www.jped.com.br
ORIGINAL
ARTICLE
Temporal
trends
of
overall
mortality
and
hospital
morbidity
due
to
diarrheal
disease
in
Brazilian
children
younger
than
5
years
from
2000
to
2010
夽
Patrícia
S.
de
A.
Mendes
∗,
Hugo
da
C.
Ribeiro
Jr.,
Carlos
Maurício
C.
Mendes
MD.PhD.UniversidadeFederaldaBahia(UFBA),Salvador,BA,Brazil
Received6September2012;accepted31October2012 Availableonline26April2013
KEYWORDS Childhooddiarrhea; Temporalseries studies; Childmortality; Hospitalization Abstract
Objective: Toverifythe temporaltrendsoftheindicators ofoverallmortalityandhospital morbidityduetodiarrhealdiseaseinchildrenyoungerthan1yearandbetween1and4years, accordingtotheregionofBrazil,between2000and2010.
Method: Thiswasanecologicalstudyoftemporalseries.Dataonhospitaladmission authoriza-tion,meanlengthofstay,andmeanvalueoftheauthorizationwereobtainedfromtheHospital InformationSystem.ThenumberofinfantdeathswasobtainedfromtheMortalityInformation System;informationonlivebirthsandthegeneralpopulationwereobtainedfromthe Informa-tionSystemonLiveBirthsandDemographicCensuses,respectively.Thesedatawereavailable atthewebsiteoftheinformaticsdepartmentoftheBrazilianUnifiedHealthSystem/Ministry ofHealth
Results: Mortalityduetodiarrhealdisease inBrazil showedadownwardstrendinbothage groups.Regardinghospitalization,therewasaslightdownwardstrendinchildrenyoungerthan 1yearandanon-significantupwardstrendbetween1-4years,withashorterhospitalstayand lowermeanvalueofhospitalstay,regardlessofageandregion.TheNorthandNortheasthad thehighestmortalityratesandthehighestpercentageofhospitalizationsinchildrenyounger than1year.TheMidwesthadthehighestmeanannualreductioninhospitalstay.
Conclusion: Currently,theindicatorsofoverallmortalityandhospitalmorbidityduetodiarrhea inBrazilianchildrenaregenerallylower,butdecreasingslowly.
©2013SociedadeBrasileiradePediatria.PublishedbyElsevierEditoraLtda.Allrightsreserved.
夽 Pleasecitethisarticleas:MendesPS,RibeiroJr.HC,MendesCM.Temporaltrendsofoverallmortalityandhospitalmorbiditydueto
diarrhealdiseaseinBrazilianchildrenyoungerthan5yearsfrom2000to2010.JPediatr(RioJ).2013;89:315---25.
∗Correspondingauthor.
E-mail:psam.ufba@gmail.com(P.S.d.A.Mendes).
0021-7557/$–seefrontmatter©2013SociedadeBrasileiradePediatria.PublishedbyElsevierEditoraLtda.Allrightsreserved. http://dx.doi.org/10.1016/j.jped.2012.10.002
PALAVRAS-CHAVE Diarreiainfantil; Estudosdeséries temporais; Mortalidadeinfantil; Hospitalizac¸ão
Tendênciatemporaldamortalidadegeralemorbidadehospitalarpordoenc¸a diarreicaemcrianc¸asbrasileirasmenoresdecincoanosnoperíodode2000a2010
Resumo
Objetivo: Conhecerastendênciastemporaisdosindicadoresdemortalidadegeralemorbidade hospitalarpordoenc¸adiarreicaemcrianc¸asmenoresdeumanoedeumaquatroanos,conforme asregiõesbrasileiras,entre2000e2010.
Método: Estudoecológicodesériestemporais.OsdadossobreAutorizac¸ãodeInternac¸ão Hos-pitalar,médiadepermanênciaevalormédiodessaautorizac¸ãoforamobtidosdoSistemade Informac¸õesHospitalares;onúmerodeóbitosinfantisfoiadquiridonoSistemadeInformac¸ões sobreMortalidade;asinformac¸õessobreosnascidosvivoseapopulac¸ãogeralforamobtidas doSistemadeInformac¸õessobreNascidosVivosedosCensosDemográficos,respectivamente. Dadosdisponíveisnoenderec¸oeletrônicodoDepartamentodeInformáticadoSistemaÚnicode Saúde/MinistériodaSaúde.
Resultados: AmortalidadepordiarreianoBrasilevidencioutendênciadedecréscimo desacel-erado em ambas as faixas etárias. Quanto à hospitalizac¸ão, houve tendência decrescente discretanosmenoresdeumanoeascendênciainsignificanteentre1-4anos,entretanto,com menorpermanênciaevalormédiodeinternamento,independentementedaidadeedaregião. Registraram-senoNorteeNordesteosmaiorescoeficientesdemortalidadeemaior porcent-agemdeinternac¸ãonosmenoresdeumano.OCentro-Oesteapresentoumaiorreduc¸ãomédia anualdotempodepermanênciahospitalar.
Conclusão: Atualmente,osindicadoresdemortalidadegeralemorbidadehospitalarpor diar-reia em crianc¸as brasileiras encontram-se, de formageral, maisbaixos, porém lentamente decrescentes.
©2013SociedadeBrasileiradePediatria.PublicadoporElsevierEditoraLtda.Todososdireitos reservados.
Introduction
Diarrhealdisease(DD)isconsideredapublichealthproblem withhighmorbidityandmortalityworldwide.1Thedemand
forhealthcareintheemergencydepartmentishigh,often resultingin hospitalizationandrisk ofdeathdue to dehy-dration.
In2008,infectiousdiseaseswereresponsiblefor approxi-matelysixmilliondeathsworldwideinchildrenyoungerthan 5years.DDwasresponsiblefor15%ofthesedeaths(1.336 million),after pneumonia (18%,1.575 million).2 InBrazil,
between1995and2005,therewere1,505,800 hospitaliza-tionsand39,421deathsofchildrenyoungerthan1yearof ageduetodiarrheaanditscomplications.3
In some countries,oral rehydration therapy (ORT) was ableyieldareductionofapproximately75%ininfantdeaths and of 61% in hospitalizations due to diarrhea between 1980and2008.4,5 However,therehasbeensome
stabiliza-tionintheratesofmorbidityandmortalitycausedbythis disease.6,7
The worldwide rates of morbidity and mortality from diarrhea, although lower, are not acceptable, consider-ing that the disease can be prevented through relatively simplepublichealthmeasures.Recentadvancesinthe pre-vention and treatment of diarrheal disease, such as the formulation of improved oral rehydration solution, zinc supplementation,rotavirusvaccines,andvitaminA supple-mentation,aresomeofthemeasuresproposedbytheWorld HealthOrganizationandtheUnitedNationsChildren’sFund (WHO/UNICEF)toreduce theseepidemiological indicators and revitalize DD control.1 However, evidence suggests a
slow global progress since 2000 regarding the implemen-tationofthenewrecommendationsforthetreatmentand preventionofdiarrhea.1
Giventheaboveconsiderationsandregionaldifferences inBrazil,thisstudyaimedtounderstandthetemporaltrends ofindicatorsofoverallmortalityandhospitalmorbiditydue to diarrheal disease in children younger than 1 year and between1and4years,accordingtotheregionsofBrazil, between2000and2010.
Methods
Thiswasanecologicalstudyoftemporalseriesperformed based on hospital morbidity data from the Brazilian Uni-fiedHealthSystem(SistemaÚnicodeSaúde-SUS),suchas hospitaladmissionauthorization(HAA),meanlengthof hos-pitalstay,andmeanHAAvalueobtainedfromtheHospital InformationSystem(SistemadeInformac¸õesHospitalares ---SIH/SUS);thenumber ofinfant deathswasobtainedfrom theMortality Information System(SistemadeInformac¸ões sobre Mortalidade --- SIM); information on live births and the generalpopulation wereobtained fromtheLiveBirth InformationSystem(SistemadeInformac¸õessobreNascidos Vivos---SINASC)anddemographiccensusesfromtheBrazilian InstituteofGeographyandStatistics(InstitutoBrasileirode GeografiaeEstatística---IBGE),respectively.Thisevaluation includedcollectionofpre-existingdata,obtainedfromthe SUSdatabase,availableonthewebsiteoftheSUS Informat-icsDepartment(DATASUS)8Thecodeusedfordataselection
origin),accordingtothetenthInternationalClassificationof Diseases(ICD-10).9
The coefficientof mortality,hospitalizationrate,mean hospital stay, mean hospitalization values, and percent-ageofhospitalizationofchildrenyoungerthan1year and between1and4yearsweredescribed,astheseagegroups arethemostvulnerabletothisdisease.
Calculationprocedures
The infant mortalityrate for diarrhea wasdefinedas the totalnumberofdeathsduetodiarrheainchildrenyounger than1yearx1,000bythenumberoflivebirths(LB).The mortalityrateofchildrenagedbetween1and4yearsdue todiarrheawasdefinedasthetotalnumberofdeathsdue todiarrheainchildren inthis agegroup x100,000by the numberofchildrenagedbetween1and4years.
The rateof hospitalization due todiarrheain children youngerthan1yearwasdefinedasthenumberof hospital-izations duetodiarrheain childrenyoungerthan1year x 1,000bythenumberoflivebirthsandtherateof hospital-izationofchildrenagedbetween1and4yearsasthenumber ofhospitalizationsduetodiarrheachildrenatthisagegroup x100,000bythenumberofchildrenagedbetween1and4 years.
Theproportionofhospitalizationsduetodiarrheain chil-dren younger than 1 year was defined as the number of hospitalizationsduetodiarrheainchildrenyoungerthan1 yearx100bythetotalnumberofhospitalizationsinchildren youngerthan1yearandtheproportionofhospitalizationof childrenagedbetween1and4yearswasdefinedasthe num-berofhospitalizationsduetodiarrheainchildreninthisage groupx100bythetotalnumberofhospitalizationsof chil-drenagedbetween1and4years.Hospitalstay andmean hospitalizationvalueduetodiarrheawereobtaineddirectly fromDATASUS.
Fortheanalysisofthetemporalseries,dynamic regres-sion models were used (regression with ARIMA errors),10
as they allow for the incorporation and adjustment of the effect of a historical seriesautocorrelation, reducing suchbias whenestimating trends. Forthetotalof the 50 series studied, those that were non-stationary were dif-ferentiated.Then,structuralparametersofauto-regression andmovingaverageswereestimated(autocorrelation(AR), differentiation (d), moving average (MA)), withan ARIMA notation(AR,d,MA),aswellastheslopesoftheregressions ()representingthechangesinaveragetrendsoftheseries, peryear.
To diagnose the best model, the Akaike’s information criterion10 thatprovidedthe least valuewas obtainedfor
each series, together withthe residual analysis, observa-tionof autocorrelation andpartial autocorrelation graphs (descriptively through the Ljung-Box test), evaluation of parameter overestimation,andcomparisonof theoriginal datawiththosepredictedbythemodels.Astheentire tar-get populationwasstudied,inferentialstatisticswerenot calculated.The statistical packageR, release2.15.1,was usedfortheanalysisofdatainthistemporalseries.11
The Committeeof Ethics in Research of theComplexo Hospitalar Universitário Professor Edgard Santos (COM-HUPES)approvedthisstudyunderprotocolNo.001/01/2012
asan addendum to a previously approved project by the samecommitteeunderprotocolNo.121/2003.
Results
According to official data from the Brazilian Ministry of Health,duringthestudy periodtherewere 22,933deaths amongchildrenyoungerthan5yearsduetoICDA09(80.3% inchildrenyounger 1year)and 1,209,622hospitalizations (62.6%in children between 1and 4years); theNortheast accountedfor57%and46%,respectively.
Coefficientsofmortalityduetodiarrhea---ICD409
There was a reduction in the number of deaths in chil-dren younger than 1 year from approximately 77% in 2000 (2,738) to 2010 (632) versus 57% (541/235) among children aged 1 to 4 years. Between the extremes of the temporal series, reductions in the infant mortality rate from 0.96/1,000 to 0.39/1,000 in the North, from 1.62/1,000to0.38/1,000intheNortheast,from0.71/1,000 to0.23/1,000intheMidwest,from0.43/1,000to0.1/1,000 intheSoutheast,andfrom0.5/1,000 to0.08/1,000inthe South were observed. Regarding the coefficient of mor-tality of children aged between 1 to 4 years, there was a reduction from 6.81/100,000 to 5.04/100,000 in the North,from6.05/100,000to2.52/100,000intheNortheast, from5.76/100,000 to3.17/100,000 in the Midwest, from 1.99/100.000 to1.13/100,000 in theSoutheast, and from 2.86/100,000to0.85/100,000intheSouth.
Figs.1Aand2Ashowthedownwardtrendincoefficients
ofmortality by age groupand by regionduring the study period.Fortheagerangeyoungerthan1year, allregions showed a slow decrease over time, ranging on aver-age from 0.03 to 0.11 deaths/year/1,000 LB (3 to 11 deaths/year/100,000 LB). Although the Northeast (1= -0.11)hadthelargestannualdecrease,itwasalsotheregion, throughoutalmosttheentireperiod,withthehighestrates ofinfantmortalityfromdiarrhea(rangingfrom1.62to0.38 deaths/1,000LB),followedbytheNorth(1=-0.06).Among older children, the mean annual reduction in the coeffi-cientofmortalityrangedfrom0.09to0.66deaths/100,000
(Tables1and2).TheNorth(5.6deaths/100,000),Northeast
(4.7 deaths/100,000), and Midwest (4.7 deaths/100,000) had, on average, higher values of mortality coefficients when compared to the South (1.4 deaths/100,000) and Southeast(1.3 deaths/100,000) during the ten-year study period.
Hospitalmorbidity
In2000, the absolute number ofHAAs paid in Brazil with ICDA09,forchildrenyoungerthan1yearwas55,161,with areductionof 52%of thisvalue in 2010(26,347) and14% increase(59,533to67,858)inchildrenagedbetween1and 4years.Therateofhospitalizationfordiarrheainchildren youngerthan1yearwasnearlystable(slightmeanannual decrease) in all regions throughout the series (Fig. 1B), withthehighest reductionrateobserved intheNortheast
Northeast : Midwest : North : South : Southeast : β1 = −0.11; Arima (5.1.0) β1 = −0.05; Arima (2.1.1) β1 = −0.06; Arima (5.1.0) β1 = −0.04; Arima (5.1.0) β1 = −0.03; Arima (5.1.0) Northeast : Midwest : North : South : Southeast : β1 = −0.45; Arima (4.1.1) β1 = −0.24; Arima (6.1.0) β1 = −0.23; Arima (4.1.1) β1 = −0.03; Arima (0.0.0) β1 = −0.38; Arima (1.1.3) Northeast : Midwest : North : South : Southeast : β1 = −1.70; Arima (5.1.0) β1 = −0.83; Arima (5.1.0) β1 = −0.75; Arima (5.1.0) β1 = −0.19; Arima (5.1.0) β1 = −0.85; Arima (5.1.0) Northeast : Midwest : North : South : Southeast : β1 = −0.09; Arima (5.1.0) β1 = −0.18; Arima (1.1.0) β1 = −0.06; Arima (5.1.0) β1 = −0.10; Arima (1.1.0) β1 = −0.07; Arima (5.1.0) 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Year Year Year 0 5.0 5.0 5.0 5.0 3.0 4 8 12 16 20 24 28 32 36 40 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Year 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 0 2 4 6 8 10 12 14 16 18 20 Rate per 1.000 NV MR per 1.000 LB % Days 0.2 0.0 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
A
C
D
B
Figure1 (A)Childmortalitycoefficientper1,000LBduetodiarrhea,accordingtotheregionsofBrazil,2000-2010.(B)Child hospitalizationrateper1,000LBduetodiarrhea,accordingtotheregionsofBrazil,2000-2010.(C)Percentageofchildren hospi-talizedduetodiarrhea,accordingtotheregionsofBrazil,2000-2010.(D)Meanhospitalstayindaysduetodiarrhea,accordingto theregionsofBrazil,2000-2010.MR,mortalityrate;LB,livebirths;1,slopecoefficient.
(1=-1.70---reductionof1.7hospitalizationsperthousand LBeachyear)(Table1).
Amongolderchildren,astabilizationtrendwasobserved, with a non-significant increase (Fig. 2B), except in the Southeast,which presented a slight decreasein hospital-izationrates (1=-0.13)(Table 2).The same pattern was observed regarding the percentage of estimated annual hospitalizations in both age groups in the last decade
(Tables 1 and 2); however, it could be observed that
the North and Northeast had more hospitalized children due to diarrhea, especially among children younger than
1 year (Figs. 1C and 2C). On average, the rate of
hos-pitalizations among children younger than 1 year in the Northeastwas10.2%(standarddeviation[SD]=1.9%);inthe North, 9.8% (SD=1.2%); in the Southeast, 4% (SD=1.3%); in the Midwest, 5.4% (SD=1.0%); and in the South, 2.2% (SD=0.3%).Forolderchildren,onaverage,therateof hos-pitalizationinthe Northeastwas10.7% (SD=1.9%); inthe North,11.6%(SD=1.5%);intheSoutheast,6.5%(SD=1.0%); in the Midwest, 8.5% (SD=1.5%); and in the South, 5.3% (SD=1.2%).
Themeanlengthofhospitalstayduringthestudyperiod was4 days in thoseyounger than 1 year and3.4 days in childrenagedbetween1 and4years,withareductionof
approximately20% (0.85 days)between 2000 and 2010 in bothagegroups.
Figs.1D and2D demonstratesthat the meanlength of
stay isdecreasing andhomogeneousinall regions,except fortheMidwest,whichpresentedaheterogeneouspattern, especially in children younger than 1 year; however, the Midwestpresentedthehighestestimatedannualdecrease, althoughnegligible,inthemeanlengthofstay(days)inboth agegroups(1=-0.18/1=-0.15)(Tables1and2).
ThemeanvaluepaidbyHAAin2000/2010wasR$405.36 (US$ 221.14)/R$ 368.43 (US$ 209.72) and R$360.12 (US$ 196.46)/R$347.62(US$197.88),forchildrenyoungerthan 1yearandbetween1and4years,respectively.This repre-sentsareductionof9.1%and3.5%,respectively,foramean stay of4days.The valueofHAAin2000wasadjustedfor inflation,basedontheyear2010.
Discussion
Mortality
Sincethe1980s,theoverallinfantmortalitytrendandthat caused by diarrhea are described as downwards, both in
12
A
B
10 9 8 7 6 5 4 3 2 1 0 2000 2001 2002 2003 2004 2005 Northeast: Southeast: β1=–0.09; ARIMA (0.1.0) β1=–0.66; ARIMA (6.1.0) β1=–0.09; ARIMA (4.0.0) β1=–0.29; ARIMA (4.1.1) β1=–0.32; ARIMA (4.1.1) South: North: Midwest: 2006 2007 2008 2009 2010 Year MR per 100.000 0 2 4 6 8 10 12 14 16 18 2000 2001 2002 2003 2004 2005 Northeast: Southeast: β1=–0.13; ARIMA (3.1.2) β1=+0.08; ARIMA (4.0.1) β1=+0.34; ARIMA (3.1.0) β1=–0.03; ARIMA (5.0.1) β1=+0.11; ARIMA (3.1.1) South: North: Midwest: 2006 2007 2008 2009 2010 Year TRate per 100.000 22 26C
D
18 14 10 8 6 4 2 0 2000 2001 2002 2003 2004 2005 Northeast: Southeast: β1=–0.09; ARIMA (4.0.1) β1=+0.26; ARIMA (4.1.1) β1=+0.41; ARIMA (4.1.1) β1=+0.35; ARIMA (5.0.1) β1=+0.46; ARIMA (5.1.1) South: North: Midwest: 2006 2007 2008 2009 2010 Year % 2.5 3.0 3.5 4.0 4.5 2000 2001 2002 2003 2004 2005 Northeast: Southeast: β1=–0.08; ARIMA (3.1.1) β1=–0.11; ARIMA (4.1.1) β1=–0.06; ARIMA (2.1.0) β1=–0.15; ARIMA (6.1.0) β1=–0.07; ARIMA (2.1.0) South: North: Midwest: 2006 2007 2008 2009 2010 Year DaysFigure2 (A)Overallmortalityrateper100,000childrenaged1to4yearsduetodiarrhea,accordingtotheregionsofBrazil, 2000-2010.(B)Rateofhospitalizationper100,000childrenaged1to4yearsduetodiarrhea,accordingtotheregionsofBrazil, 2000-2010.(C)Percentageofhospitalizationinchildren1to4yearsduetodiarrhea,accordingtotheregionsofBrazil,2000-2010. (D)Meandaysofhospitalizationinchildrenaged1to4yearsduetodiarrhea,accordingtotheregionsofBrazil,2000-2010.MR, mortalityrate;1,slopecoefficient.
Brazil3,12,13andworldwide.2,6However,ithasbeenreported
thatprogressinthedeceaseofglobalmortalityhasnotbeen accelerated,whencomparedwiththreedecadesago;7the
same was observed in Brazil in the previous decade.13 In
parallel, aslowprogress has been observed inthe imple-mentationofthenewglobalrecommendationsfordiarrhea control.1 Contaminatedwater, inadequatesanitation, and
poor hygienestillaccountfor88% of worlddeathsdue to diarrhea.1
This slowprogressin themanagementofdiarrheal dis-ease may be a factor contributing to the stabilization of infantmortality,sincethisdiseaseisthesecondinfectious causeofthisindicatorworldwide.1,2
It is fair to say that the current rates of mortality attributable to this disease in Brazil, albeit low, are still unacceptable. This is a disease transmitted via fecal-oral route; it is self-limited, preventable, of simple manage-mentat home withORS,and doesnot requiretechnology orrelativelyhighcostsforitsprevention.
In some parts of Brazil, diarrhea is still a major public health problem.3 The regional heterogeneity in
mortality rates from diarrhea described here reflects the
socioeconomicandculturalinequality,aswellasthe diffi-cultyofaccesstohealthcareandsanitation.TheNortheast and Midwest showed higher rates of reduction in these coefficients when compared to the South and Southeast. Althoughtheseregions showedhigherlevelsofthese indi-cators in 2000, which could explain the higher rates of reduction, it is possible that this downward trend is a reflection of national public health strategies, such as theFamily Health program, training programsfor profes-sionals ondiarrheal disease monitoring, rotavirus vaccine campaigns, breastfeeding encouragement programs, and vitamin A supplementation. These strategies were imple-mented in important sectors, especially after the 1990s, withpromising resultsonthe impactof overall mortality, aswell as mortality due todiarrheal disease, in children youngerthan5years.3,14---17
Itispossiblethattherecentinvestmentsaimedat reduc-ing inequalities have been an important factor on infant mortalitydue todiarrheain some regions;3 however, this
assumptionshouldbetakenwithcautionduetothe limita-tionofacademicstudiesoninfantmortalityinregionssuch astheNorthandtheMidwest.13
Table1 Estimatedcoefficientsperyearfortemporalseriesofdiarrheainchildrenyoungerthan1yearaccordingtotheregion.
Series Region AR() MA(0) ˇ 2 AICc Residual
mean
ARIMA model
Trend
Childmortalitycoefficient(per1,000LB)
North ˚1=-0.57 2=-0.87 3=-0.70 4=-0.37 5=-0.67 - ˇ1=-0.06 0.00134 36.46 0.0052 (5,1,0) Stable Northeast ˚1=+0.22 2=-0.10 3=+0.33 4=-0.76 5=+0.06 - ˇ1=-0.11 0.00467 48.65 0.00018 (5,1,0) Slight decrease Southeast ˚1=+0.30 2=-0.49 3=+0.30 4=-0.52 5=+0.76 - ˇ1=-0.03 0.00019 19.74 -0.00054 (5,1,0) Stable South ˚1=+0.29 2=+0.35 3=-0.80 4=+0.29 5=+0.36 - ˇ1=-0.04 0.000822 31.08 0.00312 (5,1,0) Stable Midwest ˚1=-0.42 2=-0.65 01=-1.00 ˇ1=-0.05 0.00039 -20.29 0.00439 (2,1,1) Stable
Hospitalizationrate(per1,000LB)
North ˚1=-0.14 2=-0.51 3=-0.50 4=-0.10 5=-0.95 - ˇ1=-0.75 0.54 104.71 0.14391 (5,1,0) Slight decrease Northeast ˚1=-0.24 2=-0.43 3=-0.42 4=-0.28 5=-0.95 - ˇ1=-1.70 0.122 91.19 0.05443 (5,1,0) Slight decrease Southeast ˚1=-0.63 2=-0.07 3=-0.73 4=-0.57 5=+0.04 - ˇ1=-0.85 0.361 91.35 0.09529 (5,1,0) Slight decrease South ˚1=-0.45 2=-0.44 3=-0.38 4=-0.28 5=+0.23 - ˇ1=-0.19 0.399 90.79 0.01630 (5,1,0) Slight decrease Midwest ˚1=-0.46 2=-0.24 3=-0.23 4=-0.03 5=-0.16 - ˇ1=-0.83 3.00 109.86 0.12281 (5,1,0) Slight decrease
Meanhospitalstay(days)
North ˚1=-0.17 2=-0.44 3=-0.47 4=-0.17 5=-0.21 - ˇ1=-0.06 0.0171 58.69 0.00941 (5,1,0) Stable
Table1(Continued)
Series Region AR() MA(0) ˇ 2 AICc Residual
mean ARIMA model Trend Northeast ˚1=-0.21 2=+0.18 3=-0.02 4=-0.24 5=+0.07 - ˇ1=-0.09 0.00763 50.03 0.00011 (5,1,0) Stable Southeast ˚1=-0.06 2=+0.16 3=+0.24 4=+0.10 5=-0.82 - ˇ1=-0.07 0.00197 43.07 -0.0073 (5,1,0) Stable South ˚1=-0.58 ˇ1=-0.10 0.0124 -5.08 0.00056 (1,1,0) Slight decrease Midwest ˚1=-0.30 ˇ1=-0.18 0.0996 15.41 -0.01159 (1,1,0) Slight decrease Percentageofhospitalization(%) North ˚1=-0.87 2=-0.46 3=-0.12 4=+0.48 01=+0.98 ˇ1=-0.23 0.349 93.83 -0.00403 (4,1,1) Slight decrease Northeast ˚1=+0.12 2=+0.27 3=-0.41 4=-0.59 01=-1.00 ˇ1=-0.45 0.112 84.74 0.08130 (4,1,1) Slight decrease Southeast ˚1=-0.48 01=+0.08 02=-0.08 03=-0.99 ˇ1=-0.38 0.0645 45.67 0.026105 (1,1,3) Slight decrease South - - ˇ1=-0.03 ˇ0=+2.32 0.062 10.11 0 (0,0,0) Stable Midwest ˚1=-0.62 2=-0.08 3=-0.34 4=-0.04 5=-0.58 6=-0.90 - ˇ1=-0.24 0.0758 173.72 0.07072 (6,1,0) Slight decrease
AR(),Autoregressiveterm;ARIMA, autoregressiveintegratedmovingaverage;cAIC,correctedAkaikeinformationcriterion;MA(0), movingaverageterm;,modelcoefficient;0,intercept;1,slopecoefficient;2,varianceaverage.
Such preventionstrategies wereconsidered modelsfor export;16 however, regarding the treatment of diarrhea,
Brazildoesnotexplicitlyadopttherecommendationof low-osmolarityORS(66countries)andzincsupplementation(46 countries). This occurs, perhaps, due to the difficulty to acquire and/or handlethe products, mainly bythe finan-cialinvestmentinvolved inthisprocess.1Accordingtothe
National Survey on Demography and Health/2006, Brazil usedthe‘‘homemadehydrationsolution’’morefrequently thanotherformsofORS(government-providedsolutionand commerciallyavailablesolutions)torevertcasesofmorbid diarrhea.18
Hospitalmorbidity
Lower numbers of deaths from diarrhea have been recorded,butwithoutaproportionaldecreaseinmorbidity attributabletothisdisease,6,18whichimposesaheavy
bur-denonthepublichealthsystem.19Thepresentstudyshowed
aslowannualreductioninratesofhospitalizationfor diar-rheainchildrenyoungerthan1 yearandaslightincrease inchildrenbetween1and4years,demonstratingthat hos-pitalizations remainedgenerally stable,despite a modest increaseinsomeregionsofthecountry(North/Northeast). National literature mentions a reduction of approxi-mately40%to60%intherateofhospitalizationinchildren youngerthan1 year,3,20 withdifferentevolutionsbetween
theBrazilianregionsandstabilizationbetweentheagesof 1and4years,20 butatdifferenttimeperiods.IntheU.S.,
diarrheawasconsideredamajorcauseofhospitalizationof children younger than 5 years in 2000, with an expected reductionafterrotavirusvaccinationmeasures.21 Thisfact
wasrecorded in Brazil,15 mainlyin children younger than
1year, althoughin thepresent study asmallreductionin hospitalization rates was observed, perhaps explained by thelow and variablerotavirus vaccine coverage achieved since2006inBrazil.8EventhoughtheBrazilianMinistryof
Table2 Estimatedcoefficientsperyearfortemporalseriesofdiarrheainchildrenaged1to4yearsdependingontheregion.
Series Region AR() MA(0) ˇ 2 AICc Residual
mean
ARIMA Model
Trend
Coefficientofmortality(per100,000inhabitants)
North ˚1=-0.75 2=-0.54 3=-0.75 4=-0.54 ˇ1=-0.09 ˇ0=+6.16 0.3930 74.69 -0.0481 (4,0,0) Stable Northeast ˚1=-0.22 2=-0.14 3=+0.17 4=-0.33 01=+0.96 ˇ1=-0.32 0.0393 69.67 -0.00922 (4,1,1) Slight decrease Southeast - ˇ1=-0.09 0.0649 6.75 0.00018 (0,1,0) Stable South ˚1=+0.37 2=+0.47 3=+0.14 4=+0.47 5=+0.38 6=-0.99 ˇ1=-0.66 0.00323 157.15 0.01405 (6,1,0) Slight decrease Midwest ˚1=-0.79 2=-0.66 3=+0.82 4=-0.90 01=-1.00 ˇ1=-0.29 0.1260 90.39 0.03066 (4,1,1) Slight decrease
Hospitalization(per100,000inhabitants)
North ˚1=-1.16 2=-1.04 3=-0.75 ˇ1=+0.34 0.201 40.66 0.0028 (3,1,0) Slight increase Northeast ˚1=+0.39 2=-0.57 3=-0.73 4=+0.26 5=-0.70 01=-1.00 ˇ1=+0.11 ˇ0=+6.94 0.0665 209.73 0.0521 (5,0,1) Slight increase Southeast ˚1=+0.56 2=+0.29 3=-0.85 01=-1.84 02=+1.00 ˇ1=-0.13 0.0418 75.71 0.002002 (3,1,2) Slight decrease South ˚1=+0.56 2=-0.55 3=-0.12 4=-0.17 01=-1.00 ˇ1=+0.08 ˇ0=+2.41 0.0782 95.41 -0.04196 (4,0,1) Stable Midwest ˚1=+0.06 2=+0.10 3=-0.41 4=-0.24 5=-0.49 01=-1.00 ˇ1=-0.03 ˇ0=+6.32 0.236 221.78 0.00157 (5,0,1) Stable
Meanhospitalstay(days)
North ˚1=-0,50 2=-0,51 - ˇ1=-0.06 0.00273 -13.95 -0.00331 (2,1,0) Stable Northeast ˚1=-0,71 2=-0,70 - ˇ1=-0.07 0.00088 -24.42 0.00339 (2,1,0) Stable Southeast ˚1=-0,16 2=-0,08 3=+0,66 01=+0.62 ˇ1=-0.08 0.00223 9.97 0.00668 (3,1,1) Stable South ˚1=-0.62 2=-0.50 3=-0.75 4=-0.73 01=-1.00 ˇ1=-0.11 0.0013 41.22 -0.00207 (4,1,1) Slight decrease
Table2(Continued)
Series Region AR() MA(0) ˇ 2 AICc Residual
mean ARIMA Model Trend Midwest ˚1=-0.40 2=+0.84 3=+0.60 4=+0.84 5=-0.99 - ˇ1=-0.15 0.00053 138.94 0.00150 (6,1,0) Slight decrease Percentageofhospitalization(%) North ˚1=-1.06 2=-0.65 3=-0.23 4=+0.36 01=+0.97 ˇ1=+0.41 0.344 93.32 -0.00133 (4,1,1) Slight increase Northeast ˚1=-0.18 2=+0.03 3=-0.25 4=-0.39 5=-0.63 01=-1.00 ˇ1=+0.46 0.27 184.02 0.15042 (5,1,1) Slight increase Southeast ˚1=+0.09 2=+0.42 3=-0.37 4=-0.68 01=-1.00 ˇ1=-0.09 ˇ0=+7.19 0.0524 97.46 -0.01261 (4,0,1) Stable South ˚1=+0.43 2=-0.22 3=-0.18 4=-0.21 01=-1.00 ˇ1=+0.26 0.531 95.44 0.11548 (4,1,1) Slight increase Midwest ˚1=-0.11 2=+0.55 3=+0.03 4=-0.43 5=-0.59 01=-1.00 ˇ1=+0.35 ˇ0=+6.30 0.224 223.85 0.06615 (5,0,1) Slight increase
AR(),Autoregressiveterm;ARIMA, autoregressiveintegratedmovingaverage;cAIC,correctedAkaikeinformationcriterion;MA(0), movingaverageterm;,modelcoefficient;0,intercept;1,slopecoefficient;2,variance.
coverageachievedwas84.4%in2009,andthemeanwasonly 58.7%inthecountrybetween2006and2010.
Inspiteofthementionedreduction,gastroenteritisand its complications are the leading cause of hospitalization ofchildrenyoungerthan5yearsduetoconditionstreated inprimarycarebetween1999and2006,withhigherrates of hospitalizationinthe North,Northeast,and Midwest.20
Thetrendinthisserieswassimilar:theNorthandNortheast accountedforthegreatestnumberofhospitalizations, espe-ciallyinchildrenyoungerthan1year,throughoutthestudy period.This ispossiblyduetosocioeconomicand cultural differenceshistoricallyfoundintheseregions,inadditionto theincreaseddemandforpublichospitalsintheseregions.18
Therecommendationofhospitalizationduringanepisode of diarrheais restricted tocomplicatedcaseswithsevere dehydration due to hypovolemic shock and high risk of death,inadditiontosmallinfants,malnourishedpatients, thosewithelevatedfecal loss,andthosewhose caregiver is unable to successfully manage the disease at home.23
Furthermore,thelowlevelofeducationincertainregions ofBrazilaccountformostofthehospitalizations.18Despite
the success of preventive strategies to control diarrheal disease in Brazil, the deficit of primary care in some regionsaffectstheredirectionofcareincasesofdiarrhea
toemergencies,19,24 with presumedrisk of hospitalization
secondarytomultifactorialcauses,favoringthe mismanage-mentofthecase(intravenousrehydration,unnecessaryuse ofdrugs,excessivetests),increasingthetimeandcostof hospitalization.4,19,25---27
Inthisstudy,themeanhospitallengthofstayduringten yearswasreducedbyonly20%(0.85days),withameanof 4daysamongchildrenyoungerthan1yearand3.4daysin thoseaged 1 to4 years. The Midwest and Southhad the highestratesofannualdecrease,afactthatneedsfurther investigationtobetterunderstandsuchbehavior.
Associated with a modest reduction in hospital length of stay,there wasa small reductionin mean hospitaliza-tionvalue(adjustedforinflationduringtheperiod),andit wasassumedthatinvestmentstargetedtoreducethelength ofhospitalizationmayrepresent areductioninthehealth systemburdenattributedtodiarrhealdisease.
In Brazil, in 2010, approximately R$ 9.8 million were spentonhospitaladmissionsofchildrenyoungerthan1year withdiarrhea,andR$23.5millionforthoseagedbetween 1and4 years;8 theseresources couldhavebeen invested
innewtherapeutic recommendationsforthemanagement ofthispathology. ItisworthmentioningthattheHAA sys-temofhospitalizationcontrolisbasedonatable ofcosts
perprocedure, not necessarily associatedwiththe actual costsofthehospitalization;28 therefore,thesevaluesmay
beunderestimated.
Researchprioritiesarebeingdefinedtoreducethe over-allmorbimortalityofchildhooddiarrheaby2015;themain areas are targeted for public health and epidemiological policiesin ordertounderstand the barriers tothe imple-mentation and to optimize the available programs and interventions.29
It appears that strategies aimed at reducingthe num-ber of hospitalizations and length of hospital stay, with consequentreduction in costand risk tothe patient, are appropriate in the management of diarrheal disease in Brazil.
Conclusion
Thelevelsofmortalityratesinchildrenfromdiarrheawere found to be lower and slowly decreasing over the study period. Hospitalization rates remained stable, and there wasaslightdecreaseinhospitallengthofstay and hospi-talcostsduringdiseasemanagement.Therewereregional differencesforallindicators,exceptforthemeanhospital stay.
Clearand sociallyregulated public policiesaiming pri-marily at reducing social differences; enabling collective sanitary measures; structuring health services; training humanresources; and encouragingand financing research inlocalhealthdiagnosis,informationsolutions,prevention, andtreatmentmustbestrategicallycoordinatedsotheycan effectivelyextendthescopeandresolutionofwhathasbeen achievedwhen addressing this specific but crucial health situation.
Conflicts
of
interest
Theauthorsdeclarenoconflictsofinterest.
References
1.WardlawT,SalamaP,BrocklehurstC,ChopraM,MasonE. Diar-rhoea: why children are still dying and what can be done. Lancet.2010;375:870---2.
2.Black RE, Cousens S, Johnson HL, Lawn JE, Rudan I, Bas-saniDG,etal.Global,regional,andnationalcausesofchild mortality in 2008: a systematic analysis. Lancet. 2010;375: 1969---87.
3.Oliveira TC, Latorre MdoR.Trends in hospital admission and infantmortalityfrom diarrhea:Brazil,1995-2005.Rev Saude Publica.2010fev;44:102---11.
4.SantoshamM,ChandranA,FitzwaterS,Fischer-WalkerC,Baqui AH,BlackR.Progressandbarriersforthecontrolofdiarrhoeal disease.Lancet.2010;376:63---7.
5.WHO. Impactof oralrehydration therapyon hospital admis-sion and case-fatality rates for diarrhoeal disease: results from 11 countries. Weekly Epidemiological Record. 1988;8: 49---52.
6.Kosek M, Bern C, Guerrant RL. The global burden of diarrhoeal disease, as estimated from studies published between 1992 and 2000. Bull World Health Organ. 2003;81: 197---204.
7.Murray CJ, Laakso T, Shibuya K, Hill K, Lopez AD. Can we achieveMillenniumDevelopmentGoal4?Newanalysisof coun-trytrendsandforecastsofunder-5mortalityto2015.Lancet. 2007;370:1040---54.
8.Brasil. Ministério da Saúde. DATASUS, Informac¸ões de saúde - Brasil. 2012 [accessed 13 aug 2010]. Available from:
http://tabnet.datasus.gov.br
9.OMS. CID-10, V. 1-Classificac¸ao internacional de doenc¸as e problemas relacionados a saúde. São Paulo: Edusp; 2000.
10.ShumwayRH, Stoffer DS. Time series analysisand its appli-cations: with R examples. 2nd ed. Nova Iorque: Springer;
2006.
11.Team RDC. R: A language and environment for statistical computing. Vienna, Austria:The R Foundation for Statistical Computing.Availablefrom:http://www.r-project.org 12.HolcmanMM,LatorreMdoR,SantosJL.Evoluc¸ãodamortalidade
infantilnaregiãometropolitanadeSãoPaulo,1980-2000.Rev SaudePublica.2004;38:180---6.
13.Duarte CM. Reflexos das políticas de saúde sobre as tendências da mortalidade infantil no Brasil: revisão da lit-eraturasobrea última década.CadSaude Publica.2007;23: 1511---28.
14.RasellaD,AquinoR,BarretoML.Reducingchildhoodmortality fromdiarrheaandlowerrespiratorytractinfectionsinBrazil. Pediatrics.2010;126:e534---40.
15.do Carmo GMI, Yen C, Cortes J, Siqueira AA, de Oliveira WK,Cortez-EscalanteJJ,et al.Declineindiarrheamortality and admissions after routine childhood rotavirus immuniza-tion in Brazil: a time-series analysis. PLoS Med. 2011;8: e1001024.
16.VictoraCG.Diarrheamortality:whatcantheworldlearnfrom Brazil?JPediatr(RioJ).2009;85:3---5.
17.EscuderMM,VenancioSI,PereiraJC.Estimativadeimpactoda amamentac¸ãosobreamortalidadeinfantil.RevSaudePublica. 2003;37:319---25.
18.Brasil. Ministério da Saúde. PNDS - pesquisa nacional de demografiaesaúdedacrianc¸aedamulher[accessed12oct 2012]. Available from: http://bvsms.saude.gov.br/bvs/pnds/ index.php
19.TiederJS,RobertsonA,GarrisonMM.Pediatrichospital adher-ence to the standard of care for acute gastroenteritis. Pediatrics.2009;124:e1081---7.
20.Moura BA, Cunha RC, da, Aquino R, Medina MG, Mota EA, MacinkoJ,et al. Themaincauses ofhospitalizationfor pri-mary health care sensitive conditions in Brazil: an analysis byagegroupsandregion.RevBrasdeSaudeMaternInfantil. 2010;10:s83---91.
21.MalekMA,CurnsAT,HolmanRC,FischerTK,BreseeJS,GlassRI, etal.Diarrhea-androtavirus-associatedhospitalizationsamong childrenlessthan5yearsofage:UnitedStates,1997and2000. Pediatrics.2006;117:1887---92.
22.Brasil. Ministério da Saúde. Secretaria de Vigilância em Saúde. Departamento de Vigilância Epidemiológica. Informe técnico doenc¸a diarréica por rotavírus: vigilância epi-demiológica e prevenc¸ão pela vacina oral de rotavírus humanoVORH, 2006[accessed12oct2012].Available from: http://portal.saude.gov.br/portal/arquivos/pdf/informerota virus02032006.pdf
23.ArmonK,Stephenson T,MacFaul R, EcclestonP,Werneke U. Anevidenceandconsensusbasedguidelineforacutediarrhoea management.ArchDisChild.2001;85:132---42.
24.PileggiC,RaffaeleG, AngelilloIF.Paediatricutilizationofan emergencydepartmentinItaly.EurJPublicHealth.2006;16: 565---9.
25.Costa AD, Silva GA. Oral rehydration therapy in emergency departments. J Pediatr (Rio J). 2011;87: 175---9.
26.Costa AD, Silva GA. Use of oral rehydration therapy in the emergency unit: a clinical-based decision? Rev Paul Pediatr. 2010;28:215---20.
27.OzuahPO, Avner JR,Stein RE. Oralrehydration,emergency physicians,and practiceparameters:a nationalsurvey. Pedi-atrics.2002;109:259---61.
28.LaForgiaGM,CouttolencBF.DesempenhohospitalarnoBrasil: embuscadaexcelência.SãoPaulo:Singular;2009.
29.FontaineO, Kosek M,Bhatnagar S, Boschi-Pinto C,Chan KY, DugganC,et al.Setting researchprioritiesto reduceglobal mortalityfromchildhooddiarrhoeaby2015.PLoSMed.2009;6: e41.
