofinformaJon.Sonearbydistrictswereexcluded.

•  SimilarexposurewiththeintervenJondistrictintermsofspecificprogramssuchasiCCMandCBNCprogramsor

•  SpecificburdenofdiseasesuchasMalaria.

•  ForSNNPregion•  SimilariJesinlengthofJmesinceiCCMandCBNCprogramstarted;•  LengthofJmesinceWDAnetworkestablished;•  PopulaJonsizeandnumberofPHCUs•  PerformanceonserviceuJlizaJon(e.g.ANC,facilitydelivery,PNCandimmunizaJon)

•  ForTigrayregion•  SimilariJesinperformanceoniCCMandCBNCprograms,•  LocaJonoftheWoredas(ifWoredaisadjacenttointervenJonWoreda),and•  ProximityoftheWoredafromtheintervenJonWoredas

•  ForAmhararegion•  SimilariJesinserviceutlisaJonratebasedonperformancedatafromEFY2008(LAFP,

MaternityservicecoverageincludingANC4+,SBA,earlyPNC,immunizaJoncoverage,accesstolatrine,modelHHsgraduated);

•  HowaccessibletheWoredaisand•  NopresenceofacJvepartnersworkingondemandgeneraJonprograms

Survey tools Ques6onnairedesign

•  DevelopedbyLSHTMresearchteam•  OthersourcesalsoprovidedguidestothequesJonnaires;

•  DemographicandHealthSurvey,•  IDEASCBNCandHouseholdSurvey.

•  Thesevarioussourcesensured•  Acomprehensivecoverage•  PracJcalityandvalidityofquesJonnaires.

Ques6onnairetransla6on•  FinaldraitranslatedbypartneruniversiJes•  Threelocallanguages.

•  Amharic,OromifiaandTigrigna

Survey tools •  Wehave8tools:

1.  HouseholdSurvey2. WomanDevelopmentArmySurvey3.  HealthExtensionWorkerSurvey4.  HealthProviderAssessmentSurvey5.  HealthCentreStaffSurvey6.  HealthCentreSurvey7.  HealthPostSurvey8. WoredaContextualSurvey

Study settings •  Thesurveywascarriedoutin52woredas,10zonesinfourregions.•  SNNP:•  SegenandDawurozones

•  Tigray:•  SouthEast,Eastern,SouthernandCentralzones

•  Amhara:•  AwiandNorthGondarzones

•  Oromia:•  GujiandWestHarargezones

StraJfythesamplingframebyintervenJonandcomparisonarea

Selectclustersfromeachstrata

STAGE1 STAGE2

Listhouseholdsinselectedcluster

Atwo-stagestraJfiedclustersamplingtechnique

Household Sampling

Selecthouseholdstobeinterviewed

Household listing

•  Firststage•  200clustersselectedfrom2007censusframe.•  -SystemaJcallyselectedwithprobabilityproporJonaltosize

•  Secondstage•  AcompletehouseholdlisJngwascarriedoutineachselectedcluster(EA)•  BasedonthehouseholdlisJng,30householdsareselectedfromeachclustersystemaJcally.

Health Providers Sampling •  Foreveryclusterof30households:

•  TheWDAleaderservingthehouseholdclusterwereselected•  Thehealthpostservingthesehouseholdswereselected.•  ThetwoHEWsservingtheselectedhealthpostwereselected.•  Thehealthcentreservingtheselectedhealthpostwereselected.•  WithintheHC,wecollectinformaJononeachmemberofhealthcenterstaffwhoworkonsickchildserviceswhetherornottheyhavebeentrainedinIMNCIandICCM.

•  Theworedaservingtheselectedhealthcentrewereselected

Sick Children Sampling Toiden6fysickchildren,foreverycluster:•  Interviewerswillconductcommunitymobiliza6on•  Interviewerscontacted1to30WDAleaders,•  Kebelefocalpersonand•  Otherkeyfiguresinthecommunity• Wesampledonaverage4sickchildrenperhealthpost.

Field Staff by Region

Region NumberofTeams Clusters NumberofHHs

Amhara 6 73 2190

Oromia 5 69 2070

SNNP 2 29 870

Tigray 2 29 870

Survey Field Staff •  Recruitment

•  45Interviewers•  15Teamleaders•  1Datamanager•  1CentralCoordinator•  4RegionalCoordinators(PhDStudents)

•  Training•  Twoweekstraining•  3daysfieldpiloJng

•  Fieldmaterials•  FieldManual•  Tablets•  GPS•  Other

Questionnaire in tablet

Questionnaire in tablet

Data Transfer to Central Office

Datatransfermechanisms1.  IFSS(EVDO)

2.  3Gmodem

3.  USBflashdisksentbyregionalcoordinators

Quality Control •  Atthefield

•  Welltrainedinterviewersandteamleaders•  Teamleadersrepeatinterview•  Regionalcoordinatorssupervision•  WeeklyProgressreportRegionalcoordinators

•  AtCentralOffice•  WeeklyQualityControlreport

•  Internalinconsistenies•  FeedbackloopfromCentralOffice

•  Frequencycheck

Quality Control

Quality Control

Current status

•  DatacollecJonexpectedtobecompletedinfirstweekofFebruary.

Next step

•  PreparingthedataforAnalysis•  Analysisworkshop(inMarch)•  PhDstudentswriJngtheirfirstpaper•  Baselinereportfor

•  Implementers•  FMoH