J.Appl.Environ.Biol.Sci., 8(5)74-80, 2018 | ISSN: 2090-4274 |
© 2018, TextRoad Publication | Journal of Applied Environmental and Biological Sciences |
www.textroad.com |
Linda Augustien Makalew1*, Kuntoro2, Soenarnatalina M.2, Bambang Widjanarko Otok3
1HealthPolytechnicManado,MinistryofHealth&PhDStudentFacultyofPublicHealth,AirlangaUniversity,Surabaya 2DepartmentofBiostatisticsandDemography,FacultyofPublicHealth,AirlangaUniversity,Surabaya,Indonesia 3LaboratoryofEnvironmentaland Health Statistic,InstitutTeknologiSepuluh Nopember,Surabaya,Indonesia
Received:January19,2018 Accepted:April12,2018
ABSTRACT
Tuberculosislungsattacked themost productive agegroup, social economyisweak and loweducation.One of themain risksassociated withthetransmissionofTBintheplaceoftheMinistryofHealth isderived frompatientswithTBthathas yettobeidentified.Asaresultthepatienthasnothadtheopportunitytoimmediatelytreated accordingtotherulesofthe PPITBthatright.Alltheplaceofhealth servicesneed toapplytheeffortofPPITBtoensurethecontinuationoftheaction immediatelydetectthepreventionandtreatmentofaperson suspected oridentified sufferTB.Thisresearch willbediscussed about the factorsthat were supposed to affect manycases for Tuberculosis Drug Sensitive Contracting (TBSO-1) which occurred in EastJavawith GeographicallyWeightedPoisson Regression (GWPR),becauseoverdispersion cases.Fromthe resultsoftheanalysisanddiscussion,obtainedtheresultthatGWPRmodelmoreappropriatetoanalyzethepatientsTBSO1 in East Java because ithas the value ofAICcsmaller than Poisson Regression.Healthyhouse, OverheadsBasic Health,Household surfing the Clean and Healthy Living Behavior (Perilaku Hidup Bersih dan Sehat: PHBS),health workers, theinhabitantsof the schools in the area for Tuberculosis Drug Sensitive Lung asregionsproneto contracting (TBSO-1)affectpatientsTBSO-1 on 4 groupsofdistrictsin EastJava.Thedominantfactor ininfluencing TBSO-1 inall districtsin East JavaisthepercentageofhealthyhouseandthepercentageofhouseholdsPHBS,except in Nganjuk.The number of patients with TBSO-1 Pacitan District in addition to influenced by thepercentage of healthy houseand the percentageofhouseholdsPHBSwasalsoinfluenced bytheratioofoverheadsBasicHealth.Soalsoin Ponorogodistrict, Lumajang,Bangkalan Sumenepandalsoinfluenced bythepercentageofthepopulation ofschools. KEY WORDS:TBSO-1,PoissonRegression,GWPR,AICc
INTRODUCTION
Indonesia isacountrywith thepatientTB lungstop 3in theworld after India andChina. Estimatednumber of patientswithTBofthelung in Indonesiaaround Tenpercentofthetotalnumberofpatientswith TBofthelungin theworld [1].According[2],theprevalenceofcasesofTBofthelungin Indonesiaforallageis0.4%,andthereare9(nine)province experienced inotherwordstheprevalenceofhimthan withtheyear2007.Indonesiaaroughestimationthereare115new patientsTBpositiveLungevery100,000inhabitants,mostproductiveagegroup(15-50 years),socialeconomyisweak and loweducation [3].Thestatusoftheeconomyverycloselyalsowith contractingTB,becausethesmallincomemakepeople could not live worthy to meet the conditions of health. TBSO patients with low economic level, find difficulties in the requirementsofhealthyhouseorbalanced nutrition [4].
According [5],that from 378 Respondents with the TBSO throughout the year 2010 with 2014 on 7 Hospital Education in Korea,obtained57.1%patientswith TBofprimarySO.ClinicalmanagementsTBSOusesdrugsanti-TBline Iand IIlinecausing theproblemoftoleranceand sideeffects.[6]the100 peoplewhoarealllocated living inashantytown and denselypopulated with personalhygienebad,which isthesourceofinfectionMycobacteriumtuberculosis.
TheriskofcontractingTBpatientslungscan through dropletinfection.Dropletinfectiondropletnucleifromwhich containsgermsTB(Mycobacteriumtuberculosis)can slurpbythehealthy.Theenvironmentaroundthehousecontributethe survival of Mycobacteriumtuberculosis.[7],to active case findings research done by takingbhayangkaraaround the patientswith TBofthelungthatis200metersobtainedpatientswithTBoftheLungnew32 householdsfrom364.
According [8],on theresultsofresearchtogetthelighting,humidityandventilationistheenvironmentalfactorsthat affectsignificanthouseson thegenesisoftuberculosis.[9],in hisresearch titledmultivariatanalysisoftuberculosis2012 in Surabaya City, wrote that thehealthy house, Means of Clean Water and Sanitation Projectis a factor that contribute to tuberculosis.Thesamethingwasalsoobtained fromtheresultsofresearchfallow,[10]Whosaysthatthechildrenaround theenvironmentpatientswithTBoftheLungvulnerablecontractinganddhostsneighborhood,densitytheplaceofshelter verycontributetothegenesisoftuberculosis.
Corresponding Author: LindaAugustien Makalew,DoctoralProgrammeFacultyofCivilEngineering,BrawijayaUniversity,Malang, Emailaddress:augustienlinda@yahoo.com
[6],inSouthAfricadifficultiesreachbasichealthfacilitiesisoneofthefactorsthatmakeapersonslowlydetected as fortuberculosis.[7],writewithsanitation housethatlowintheneighborhood in MaharashtraIndia,makechildren vulnerable contracting Mycobacteriumtuberculosis.According [8],ontheresultsofresearchtogetthelighting (OR=3,286),Humidity (OR=3,202),ventilation(OR=4,144).SanitationhousecontributethesurvivalofMycobacteriumtuberculosis. [11],wrote thatthemostdominantrisk factoriseducation.Tosupporttheglobaltuberculosiscontroland recognitionprogramasearly as possible TB lungson the elementaryschool and the utilization of media information need to be increased in order to decreasethecasesanddeathsduetoTBlungsespeciallyonproductiveage.
TheproblemoftreatmentTBbecomean importantpublichealth problemand need tobesolved soon.Thereforethis research examines therelationship between thenumber of cases ofpatients with TBSO-1 in East Java with the variables predictorswhich allegedlyinfluencewith howtoget thebestrelationshipmodel usingGeographicallyWeighted Poisson Regression(GWPR).
METHODOLOGY
Thisresearchusing secondarydataobtained fromtheprofiledatathehealthoftheprovinceofEastJava2015and ReportingDataP2TBEastJavaProvincialHealth Office2015.Thevariabledataisexaminedin theformoftheaddressand thedatetheenactment ofrespondentsaspatientsTBSO-1 in East Java Province[12].Thevariablesused in thisresearch consistsofoneresponsevariable(Y),thenumberofpatientswith TBSO-1and4 variablespredictors(Z),thepercentageof healthyhouse(Z1),theratiooftheBasicHealth Overheads(Z2),thepercentageofhouseholdsPHBS(Z3)and thepercentage ofthepopulation ofschools(Z4)and thelayoutofthelatitudesouth(UI)andEastlongitudelayout(vi).
Thestepsdonein theanalysisofthedatatoachievethegoalofresearch [13][14][15]
RESULTS AND DISCUSSION
Characteristics ofpatients with TBSO-1 East Java Province 2015 in Regency/Cityin East Java Province consistsof thepercentageofhealthyhouse(Z1),theratiooftheBasicHealth Overheads(Z2),thepercentageofhouseholdsclean and healthylifebehavior (PHBS)(Z3)andthepercentageofthepopulation ofschools(Z4).Thedescription ofeach research variableisasfollows.
Table 1. A description of the research Variables
Based on Table1 it is knownthatthe averagenumber ofpatients with TBSO-1 in East Java as much as 865caseswith varianceof663 cases.Theaveragepersentasehealthyhouse(Z1)of66.41%with varianceof21.3%,an averageoftheratio ofthe basichealth overheads(Z2)of 69.18% with variance of 16.85%, theaveragepercentageof households ber PHBS (Z3)of48.63%with variance15.59percentand theaveragepercentageofthepopulation ofschools(Z4)of84.53%with variance15.13%.Thisshowstheexistenceofoverdispersion TBSO-1.
Next,multicolinearityexamination onthevariablespredictorsbased onbased onthecorrelationbetweenand thevalue ofVIFeach ofwhichisshownin Table2andTable3.
Thetable4.2 stating thevalueofthecorrelationbetweenthevariablespredictors.Thevalueofthegreatcorrelation there isbetweenavariablepercentageofhealthyhouse(Z1)with theratiooftheBasicHealthOverheads(Z2)of0.362(p-value = 0.025), a variablepercentage of healthy house (Z1)and thepercentage of households PHBS (Z3)of 0.626 (p-value = 0.626).Thisindicatesthemulticolinearitybetweenavariablepercentageofhealthyhouse(Z1),theratiooftheBasicHealth Overheads(Z2),thepercentageofhouseholdsPHBS(Z3).
Othercriteriathatcan seemultikolinearitascaseisthevalueoftheVarianceInflationFactor(VIF).VIFvaluesineach ofthevariablespredictorscan beseenin thetable4.3
Table3indicates thatthereisno predictors variables which VIF value more than 10, so that there will be no cases multicolinearity.Thencan beused poissonregression modelTBSO-1 byinvolving thepercentageofhealthyhouse(Z1),the ratiooftheBasicHealth Overheads(Z2),thepercentageofhouseholdsPHBS(Z3)and thepercentageofthepopulation of schools(Z4).
Theresultsoftheparameterestimation valuereached convergenceafteriteration5.Next,testisdonesimultaneously parameterstoknowiswhether or nottheinfluenceoftheindependent variablesagainst thedependentvariableswith the hypothesisasfollows:
H0 : 0
1234
H1 :mostnoone j 0,J=1,2,3,4
2
The value of thedevianceon this analysis of 12984 and 47.3999 , Then reject H0 because
(33;0,05)
Dˆ 2 Soitcanbeconcluded thatthereareatleastoneindependentvariablesthataffectthesignificant
hitung v;
impact on thedependent variables. Then thetest isdonepartiallyparametersto knowtheinfluence of each independent variables. H0 : 0 (Variablesto-Idonotaffectsignificant)
j
H1 : j 0,(Thevariablesto-igivesignificantinfluence) J=1,2,3,4
Using theMLEmethod obtained theestimation ofparametersasfollows:
Table 4. Partial test Poisson Regression parameters on the TBSO-1
Parameters estimator StandardError Z P-value
Table4.Showthat |Z | Z ,where Z 1.96 ,sothatonasignificant5percentdeclineH0which
hitung
/20,025 meansavariablepercentageofhealthyhouse(Z1),theratiooftheBasicHealthOverheads(Z2),thepercentageofhouseholds PHBS(Z3)andthepercentageofthepopulationofschools(Z4)influentialsignificantonthenumberofpatientswithTBSO12015.WhileforthevariableratioofBasicHealthOverheads(Z2)isnotsignificantinaffectingthenumberofpatientswith TBSO-12015,becausethevalueofZsmallerthan1.96orp-value=0.748greaterthan0.05.Sopoissonregressionmodel obtainedisasfollows:
ˆ ,0001
exp 7.707 0,009 Z 0 Z 0,022 Z 0.017 Z4
1 23
Increase or decrease the number of patients with TBSO-1each district in East Java 2015 depending of the value of the coefficientofeachvariablethatinfluence.Furthermoredoneoverdispersicaseexaminationonpoissonregressionmodelthat ispresentedintable5.
Table 5shows that the value of thedeviance/db of51.8969greater than 1so that it can be concluded on poisson regressionmodelnumberofpatientswithTBSO-1eachdistrictinEastJava2015happenedoverdispersi.
TheanalysisusingtheGWPRmethodaimstoknowthevariablesthataffecttheprevalenceofGenesisdiseaseTBSO-1oneachobservationlocationwhichisintheDistrictoftheprovinceofEastJava.Followingthemodelingthenumberof patientswithTBusingGWPRmethod.
ThefirststepisdonetogetGWPRmodelistodeterminethecoordinatesofthepointlatitudeandlongitudeoneach locationtocountthedistanceeuclidean,anddeterminethe optimumbandwidthvaluesbasedonthecriteriaAICc.Thenext stepistodeterminethematrixweightedwithkernelfunction.
Thematrixpembobotobtainedforeachlocationandthenusedtoformamodel,sothatobtainedthemodelvaryin eachlocationofobservation.TheestimationofmodelparametersGWPRservedintable6below.
Table 6.The estimation of Model ParametersGWPR
ThenumberofpatientswithTBSOmodeling-1inRegency/CityoftheprovinceofEastJavausingGeographically WeightedPoisson Regressionapproach(GWPR) towhileis amodel that better ifcomparedwith thepoisson regression model.
TestingthehypothesisGWPRmodelconsistsoftwopeengujian,namelysuitabilitytestGWPRmodelandtestthe significanceoftheparametersGWPRmodel.ThefollowingistheresultsofthehypothesistestingGWPRmodel:
H0: | ( , ) u v k i i k | (Thereisnosignificantdifferencebetween thepoissonregressionmodel (global)andGWPR | |||||
---|---|---|---|---|---|---|---|
model) | |||||||
H1: | There is | at | least | one ( , )u v ;k= i i k | 1,2,….,4 (There is | a | significant difference between the poisson |
regressionmodel(global)andGWPRmodel) |
Table 7. Test the suitability of the Deviance with Pembobot GWPR Model
Table7. shows with weighted adaptive bi-square value differencedeviance/d0f of 393.465
22
ˆ
and 47.3999 ,ThenRejectH0because D Soitcanbeconcludedthemodelnumber
(33;0,05) hitung v; of the patients TBSO-1is GWPR.To pembobot adaptive Gaussianvalue ofdeviance/d0f difference of 407.514
22
ˆ
and 37.6525 ,ThenRejectH0because D Itcanbeconcludedthatthemodelofthe
(25;0,05) hitung v; number ofpatients withTBSO-1 eachdistrict inEast Java 2015isGWPR. Thevalueofthe smallest AICc on weighted AdaptiveGaussianamounting3311.965,sopembobotGWPRmodelusingadaptiveGaussian.
ThenextstepistestingthesignificanceofmodelparametersGWPRpartiallytoknowtheparametersthataffectthe numberofpatientswithTBSO-1ineachlocationofobservation.Thehypothesisthatisusedisasfollows:
uv
H0: (,) 0
k ii
H1: (,) 0;I=1,2,....,38;k=1.2,…,4
uv
k ii
With equal significance () of 5%, value t 2,037 .The following variables predictors whichaffect
(0,025;32)
significantlyoneachlocationofobservationpresentedinTable8.
Table 8. The value of T-count on Variable Parameter Coefficient predictors in each District in East Java Using Adaptive Gaussian
28"Pati. Pamekasan" | -11.0467 | -15.3865 | 28.0880 | -8.1257 |
---|---|---|---|---|
29"Pati. Sumenep" | -21.7194 | -0.2631 | 29.4300 | -5.4905 |
30"Kediri City" | 13.5314 | -49.8751 | -8.7530 | -1.4511 |
31"City of Blitar" | -21.7180 | -13.7498 | -3.0941 | -16.1372 |
32"Malang city" | -1.6011 | -12.7815 | 21.4911 | -14.8124 |
33"Probolinggo town" | -18.4198 | 20.2722 | 17.4585 | -51.5702 |
34"Pasuruan" | 7.8702 | -16.9734 | -5.5136 | 2.9481 |
35"City Mojokerto" | -22.2440 | 0.1191 | 9.8328 | -10.1911 |
36"Madiun" | -18.0865 | 1.6857 | 3.8886 | -10.6461 |
37"Surabaya City" | 9.7423 | -13.6207 | -3.6997 | -4.1342 |
38"Batu" | -15.8890 | -11.6251 | 7.3548 | -4.3164 |
Basedon thetable8,with Bandwidthsize=4can beknownthatallthelocation oftheobservationidentifiedthe variablessignificantlyinfluencedinallDistrictsinEastJavaProvinceclusteringbecome4grouponthefollowingimage.
Picture 3. The mapping of the number of patients with TBSO-1 District in East Java based significant Variable
Thepercentageofhealthyhouse(Z1),theratiooftheBasicHealthOverheads(Z2),thepercentageofhouseholdsber PHBS(Z3)andthepercentageofthepopulationofschools(Z4)affectthenumberofpatientswithTBSO-1inTrenggalek, Tulungagung, Blitar, Kediri, Malang, Jember, Banyuwangi, Bondowoso, Situbondo, Pasuruan, Probolinggo, Sidoarjo, MojokertoJombang,Madiun,Magetan,Ngawi,Bojonegoro,TubanLamongan,Gresik,Sampang,Pamekasandistrict,and Kediri City, Blitar City, Malangcity, Probolinggo, Pasuruan, CityMojokerto, Kota Madiun cityofSurabaya, Batu. The percentageofhealthyhouse(Z1),theratiooftheBasicHealthOverheads(Z2),thepercentageofhouseholdsberPHBS(Z3) affect the number of patients with TBSO-1in Pacitan district.The percentage of healthyhouse (Z1), the percentage of householdsberPHBS(Z3)andthepercentageofthepopulationofschools(Z4)affectthenumberofpatientswithTBSO-1 inPonorogo,Lumajang,BangkalanandSumenepdistrict.
CONCLUSION
BasedontheresultsoftheanalysisandthediscussioncanbetakenusaconclusionthatGWPRmodelwithadaptive Gaussianpembobot functionmoreappropriatetoanalyzethenumberofpatientswithTBSO-1inEastJavabecauseithas the value ofAICcsmaller.The classification ofthe number ofpatients with TBSO-1 in East Java based on the variablespredictorssignificantlythereare4groups.ThedominantfactorininfluencingTBSO-1inalldistrictsinEastJava isthepercentageofhealthyhouse(Z1)andthepercentageofhouseholdsPHBS (Z3), except in Nganjuk.Thenumberof patientswithTBSO-1PacitanDistrictinadditiontoinfluencedbythepercentageofhealthyhouse(Z1)andthepercentage ofhouseholdsber PHBS(Z3)wasalsoinfluencedbytheratiooftheBasicHealth Overheads(Z2).SoalsoinPonorogo, Lumajang,BangkalanSumenepdistrictandalsoinfluencedbythepercentageofthepopulationofschools(Z4).
REFERENCES
[1] KementerianKesehatanRI,2016.ProfilKesehatanIndonesia2015thed.KementerianKesehatanRI,ed.,Jakarta.
[2] KementerianKesehatanRI,2014.Pedoman NasionalPengendalianTuberkulosis2014thed.T.N.Dinihari&Siagian Vanda,eds.,Jakarta.
[3] Murray,C.J.L.etal.,2014.Global,regional,andnationalincidenceandmortalityforHIV,tuberculosis,andmalaria during1990–2013:asystematicanalysisfortheGlobalBurdenofDiseaseStudy2013.Lancet,384(9947),pp.1005– 1070.
[4] Erawatyningsih,E., Purwanta &Subekti, H.,2009. Faktor-faktor yangmempengaruhi ketidakpatuhan berobat pada penderitatuberkulosisparu.Berita Kedokteran Masyarakat,25(3),pp.117–124.
[5] Mok,J.H.etal.,2017.AdditionalDrugResistancePatternsamongMultidrug-ResistantTuberculosisPatientsinKorea: ImplicationsforRegimenDesign.,(December2014),pp.636–641.
[6] Kerkhoff, A.D. et al., 2015. The predictive value of current haemoglobin levels for incident tuberculosis and/or mortalityduringlong-termantiretroviraltherapyinSouthAfrica:acohortstudy.BMCMed,13,p.70
[7] Rakotosamimanana, S. et al., 2014. Spatial analysis of pulmonary tuberculosis in antananarivo madagascar: Tuberculosis-relatedknowledge,attitudeandpractice.PLoS ONE,9(11).
[8] Fatimah,S.,2008.FaktorKesehatanLingkunganRumahYangBerhubunganDenganKejadianTbParuDiKabupaten Cilacap (Kecamatan: Sidareja, Cipari, Kedungreja, Patimuan, Gandrungmangu, Bantarsari) Tahun 2008. Jurnal KesehatanUNDIP.
[9] Gultom,Z.A.,2012.PemetaanPenyakitTuberkulosisdiKotaSurabayaTahun2012. (1),pp.1–11.
[10] Saptawati, L. et al., 2012. Evaluasi Metode FastPlaqueTB Untuk Mendeteksi Mycobacterium Tuberculosis Pada SputumDiBeberapaUnitPelayananKesehatanDiJakarta-Indonesia.JurnalTuberkulosisIndonesia,8,pp.1–6.
[11] Nurjana,M.A.,2015.FaktorRisikoTerjadinyaTubercolosisParuUsiaProduktif(15-49Tahun)diIndonesia.Media Penelitian dan PengembanganKesehatan,25(3),pp.163–170.
[12] Dinas Kesehatan Provinsi Jawa Timur, 2016. Profil Kesehatan Provinsi Jawa Timur Tahun 2015 2016th ed. KementerianKesehatanRI,ed.,Surabaya.
[13]Cameron,A.C.&Trivedi,P.K.,1988.Regressionanalysisofcountdata.EconometricSocietyMonographs,1988(17), p.411.
[14] LeSage, J.P., 1999. The Theory and Practice of Spatial Econometrics. International Journal of Forecasting, 2(2), pp.245–246.
[15] SelviaNurulQomari,BambangWidjanarkoOtok,HariBasukiN,AriefWibowo.(2017).ModelingNumberofTetanus NeonatorumCasesUsingGeneralizedPoisson RegressionandZeroInflatedGeneralizedPoisson.J.Appl.Environ. Biol.Sci.,7(8)165-169,2017