The value of telemonitoring and ICT-guided
disease management in heart failure: Results
from the IN TOUCH study
Imke Kraai, Arjen de Vries, Karin Vermeulen, Vincent van Deursen, Martje van der Wal,
Richard de Jong, Rene van Dijk, Tiny Jaarsma, Hans Hillege and Ivonne Lesman
Linköping University Post Print
N.B.: When citing this work, cite the original article.
Original Publication:
Imke Kraai, Arjen de Vries, Karin Vermeulen, Vincent van Deursen, Martje van der Wal,
Richard de Jong, Rene van Dijk, Tiny Jaarsma, Hans Hillege and Ivonne Lesman, The value
of telemonitoring and ICT-guided disease management in heart failure: Results from the IN
TOUCH study, 2016, International Journal of Medical Informatics, (85), 1, 53-60.
http://dx.doi.org/10.1016/j.ijmedinf.2015.10.001
Copyright: 2015 The Authors. Published by Elsevier Ireland Ltd. This is an open access article
under the CC BY-NC-ND license
http://www.elsevier.com/
Postprint available at: Linköping University Electronic Press
ContentslistsavailableatScienceDirect
International
Journal
of
Medical
Informatics
jo u r n al h o m e p a g e :w w w . i j m i j o u r n a l . c o m
The
value
of
telemonitoring
and
ICT-guided
disease
management
in
heart
failure:
Results
from
the
IN
TOUCH
study
Imke
Kraai
a,∗,
Arjen
de
Vries
a,
Karin
Vermeulen
b,
Vincent
van
Deursen
a,
Martje
van
der
Wal
d,
Richard
de
Jong
e,
René
van
Dijk
c,
Tiny
Jaarsma
d,
Hans
Hillege
b,
Ivonne
Lesman
baDepartmentofCardiology,UniversityMedicalCenterGroningen,UniversityofGroningen,TheNetherlands bDepartmentofEpidemiology,UniversityMedicalCenterGroningen,UniversityofGroningen,TheNetherlands cDepartmentofCardiology,CavariClinicsGroningen,TheNetherlands
dLinköpingUniversity,FacultyofHealthSciences,DepartmentofSocialandWelfareStudies,Linköping,Sweden eDepartmentofCardiology,WilhelminaHospitalAssen,TheNetherlands
a
r
t
i
c
l
e
i
n
f
o
Articlehistory:
Received19January2015
Receivedinrevisedform7October2015 Accepted8October2015 Keywords: Heartfailure Diseasemanagement Telemonitoring ICT-guideddisease-management Computerdecisionsupport
a
b
s
t
r
a
c
t
Aim:Itisstillunclearwhethertelemonitoringreduceshospitalizationandmortalityinheartfailure(HF) patientsandwhetheraddinganInformationandComputing Technology-guided-disease-management-system(ICT-guided-DMS)improvesclinicalandpatientreportedoutcomesorreduceshealthcarecosts. Methods:AmulticenterrandomizedcontrolledtrialwasperformedtestingtheeffectsofINnovative ICT-guided-DMScombinedwithTelemonitoringinOUtpatientclinicsforChronicHFpatients(INTOUCH) withintotal179patients(meanage69years;72%male;77%inNewYorkHeartAssociationClassification (NYHA)III–IV;meanleftventricularejectionfractionwas28%).Patientswererandomizedto ICT-guided-DMSortoICT-guided-DMS+telemonitoringwithafollow-upofninemonths.Thecompositeendpoint includedmortality,HF-readmissionandchangeinhealth-relatedqualityoflife(HR-QoL).
Results:Intotal177patientswereeligibleforanalyses.Themeanscoreoftheprimarycompositeendpoint was−0.63inICT-guided-DMSvs.−0.73inICT-guided-DMS+telemonitoring(meandifference0.1,95% CI:−0.67+0.82,p=0.39).All-causemortalityinICT-guided-DMSwas12%versus15%in ICT-guided-DMS+telemonitoring(p=0.27);HF-readmission28%vs.27%p=0.87;all-causereadmissionwas49% vs.51%(p=0.78).HR-QoLimprovedinmostpatientsandwasequalinbothgroups.Incrementalcosts wereD1360infavorofICT-guided-DMS.ICT-guided-DMS+telemonitoringhadsignificantlyfewer HF-outpatient-clinicvisits(p<0.01).
Conclusion:ICT-guided-DMS+telemonitoringforthemanagementofHFpatientsdidnotaffectthe pri-maryandsecondaryendpoints.However,wedidfindareductioninvisitstotheHF-outpatientclinicin thisgroupsuggestingthattelemonitoringmightbesafetouseinreorganizingHF-carewithrelatively lowcosts.
©2015TheAuthors.PublishedbyElsevierIrelandLtd.ThisisanopenaccessarticleundertheCC BY-NC-NDlicense(http://creativecommons.org/licenses/by-nc-nd/4.0/).
1. Introduction
Increasing shortage of resources, growing costs, increase in expensivetreatmentsanddownwardspressureonhealthcare bud-gets,necessitatesathoroughreviewofmanagement ofpatients withheartfailure(HF).Heartfailureisachronicdebilitating con-ditionwithapoorprognosisandcanbedefinedasanabnormality
∗ Correspondingauthorat:UniversityofGroningen,UniversityMedicalCenter Groningen,DepartmentofCardiology,HPCAB41,POBox30.001,9700RB Gronin-gen,TheNetherlands.Fax:+31503618062.
E-mailaddress:i.h.kraai@umcg.nl(I.Kraai).
ofthecardiacstructureorfunctionleadingtofailureoftheheartto deliveroxygenandcausesacascadeofsyndromes,symptomsand complaints[1].AfirststepinefficientlyorganizingHF-carewas takenbyimplementingspecializedHF-outpatientclinics,witha strongcollaborationbetweenHF-nursesandcardiologists[2,3].In theseHF-outpatientclinics,DiseaseManagementPrograms(DMP) wereintroduced[4,5]andoverthelastdecade,theseclinicsbecame ‘careasusual’inseveralEuropeancountries[6].Theeffecton qual-ityofcareandclinicaloutcomesinHF-clinicsintheNetherlands wasinvestigatedbytworandomizedstudies[7,8].Althoughboth studieshavecontributedtothequalityofDMPinHF-outpatient
http://dx.doi.org/10.1016/j.ijmedinf.2015.10.001
1386-5056/©2015TheAuthors.PublishedbyElsevierIrelandLtd.ThisisanopenaccessarticleundertheCCBY-NC-NDlicense(http://creativecommons.org/licenses/ by-nc-nd/4.0/).
54 I.Kraaietal./InternationalJournalofMedicalInformatics85(2016)53–60
clinics led by specialized HF-nurses, the key characteristics of patientswhobenefitsmostfromaDMPremainedunclear[9].
Althoughtelecardiologyseemstobealargesuccessforthe management of pace-makers and defibrillators, the useof tele cardiologyinthefollow-upoftechnologicalfunctioningof pace-makersand defibrillators isquite differentfromtheusein the monitoringofsymptomsofdeteriorationinpatients.Guidelines suggesttoincludecareful monitoringofHFpatientsinaddition toregularoutpatientclinicvisitsbymeansoftelephonic assess-mentsortelemonitoring[1].Onacaregiverlevel,cardiologistsand HF-nurseshavehighexpectationsoftelemonitoringbyexpecting ittoimprovequalityofcare,reducecostsandimprovespatients’ Health-RelatedQualityofLife(HR-QoL)byincreasingfeelingsof controlandempowermentofthepatient[10].Ontheotherside,the misalignmentbetweenpatientsexpectationsregardingthe contin-uouslymonitoringofthedataandtheactuallynon-continuously monitoringactivities from the professionals providing the ser-viceisacauseofconcern.Onapatientlevel,patientsexperience byusing telemonitoring anincrease in their self-careactivities
[11],anincreaseintheirunderstandingofHFsymptomsandan improvementintheirfeelingofsafety[12].Furthermore,patients experiencea highlevel of satisfaction withtelemonitoring and havethefeelingthatcare-giverscontinuouslymonitorthe trans-mitted parameters (despite the knowledge that monitoring by care-giversisnon-continuously)[13].Onapopulationlevel,results regardingreductioninhospitalizationandmortalityratesbythe useoftelemonitoringinHFpatientscomparedtousualcareare ambiguous[14–19].Itissuggestedthatthedevelopmentoftools toautomaticallyanalyzethedataandprovideadvicetopatients andcaregiversregardingthetreatmentofthepatientwouldbea revolution[18].However,untilnowitisnotclearwhether tele-monitoring,whendeliveredasanintegratedapproachaddedtoan informationandcommunicationtechnologyguideddisease man-agementsystem(ICT-guided-DMS),improvesclinicalandpatient reportedoutcomes or reduces healthcare costs. The valueof a ComputerDecisionSupportSystem(CDSS)incorporatedintoan ICT-guided-DMStofacilitatehealthcareprofessionalsinto optimiz-ingtreatmentandcareisalsonotknown.Therefore,theaimofthis studyis toassesstheeffectoftelemonitoringontop ofan ICT-guided-DMSwithanCDSSinpatientswithworseningHFonthe combinedendpointofdeath,readmissionandHR-QoL,compared topatientstreatedwithICT-guided-DMSandCDSSalone.
2. Methods
2.1. INTOUCHstudy
TheINTOUCH(thevalueofINnovativeICT-guideddisease man-agementcombinedwithTelemonitoringinOUtpatientclinicsfor Chronic Heart failure patients) was a multicenter, randomized controlledstudydesignedtoinvestigatetheeffectsandcostsof ICT-guided-DMSwithCDSS,versusICT-guided-DMSwithCDSSand telemonitoringinHF-patientsonacompositeendpointof mortal-ity,HF-readmissionandchangeinHR-QoL.Therationalanddesign havebeendescribedelsewhere[20].Inshort,thestudyconsisted ofone control groupand two intervention groups. The control groupconsistedofseparatehospitalsinwhichpatientsreceived usualcarewithoutanyuseoftelemonitoringdevicesfortheentire study.Theinterventiongroupconsistedofonegroupreceiving ICT-guided-DMSwithCDSSwithouttelemonitoring,theothergroup receivedICT-guided-DMSwithCDSSandtelemonitoring,fromnow ondescribedasICT-guided-DMSgroupandtelemonitoringgroup respectively.InDecember2010,adjustmentstothestudy proto-colwerenecessaryfortworeasons.Itwasimpossibletorecruit hospitalsfortheusualcarepartofthestudy(thecontrolgroup)
becauseusing anyformof telemonitoring in thecontrolgroup duringthecompletestudyperiod wasanexclusion criterionto participateasacontrolhospitalinthisstudy.Thesecondreason fortheadjustmentofthestudyprotocolwasthelowinclusionrate ofpatientswithanHF-admissionintheinterventionhospitals.The adjustmentstothestudyprotocolwere(1)tohavearandomized interventionstudywithoutcontrolgroupand(2)broadeningofthe inclusioncriteriatoinclusionofpatientswithHF-deterioration vis-itingtheHFclinicandoutpatientHFclinicandneededtobetreated withextradiuretics.Inaddition,theinclusionperiodwasextended withninemonths.Asaresult ofthis thestudydesign changed fromastudywith3arms(acontrolgroupandtwointervention groups)toastudywith2interventionarmsandnocontrolgroup. PatientswererecruitedintenDutchhospitalsduringaperiodof intotal25months(December2009–January2012).Themedical ethicalcommitteeapprovedtheprotocolandtheadjustmentsto theprotocolwhichhadnoconsequencesforthefeasibilityofthe ongoingstudy(ABR:NL26271.042.08).Allpatientsprovided writ-teninformedconsentand ourstudy conformstotheprinciples outlinedintheDeclarationofHelsinki.
2.2. Patientpopulation
AllincludedpatientshadworseningHFbasedontypicalsigns andsymptomsoffluidretention.Inclusioncriteriawere: admis-siontotheintensivecare/coronarycareunitorcardiologyward orvisiting theoutpatient HF-clinicand inneedoftreatmentor adjustmentwithoralor intravenousdiuretics,aged18 yearsor older,evidenceofstructuralunderlyingheartdisease,documented reducedleftventricularejectionfraction(LVEF)≤45%.Exclusion criteriawere:myocardialinfarctioninthelastmonth,cardiac inva-siveinterventioninthepast6monthsorplannedinthenext3 months,hemodialysis,useofothertelemonitoringsystems, inabil-ityorunwillingnesstogiveinformedconsent.
2.3. Randomization
Randomization to one of the intervention groups was per-formed within 2 weeks after inclusion. One group received ICT-guided-DMS, the other group received telemonitoring. The computer-generated randomization scheme used random per-mutedblocksof2:1(originalprotocol)and1:1(adjustedprotocol December2010)stratifiedbycentretoensurebalancedassignment ofpatientstoeachgroupinthetenparticipatingcentres.
2.4. Intervention 2.4.1. ICT-guided-DMS
Allpatientsinboth groups. Thefunctionof theCDSS inthe DMSwastoprovidedadvicetothehealthcareproviders accord-ingtotheactualESC-HFguidelines,regardingtheup-titrationof HF-medicationtooptimalindividualdoses.TheESC-HFguidelines wereprogrammedintotheDMSasworkabledigitalflow-charts. TheadviceoftheCDSSwasbasedontheinputofdatafromnursing assessment,physicalexamination,medicalhistory,laboratoryand questionnaires;ifforexamplethevitalsignsandlaboratory val-uesofthepatientwerewithinthepre-definedrangesofthedigital flow-chart,theCDSSgeneratedanadvicetofurtherup-titratethe medicationtooptimaldosesaccordingtotheHF-guidelines.The dataregardingthenursingassessment,physicalexaminationand medicalhistorywasaddedmanuallytothesystembythenurse. ThedataregardingthelaboratorysystemwasloadedintotheCDSS automatically.Thequestionnairescouldbepresentedtothepatient electronicallyoronpaper.Ifthequestionnairewasfilledinbythe patientelectronically,thedatawasloadedintotheCDSS automat-ically.Ifthequestionnairewasfilledinbythepatientonpaper,the
datawasenteredintotheCDSSmanuallybythenurse. Further-more,thedatafromthehealth-monitorloadedautomaticallyinto theCDSS.BothgroupsreceivedcounsellingonHF-symptom man-agementand improvementofthenon-pharmacologicalregimen basedonprotocolsandrecentHF-guidelines[20].Patientsallocated tothetelemonitoringgroupwereonlyallowedtovisitthe cardi-ologistorHF-nurseincaseofanabsoluteneedforintervention. TheICT-guided-DSMgroupfollowedthenormalHF-routineofthe individualhospitals,likeanyotherHF-patient,withoutlimitations tothevisits.
2.4.2. Telemonitoring
Patientsinthetelemonitoringgroupreceivedtelemonitoring devices athome consistingof a weighingscale, blood pressure equipment,anECG-deviceandahealth-monitor.Theinstruction wastorecordweightandbloodpressureonceadayandanECG in case of starting or up-titrationof Beta-blockers. The health-monitorwasaninteractivemonitorwhichreceivedthedatafrom theweighingscale,bloodpressureequipmentandifapplicablethe ECG-deviceautomaticallyattheendofthemeasurementbymeans ofawirelessbluetoothconnection.Afterreceivingthedatafromthe abovementioneddevices,thehealth-monitorgeneratedstandard health-relatedquestionsregardingthepatients’healthstatus.The HF-nursecouldaddpre-definedandindividualrangesofweight, bloodpressureorheartrhythmintheCDSS.Incaseofdeviationof thepre-definedrangesthehealth-monitorathomegenerated auto-maticallysupplementaryquestionstoevaluatetheactualhealth situation.Thedataoftheweighingscale,bloodpressure equip-ment andif applicabletheECG-device andthe answersonthe health-relatedquestionsweredirectlytransmittedbythe health-monitorthroughtheGPRSnetworktotheDMSsystemlocatedin thehospitalandloadedintheCDSS.Incaseofdeviationofthe pre-definedrangestheHF-nursewasinformedautomaticallybymobile phoneandemail(accordingtothestudyprotocol).Topreventan increasedworkloadfortheHF-nurseandtopreventanincorrect perceptionofthepatientthatthetelemonitoringmeasurements werecontinuouslymonitored,thepatientswereinformedthatthey hadtoperformthemeasurement everymorning(includingthe weekends)beforeacertaintime.Iftherewasadeviationofthe pre-definedranges,theHF-nursecontactedthepatientwithintwo hoursanddiscussedthesymptomsandtreatmentwiththepatient
[21].
2.5. Datacollection
Thechangein HR-QoL betweenbaselineand theend ofthe studywasassessedbytheMinnesotaLivingwithHF Question-naire(MLHFQ).TheMLHFQisa21-itemdisease-specificHR-QoL instrumentwithscoresforeachitemrangingfrom0to5,a sum-maryscorerangingfrom0to105and2subscales.Higherscores representsworseHR-QoL[22].
Dataonhospitalizationsandmortalitywerecollectedfromthe medicalrecordsduringthe9monthsoffollow-up.Thereasonof hospitalreadmissionordeathandthedateoftheeventwere adju-dicatedbyan independentendpointcommitteeblindedforthe groupassignment.Thedayslosttodeathweredefinedasthe num-berofdayslostbecauseofdeathduringthe9monthsoffollow-up. Data oncosts; for thecost minimization analysesa distinc-tionwasmadebetweeninterventioncostsandresourceutilization costs.Dataonresourceusewerecollectedfrommedicalrecordsand questionnaires.Thecostoftheinterventionwasdefinedasthecosts oftheICT-guided-DMSwithCDSS,thecostsofthetelemonitoring devicesandhandlingofthealarms(personnel)andwerecalculated asafixedpriceoverthefollow-upperiod.Resourceutilizationcosts werecalculatedbycollectingdataonscheduledandnon-scheduled outpatientclinicvisitsandhospitaladmissions(HFand
non-HF-Table1
Scoringsystemfortheprimarycompositeendpoint[25].
Endpoint Score
Death(atanytimeduringstudy) −3
Survivaltoendofstudy 0
Firstreadmissionforheartfailureduringthestudyperiod −1 Noreadmissionforheartfailureduringthestudyperiod 0 Changeinqualityoflifeat9months:
•Improvement≥20units +2
•Improvementby10until19units +1
•Noimprovementby−9until+4units 0
•Worseningby+5until+9units −1
•Worseningby≥10units −2
Possiblescore −6to+2
related).Inaddition,apatientMTA-questionnaireadaptedfromthe iMTA/TiC-P[23]wasadministeredat4.5and9monthsfollow-upto collectcomplementarydata(e.g.,GP,dietician,physiotherapist vis-its,homecare,andnursinghome[daycare,andadmissions]).Unit costswereestimatedbyusingtheDutchguidelinesforcoststudies
[24]andinflatedtothepricelevelof2012usingageneralconsumer priceindex(http://www.CBS.nl).Thetimehorizonofthecost anal-ysiswas9months.Indirectcosts,suchasproductivitylosses,were notcalculated.
2.6. Endpoints
Theprimaryendpointwasacompositeweightedscore consist-ingofavalueformortality,HF-readmissionandchangeinHR-QoL betweenbaseline and end of studymeasured withtheMLHFQ withapossiblerangeof–6to+2.AHF-readmissionwasdefined asanunplannedovernighthospitalstayduetoprogressionofHF ordirectlyrelatedtoHF.Thescoringsystemoftheprimary com-positeendpointwasadaptedfromtheA-HeFTstudy[25](Table1). Theendpointhasbeenusedinseveralotherstudiesandisalso rec-ommendedinarecentstatementbytheHeartFailureAssociation
[26].Incaseofmissingdatafortheprimarycompositeendpoint, theworst-casescenariowasusedfortheanalyses.Patientslostto follow-upwereassumedtohavedied(−3),andpatientswithout aHR-QoLscorereceivedtheworstscoreforthatcomponent(−2). Secondaryendpointsofthestudyweretheseparatecomponents oftheprimaryendpoint(mortality,HF-readmissionandchangein HR-QoL).Othermajorsecondaryendpointswerethetotal num-beranddurationofallhospitaladmissions,numberofvisitstothe outpatientHF-clinicandcostanalyses.
2.7. Samplesize
Weexpectwith80%powertodetectsuperiorityfor telemon-itoringusinga one-sided,two-samplet-test,when130patients areincludedinthetelemonitoringgroupand90patientsinthe ICT-guided-DMSgroup.Thetruedifferencebetweenbothgroups wasassumedtobe0.8regardingtheprimarycompositeendpoint score.Thelargestdifferencebetweenbothgroupswasexpectedto beobservedintheHR-QoLdomain.
2.8. Statisticalanalyses
Allanalyseswereconductedaccordingtotheintention-to-treat principle (n=177). Descriptive statistics, independent t-tests in case of normal distributions, Mann–WhitneyU-tests in case of skeweddistributionsforcontinuousvariablesandChi-squaretests forcategoricalvariableswereusedtocomparethedemographic and clinical characteristics between both groups. To compare the primary composite endpoint score between both groups a
56 I.Kraaietal./InternationalJournalofMedicalInformatics85(2016)53–60
one-sidedtwo-samplet-testwasperformedand95%confidence intervalswereconstructedtodescribethetreatmentdifferences. KaplanMeiercurveswereconstructedforthedifferencesintimeto mortalitybetweenbothgroups.Costspercategorywerecalculated bymultiplyingresource usewiththecost peritem. Meansper groupandincremental costswerecalculatedbased onthetrial data.Bootstrap simulations(5000 replications ofthetrialdata) wereperformedtoestimatetheconfidenceintervalssurrounding theincrementalcosts(2.5thand97.5thpercentile).Thecostdata forhospitaladmissionswerehighlyskewedduetoonepatientwith anextremeadmissionduration(126days)therefore,weperformed thisanalysiswherebythemeanadmissiondurationforthispatient wasreplacedwiththegroupmeanvalue.Analyseswereperformed usingPASW-version18.0,Excel-version2003andR-version2.15.1. 3. Results
Overall, 179patients were randomized into thestudy. Two patientsdidnotfulfiltheinclusioncriterionofworseningHFand werethereforenot partoftheanalysis.Intotal,177HF-patients wereanalyzed,ofwhich83patientswererandomizedintothe ICT-guided-DMSgroupand94patientstothetelemonitoringgroup. Twenty-threepatients(13%)wereincludedfromtheoutpatient HF-clinic,theothersfromahospitaladmissionforHF.Thedistribution ofthenumberofpatientsinbothgroupswasnotequal,duetothe designofthestudy(seepoweranalysis)andtheforcedadjustment ofthestudyprotocol(changeinrandomizationblocks).
3.1. Patientcharacteristics
Overall,themeanageofthepatientswas69(±12)years,72% wasmale(n=128)andthemeanLVEFwas28%(±9%).During inclu-sionmostpatients(77%,n=133)wereclassifiedasNYHAIII–IV while23%(n=39)werein NYHAII.Intotal31% ofthepatients (n=54)werenewlydiagnosedwithHF(diagnosis<6months),the averageperioddiagnosedwithHFwas4.6yearsforthe ICT-guided-DMSgroupand5.5yearsforthetelemonitoringgroup(p=0.41). Otherbaselinecharacteristicssuchascomorbidityandprescribed medicationwerecomparablebetweenbothgroups(Table2).The adherenceof the patientsto telemonitoring (assessed by daily weighingandmeasuringofbloodpressure)hadamedianof95% witharangefrom87to99%forthetotalstudyperiod(notshown). 3.2. Primaryendpoint
Themean composite endpoint score in theICT-guided-DMS groupof−0.63wasnotsignificantlydifferentfromthescoreof −0.73inthetelemonitoringgroup(meandifference0.1;95%CI: −0.67+0.82,p=0.39)(Table3).Intotal,8%ofthepatients(n=14) hadtheworstpossiblescoreof−6(6patientsinthe ICT-guided-DMSgroupand8patientsinthetelemonitoringgroup).Intotal,27% ofthepatients(n=47)hadthebestpossiblescoreof+2(19patients intheICT-guided-DMSgroupand28patientsinthetelemonitoring group),withnostatisticaldifferencesbetweenthegroups(p=0.39). 3.3. Secondaryendpoints
3.3.1. Mortality
All-cause mortality in the ICT-guided-DMS group was 12% (n=10) and in the telemonitoring group 15% (n=14), p=0.27 (Table3).Theanalysesoftimetodeathshowedahazard-ratio(HR) of1.25(95%CI0.52–3.00,p=0.62).Attheendofthefollow-up,the totalnumberofdayslosttodeathwas156days(range114–206) fortheICT-guided-DMSgroupand128days(range84–217)for thetelemonitoringgroup(p=0.52).KaplanMeiersurvival
analy-Table2
Baselinecharacteristicsofthepatientsaccordingtotheassignedtreatment(n=177). ICTDMSn=83 ICTDMSwith
TMn=94 p-Value Demographics Age(yrs) 69 ±11 69 ±12 0.98 Malesex 62(75) 66(70) 0.51 LVEF% 28 ±9.0 27 ±9.9 0.57
DurationofdiagnosisHF(yrs) 4.6 ±5.6 5.5 ±6.5 0.41 NewdiagnosisofHF 25(30) 29(31) 0.92 Etiology Ischemic 35(42) 45(48) 0.45
Historyofmyocardialinfarction 25(30) 35(37) 0.32 Co-morbidities Hypertension 29(35) 26(28) 0.30 COPD 17(21) 21(22) 0.76 Atrialfibrillation 32(40) 47(48) 0.13 Diabetes 29(35) 32(34) 0.26 Stroke 6(7) 12(13) 0.22 Oncology 12(15) 12(13) 0.74 Clinicalvariables NYHAclassification 0.92 II 18(22) 21(23) III 49(60) 51(57) IV 15(18) 18(20) Heartrate(bpm) 84 ±20 81 ±19 0.42 Weight(kg) 84 ±18 81 ±17 0.40 LBTB 16(19) 31(33) 0.04 SBP(mmHg) 123 ±21 117 ±20 0.09 DBP(mmHg) 74 ±13 71 ±11 0.06 Cardiovascularinterventions CABG 15(18) 22(23) 0.38 PCI 13(16) 18(19) 0.54
ICD(CRTnotincluded) 10(12) 16(17) 0.35 Laboratory Hemoglobin,mmol/L 8.36 ±0.98 8.24 ±1.16 0.52 Sodium,mmol/L 141 ±3.07 141 ±4.33 0.94 Creatinine,mol/L 135 ±70.4 134 ±64.7 0.90 Nt-proBNP,ng/L 3672 (384–22929) 4394 (598–28887) 0.34 MeanEgfr,mL/min/1.73m2 66.3
±26.6 65.3 ±29.8 0.82 Medication Diuretica 69(89) 72(87) 0.74 ACE-inhibitor 46(59) 50(60) 0.87
Angiotensinreceptorblocker 21(27) 20(24) 0.68
Betablocker 54(69) 53(64) 0.47
Aldosteronantagonist 45(58) 49(59) 0.86 Note:valuesaregivenasmean±SD,number(percentage),ormedian(25–75 per-centiles).
Variablesweretestedtwo-sided.
ICTDMS=ICT-guided-diseasemanagementsystemwithcomputerdecision sup-port.
ICTDMSwithTM=ICT-guided-diseasemanagementsystemwithcomputer deci-sionsupportandtelemonitoring.
yrs=years;LVEF=leftventricularejectionfraction;HF=heartfailure;NYHA=New YorkHeartAssociation;bpm=beatsperminute;kg=kilograms;LBTB=left bun-dlebranchblock;SBP=systolic bloodpressure;DBP=diastolicbloodpressure; CABG=coronary artery bypass graft; PCI=percutaneous coronary interven-tion; ICD=internal cardiac defibrillator; NT-pro-BNP=N-terminal prohormone of brain natriuretic peptide; eGFR=estimated globular filtration rate: ACE-inhibitor=angiotensineconvertingenzymeinhibitor.
Table3
Primaryandsecondaryoutcomes:compositeendpointscoreandtheseparatecomponentsofthecompositeendpoint(n=177).
ICTDMSn=83 ICTDMSwithTMn=94 CI95% p-Value
Primaryendpoint Compositeendpoint −0.63 ±2.37 −0.73 ±2.59 -0.67+0.82 0.39 Secondaryendpoints Mortality 10(12) 14(15) −0.14+0.07 0.27 Re-admissionHF 23(28) 25(27) −0.12+0.14 0.87 ChangeinHR-QoL −14.63±25.14 −13.97±22.31 −8.68+7.36 0.63
Note:variablesweretestedone-sided.
Valuesaregivenasmean±SDornumber(percentage).
ICTDMS=ICT-guided-diseasemanagementsystemwithcomputerdecisionsupport.
ICTDMSwithTM=ICT-guided-diseasemanagementsystemwithcomputerdecisionsupportandtelemonitoring.
Table4
Numberofpatientswithhospitalizations,numberofhospitalizations,numberof outpatientvisitsandchangeinHR-QoL(n=177).
ICTDMS n=83
ICTDMSwith TMn=94
p-Value
Numberofpatientswithhospitalizationsn(%)
All-causehospitalization 41(49) 48(51) 0.78 Cardiovascular-relatedhospitalizations 15(18) 17(18) 0.90 Heartfailure-relatedhospitalizations 23(28) 25(27) 0.87 Totalnumberofhospitalizations
Allcause 76 96 0.58
Cardiovascular-related 23 19 0.87
Heartfailure-related 35 37 0.93
Totalnumberofvisitstooutpatientclinics(median,IQR)
All-cause 7(4–12) 6.5(4–9) 0.28 Heartfailure-related 4(0–6) 2(0–5) 0.02 ChangeinHR-QoL 0.71 Improvement≥20units 21(25) 30(32) Improvement10–19units 14(17) 7(7) Noimprovement 13(16) 18(19) Worsening 14(17) 1(15) Notavailablea 21(25) 25(27)
Note:valuesaregivenasnumber(percentage)ormedian(25–75thpercentiles). Variablesweretestedtwo-sided.
ICTDMS=ICT-guided-diseasemanagementsystemwithcomputerdecisionsupport. ICTDMSwithTM=ICT-guided-diseasemanagementsystemwithcomputerdecision supportandtelemonitoring.
HR-QoL:health-relatedqualityoflife. IQR=interquartilerange.
aOfn=46,intotal24patientshavedied.
sisdidnotshowastatisticallysignificantdifferencebetweenboth interventions(logrank2=0.025,p=0.88).
3.3.2. Readmission
Readmission for HF in the ICT-guided-DMS group was 28% (n=23)versus27%(n=25)inthetelemonitoringgroup.All-cause readmissionwas49%(n=41)intheICT-guided-DMSgroupand 51%(n=48)for thetelemonitoringgroup.The differenceswere non-significantbetweenbothgroups(Tables3and4).
3.3.3. ChangeinHR-QoL
At9monthsthemeanchangeinscorewas−15unitsinthe ICT-guided-DMSgroupversus−14unitsinthetelemonitoringgroup (95%CI:−8.7+7.4,p=0.63)(Table3).TheHR-QoLwasimproved with≥20unitsin29%(n=51)ofthepatients(21patientsinthe ICT-guided-DMSgroupand30patientsinthetelemonitoringgroup).In total12%(n=21)improvedbetween10and19units(14patients intheICT-guided-DMSgroupversus7patientsinthe telemoni-toringgroup).Intotal35%ofthepatientsdidnotimproveorhad worseHR-QoLattheendofthestudy,withnodifferencesbetween bothgroups(Table4).Regardingbothsub-scoresoftheMLHFQ,no significantdifferenceswereseenbetweenthegroups.
Table5
Bootstrapsimulations(5000replicationsofthetrialdata)toestimatetheconfidence intervalssurroundingtheincrementalcosts(2.5thand97.5thpercentile)ofthe interventions.
Costs ICT-DMS ICT-DMSwithTM Difference CI95% Intervention D37a D1766b D1729 Constant
Re-admission D3213 D2427 D786 −2875+1525 Out-of-hospitalcare D1152 D1642 D490 −1297+2880 Outpatientclinic D603 D530 D73 −261+110
Total D5006 D6366 D1360 −2263+5221
Note:ICT-DMS=ICT-guided-diseasemanagementsystemwithcomputerdecision support.
ICT-DMSwithTM=ICT-guided-diseasemanagementsystemwithcomputer deci-sionsupportandtelemonitoring.
aSoftware.
bSoftware,telemonitoringdevicesandthehandlingofincomingdata.
3.3.4. HF-outpatientclinicvisits
ThemediannumberofvisitstotheHF-outpatientclinicwas4for theICT-guided-DMSgroupversus2forthetelemonitoringgroup (p<0.02)(Table4).Toanyoutpatientclinicamedianof approxi-mately7visitsforbothgroupswasobserved(p=0.28).
3.3.5. Costanalysis
ThemeantotalcostswasD5006perpatientforthe ICT-guided-DMSgroupandD6366perpatientforthetelemonitoringgroup (Table5).IncrementalcostswereD1360infavourofpatientsin theICT-guided-DMSgroup.Inthisscenariothecostsforoutpatient visitswereD530forthetelemonitoringgroupversusD603forthe ICT-guided-DMS.HospitaladmissionswereD3213for ICT-guided-DMSandD2427forthetelemonitoringgroup.Inbothgroups,the highest proportionofcostsconsisted ofcostsforadmissions to a hospitalward.Costsmadeoutside thehospital(GP, dietician, physiotherapistvisits,homecareandnursinghome[daycare,and admissions])wereD1152intheICT-guided-DSMgroupandD1642 inthetelemonitoringgroup.Themajorcostdriverinthis cate-gorywasnursinghomeadmissions.Thehandlingoftheincoming telemonitoringdata(alerts,viewingtheICT-guided-DMS,making telephonecallstopatients,reportingtheaction/interventionand performingfollow-up)wasestimatedat12hforeachpatientfor thetotalstudyperiod,whichwascalculatedtobeD364basedon salaryofthenurses.
4. Discussion
In this study we found no additional benefit from adding telemonitoringtoanICT-guided-DMSwithCDSSwithregardto thecompositeendpointscore;all-causemortality,HF-readmission and changein HR-QoL.Weexpectedtofindadifferenceof0.8, howeverwefoundadifferenceofonly0.1.Anexplanationforthis absenceofeffectcouldbethatthestudypopulationwasrelatively “healthy”comparedtootherHF-studieswithacomparablepatient population [27], making a significant or clinical difference for
58 I.Kraaietal./InternationalJournalofMedicalInformatics85(2016)53–60
furtherimprovementinoutcomeratherdifficult.Only16patients hadmorethan1HF-readmission(9%),andarelativelargenumber ofpatients(31%)wasnewlydiagnosedwithHFandthereforedid nothaveahistoryoffrequentreadmissions.Frequentreadmissions areknownasunfavourableforoutcome[28,29],anditisknown that newly diagnosed patients stay in a stable condition for a periodaftertheyreceivingdiagnosisandcare.Alsothemortality rateinourstudywasrelativelylow(13%)andalmostallpatients (96%)receivedonlyscheduled visitstotheHF-clinic, indicating thattherewasnoneedforacutecareforworseningHF.Although in both groups the HR-QoL increased, no differences in score betweenbaselineand9monthswasseenbetweenbothgroups. Moreover,itmightbepossiblethattheeffectoftheintervention wasnotlargeenoughtobeabletodiscriminate.
Tofindaneffectoftelemonitoringonmortality,readmissions and,HR-QoLprobablyrequiresamoredefinedandcoherentgroup ofpatientsthatcouldbenefitfromthisnewintervention[30].It seemsreasonabletoassumethatothervariablessuchasage, socio-economicstatus orseverityof thediseasein termsof frequent readmissionsbeforestartingwithtelemonitoringmightaffectthe benefitoftelemonitoring,perhapsmorethanweinitiallycalculated anddefinedinthestudyprotocol.However,previousstudiesdo notshowspecificprofilesofpatientswhocouldbenefitmorefrom telemonitoring,withtheexception forpatientswithdepressive symptoms[30].
Wefoundaveryhighadherenceamongpatientswiththe tele-monitoringsystemduringtheentirestudyperiod,which shows thatourtelemonitoringsystemwaswelltoleratedbythepatient incomparison withsystems usedin otherstudies[19,30]. Fur-thermoretherewereissueswithadherencebynurses,especially regardingtotheup-titrationofHF-medicationduringthefollow-up periodinbothgroups.TheCDSSfunctionalityforup-titrating HF-medicationwasexperiencedasdifficult,whichmightbereflected bytherelativelylowpercentageofprescribedACE-inhibitorusedat theendofthefollow-upperiod.Althoughthepercentageof Beta-blockersandMineralocorticoidreceptorantagonistsareconform theobservedpercentagesintheEUROHeartsurveyIIstudy[31],the percentageofACE-inhibitorsattheendofthestudy(59%vs.60%)is lowerthantheobserved80%intheEUROHeartsurveyII.Themain reasonforthelackofuseoftheCDSSwasthattheusabilityand acceptabilitywereconsideredpoor,primarilydueto organization-andsystem-relatedbarriers,whichareknownasmajorobstacles inapplyingCDSSinhealthcare[32–34].
Aknownpersonal-relatedbarriertouseaCDSSisalackoftrust, whichwasalsoobservedinsomeparticipatinghospitalswhereHF nurseswerenotallowedtotitratemedicationthemselvesdueto hesitationsofthecardiologists.Finally,thelackofusewasalso attributedtothefact thattheCDSS wasnotintegratedintothe existingelectronicpatientrecordandthereforecausedadditional work.
Despitethefactthatwecouldnotfindadifferenceinoutcome infavourofthetelemonitoringgroup,wefoundthatthenumberof HF-relatedvisitstotheoutpatientHF-clinicwasstatistically signif-icantlowercomparedtopatientswhodidnotusetelemonitoring. ThisdifferenceinHF-relatedoutpatientHF-clinicvisitswaspartly protocol-driven;patientsinthetelemonitoringgroupwereonly allowedtovisitthecardiologistofHFnurseincase ofan abso-lute needfor intervention. Moreover,this indicatesalso that it wasnotnecessarytoschedulemoreHF-relatedvisitswhen tele-monitoringisbeingused.Theoutcomeinterms ofreadmission andmortalitywassimilarinbothgroups.Importantlyinthis con-text,theoverallpatientadherenceofusingtelemonitoringathome wasveryhigh(thetelemonitoringdeviceswereusedin95%ofthe patientsduringthewholestudyperiod)comparedtootherstudies
[19,30]inwhichpatientadherenceofusingtelemonitoringoverall wasnothigherthan30–40%orthatpatientadherencecompliance
ratesdecreasedafteraddingbloodpressuremeasurementtodaily weighing[35].Thisindicatesthatthedevicesusedforthisstudy, combinedwiththeprotocolofdailymeasurements,wereverywell acceptedbypatients.Furthermore,itwasexplainedtothepatients thatthetelemonitoringmeasurementswould onlybereviewed onceadaybytheHFnurse,namelyinthemorningaftera cer-taintimeandincaseofadeviationofthepredefinedrangesthe HFnursewouldcontactthepatientwithin2h.Thisdefined time-rangefortheuseoftelemonitoringpreventedthattheHFnurse neededtobeavailableduringthewholedayandoutside office hours.Thistime-rangewasalsoapplicableintheweekends. There-foretheHFnursecouldschedulethecontacts(ifnecessary)with thepatientsinthemorningduringtheirnormalworkroutineand itwasnottoomuchofaworkloadintheweekends.Thisprevented anincreasedworkloadfortheHF-nurseandtheperceptionofthe patientthatthetelemonitoringmeasurementswerecontinuously monitored[13].
4.1. Practicalimplications
Taking into consideration the advances and possibilities of telemonitoringandthelimitedincreaseincosts(D1360),an inter-ventionwithtelemonitoring mightbe anoption for caregivers. Forpatientswhodo nothave directordifficultaccesstoa HF-outpatient clinic(e.g.,longdistance totravel, theinabilityof a patientto visit theHFoutpatient clinic) or in case of prevent-ingregularHF-relatedvisitstotheHF-outpatientsclinicjustfor up-titrationof medication orassessment of physical condition, telemonitoringcouldbeasafeandefficientalternativefor visit-ingtheHF-outpatientclinic.Telemonitoringcouldbeasignificant toolforre-organizingHF-caretobemoreefficient,anecessitydue totheunderstaffedcareofagrowingandagingpopulation.
4.2. Limitations
Despiteanextendedinclusionperiodcombinedwithan adjust-mentofthestudyprotocol,wewerenotabletoincludethenumber ofpatientsneededascalculated.Thiscouldhaveinfluencedthe outcomeintermsofalackofsamplestrength.However,the calcu-latedp-valuesdidnotshowatrendor‘near’significance.Inaddition therewerenonumericaldifferencesbetweenbothgroups indicat-ingthatalargerpopulationwouldprobablynotmakeadifference inoutcome.Inthedesignofthestudy,acosteffectiveness anal-ysisand/orcostconsequenceanalysiswasplanned.Theplanned analyseswerenotperformedsincenodifferenceinprimaryand secondaryclinicalendpointswerefound,thereforeweonly per-formedacostminimizationanalysis.Thisstudywasdesignedas aclinicalstudy,thereforesometechnicalaspectsmightbe under-exposed.Finally,wehavenoinformationonhowmanypatients werenoteligibleforthisstudybecausethiswasnotpartofthe protocol.
5. Conclusions
ICT-guideddiseasemanagementincombinationwith telemon-itoring, usedinthefollow-up of HF-patients,didnotaffectthe primary(composite)endpointofmortality,HF-readmissionand HR-QoL,northeseparatedindividualoutcomesofthiscomposite endpoint.However,wedemonstratedthattelemonitoringissafe andcanreduceHF-relatedvisitstotheHF-outpatientclinic, keep-ingHF-careaccessible.CostshoweverwereD1360,-higherwith telemonitoring.Theadherenceofpatientsinusing telemonitor-ingwasveryhigh,indicatingthatthedevicesusedforthisstudy, in combination with daily measurements, were well accepted
Summarypoints
Whatwasalreadyknownonthetopicofthestudy:
-Onacaregiverlevel,cardiologistsandHF-nurseshavehigh expectations oftelemonitoringby expectingit to improve qualityofcare,reducecostsandtoimprovepatients’ Health-Related Qualityof Life(HR-QoL) byincreasing feelingsof controlandempowermentofthepatient.
-Onapopulationlevel,resultsregardingreductionin hospi-talizationandmortalityratesbyusingtelemonitoringinHF patientscomparedtousualcare,areambiguous.
Whatthisstudyaddedtoourknowledge:
-ItissafetousetelemonitoringanditcanreduceHF-related visitstotheHF-outpatientclinic,keepingHF-careaccessible. Howeverthecostswere&z.euro;1360,-higherwith telemon-itoring.
-The adherence of patients using telemonitoring was very high indicatingthatthedevicesused incombinationwith dailymeasurementswerewellacceptedbypatients. -Wefoundnoadditionalbenefitfromaddingtelemonitoring
toanICT-guided-DMSwithCDSSwithregardtothe compos-iteendpointscore;all-causemortality,HF-readmissionand changeinHR-QoL.
bypatientsandthereforeacceptablewhenimplementingregular care.
Authors’contributions
ImkeKraai,ArjendeVries,HansHillege,andIvonneLesman: designing,conducting,analyzingandinterpretingdataand draft-ingthearticle.VincentvanDeursen:conductingsurvivalanalyses. KarinVermeulen:analyzingandconductingcostanalyses.Allother authorsmadesubstantialcontributionstothedesignofthestudy, interpretingdata, draftingthearticletorevisingit criticallyfor importantintellectualcontent.
Funding
ThisprojectwasfundedbytheDutchMinistryofHealth, Depart-mentofPharmaceuticalAffairsandMedicalTechnology(GMT).
Conflictsofinterest
RenévanDijk:MedicaladvisorCURIT.Allotherauthors:none declared.
Acknowledgments
WethankalltheparticipatingDutchheartfailureclinicsinthis studyfortheirinvolvement,workandcooperation:Antonius Hos-pital,Sneek;CatharinaHospital,Eindhoven;Canisius-Wilhelmina Hospital, Nijmegen; Deventer Hospital, Deventer; Diakonessen Hospital, Utrecht; HAGA Hospital, Den Haag; Martini Hospital, Groningen; Medical Center Leeuwarden, Leeuwarden; Rijnland Hospital, Leiderdorp and University Medical Center Groningen, Groningen.We alsowould liketothankI.C.C.vanderHorstfor hisworkregardingtheendpointcommittee.
AppendixA. Supplementarydata
Supplementarydataassociatedwiththisarticlecanbefound, intheonlineversion,athttp://dx.doi.org/10.1016/j.ijmedinf.2015. 10.001.
References
[1]J.J.McMurray,S.Adamopoulos,S.D.Anker,A.Auricchio,M.Bohm,K. Dickstein,etal.,ESCguidelinesforthediagnosisandtreatmentofacuteand chronicheartfailure2012:thetaskforceforthediagnosisandtreatmentof acuteandchronicheartfailure2012oftheEuropeanSocietyofCardiology. Developedincollaborationwiththeheartfailureassociation(HFA)oftheESC, Eur.J.HeartFail.14(2012)803–869.
[2]T.Jaarsma,A.Stromberg,HeartfailureclinicsinEurope,Prog.Cardiovasc. Nurs.15(2000)67–68.
[3]T.Jaarsma,B.Tan,R.J.Bos,D.J.vanVeldhuisen,Heartfailureclinicsinthe Netherlandsin2003,Eur.J.Cardiovasc.Nurs.3(2004)271–274. [4]A.Takeda,S.J.Taylor,R.S.Taylor,F.Khan,H.Krum,M.Underwood,Clinical
serviceorganisationforheartfailure,CochraneDatabaseSyst.Rev.9(2012) CD002752.
[5]G.Ellrodt,D.J.Cook,J.Lee,M.Cho,D.Hunt,S.Weingarten,Evidence-based diseasemanagement,JAMA278(1997)1687–1692.
[6]F.A.McAlister,F.M.Lawson,K.K.Teo,P.W.Armstrong,Asystematicreviewof randomizedtrialsofdiseasemanagementprogramsinheartfailure,Am.J. Med.110(2001)378–384.
[7]T.Jaarsma,M.H.VanDerWal,J.Hogenhuis,I.Lesman,M.L.Luttik,N.J.Veeger, etal.,DesignandmethodologyoftheCOACHstudy:amulticenter
randomisedcoordinatingstudyevaluatingoutcomesofadvisingand counsellinginheartfailure,Eur.J.HeartFail.6(2004)227–233.
[8]P.W.Bruggink-AndredelaPorte,D.J.Lok,J.vanWijngaarden,J.H.Cornel,D. Pruijsers-Lamers,D.J.vanVeldhuisen,etal.,Heartfailureprogrammesin countrieswithaprimarycare-basedhealthcaresystem.Areadditionaltrials necessary?DesignoftheDEAL-HFstudy,Eur.J.HeartFail.7(2005) 910–920.
[9]L.A.Savard,D.R.Thompson,A.M.Clark,Ameta-reviewofevidenceonheart failurediseasemanagementprograms:thechallengesofdescribingand synthesizingevidenceoncomplexinterventions,Trials12(2011), 194-6215-12-194.
[10]A.E.deVries,M.H.vanderWal,M.M.Nieuwenhuis,R.M.deJong,R.B.vanDijk, T.Jaarsma,etal.,Healthprofessionals’expectationsversusexperiencesof internet-basedtelemonitoring:surveyamongheartfailureclinics,J.Med. InternetRes.15(2013)e4.
[11]E.Seto,K.J.Leonard,J.A.Cafazzo,J.Barnsley,C.Masino,H.J.Ross,Developing healthcarerule-basedexpertsystems:casestudyofaheartfailure telemonitoringsystem,Int.J.Med.Inform.81(2012)556–565.
[12]P.Lynga,B.Fridlund,A.Langius-Eklof,K.Bohm,Perceptionsoftransmissionof bodyweightandtelemonitoringinpatientswithheartfailure?Int.J.Qual. Stud.HealthWell-being8(2013)21524.
[13]P.Fairbrother,J.Ure,J.Hanley,L.McCloughan,M.Denvir,A.Sheikh,etal., Telemonitoringforchronicheartfailure:theviewsofpatientsandhealthcare professionals—aqualitativestudy,J.Clin.Nurs.23(2014)132–144. [14]C.Klersy,A.DeSilvestri,G.Gabutti,F.Regoli,A.Auricchio,Ameta-analysisof
remotemonitoringofheartfailurepatients,J.Am.Coll.Cardiol.54(2009) 1683–1694.
[15]J.P.Riley,M.R.Cowie,Telemonitoringinheartfailure,Heart95(2009) 1964–1968.
[16]S.C.Inglis,R.A.Clark,J.G.Cleland,Cochranesystematicreviewteam. Telemonitoringinpatientswithheartfailure,N.Engl.J.Med.364(2011) 1078–1079,authorreply1079-80.
[17]F.Koehler,S.Winkler,M.Schieber,U.Sechtem,K.Stangl,M.Bohm,etal., Telemedicalinterventionalmonitoringinheartfailure(TIM-HF),a randomized,controlledinterventiontrialinvestigatingtheimpactof telemedicineonmortalityinambulatorypatientswithheartfailure:study design,Eur.J.HeartFail.12(2010)1354–1362.
[18]C.Sousa,S.Leite,R.Lagido,L.Ferreira,J.Silva-Cardoso,M.J.Maciel, Telemonitoringinheartfailure:astate-of-the-artreview,Rev.Port.Cardiol. 33(2014)229–239.
[19]S.I.Chaudhry,J.A.Mattera,J.P.Curtis,J.A.Spertus,J.Herrin,Z.Lin,etal., Telemonitoringinpatientswithheartfailure,N.Engl.J.Med.363(2010) 2301–2309.
[20]A.E.deVries,R.M.deJong,M.H.vanderWal,T.Jaarsma,R.B.vanDijk,H.L. Hillege,ThevalueofinnovativeICTguideddiseasemanagementcombined withtelemonitoringinoutpatientclinicsforchronicheartfailurepatients. DesignandmethodologyoftheINTOUCHstudy:amulticenterrandomised trial,BMCHealthServ.Res.11(2011)167.
[21]A.E.deVries,M.H.vanderWal,W.Bedijn,R.M.deJong,R.B.vanDijk,H.L. Hillege,etal.,Follow-upandtreatmentofaninstablepatientwithheart failureusingtelemonitoringandacomputeriseddiseasemanagement system:acasereport,Eur.J.Cardiovasc.Nurs.11(2012)432–438. [22]L.D.Scott,K.Setter-Kline,A.S.Britton,Theeffectsofnursinginterventionsto
enhancementalhealthandqualityoflifeamongindividualswithheart failure,Appl.Nurs.Res.17(2004)248–256.
60 I.Kraaietal./InternationalJournalofMedicalInformatics85(2016)53–60 [23]E.Linjer,J.Jornmark,T.Hedner,B.Jonsson,Stophypertension-2group.
Predictorsforhighcostsofhospitalcareinelderlyhypertensivepatients, BloodPress.15(2006)245–250.
[24]S.S.Tan,C.A.Bouwmans,F.F.Rutten,L.Hakkaart-vanRoijen,Updateofthe Dutchmanualforcostingineconomicevaluations,Int.J.Technol.Assess. HealthCare28(2012)152–158.
[25]J.A.Franciosa,A.L.Taylor,J.N.Cohn,C.W.Yancy,S.Ziesche,A.Olukotun,etal., African-Americanheartfailuretrial(A-HeFT):rationale,design,and methodology,J.Card.Fail.8(2002)128–135.
[26]F.Zannad,A.A.Garcia,S.D.Anker,P.W.Armstrong,G.Calvo,J.G.Cleland,etal., Clinicaloutcomeendpointsinheartfailuretrials:aEuropeanSocietyof CardiologyHeartFailureAssociationconsensusdocument,Eur.J.HeartFail. 15(2013)1082–1094.
[27]T.Jaarsma,M.H.vanderWal,I.Lesman-Leegte,M.L.Luttik,J.Hogenhuis,N.J. Veeger,etal.,Effectofmoderateorintensivediseasemanagementprogram onoutcomeinpatientswithheartfailure:coordinatingstudyevaluating outcomesofadvisingandcounselinginheartfailure(COACH),Arch.Intern. Med.168(2008)316–324.
[28]B.M.Massie,P.E.Carson,J.J.McMurray,M.Komajda,R.McKelvie,M.R.Zile, etal.,Irbesartaninpatientswithheartfailureandpreservedejectionfraction, N.Engl.J.Med.359(2008)2456–2467.
[29]J.G.Cleland,M.Tendera,J.Adamus,N.Freemantle,L.Polonski,J.Taylor,etal., Theperindoprilinelderlypeoplewithchronicheartfailure(PEP-CHF)study, Eur.HeartJ.27(2006)2338–2345.
[30]F.Koehler,S.Winkler,M.Schieber,U.Sechtem,K.Stangl,M.Bohm,etal., Impactofremotetelemedicalmanagementonmortalityandhospitalizations inambulatorypatientswithchronicheartfailure:thetelemedical
interventionalmonitoringinheartfailurestudy,Circulation123(2011) 1873–1880.
[31]M.S.Nieminen,D.Brutsaert,K.Dickstein,H.Drexler,F.Follath,V.P.Harjola, etal.,EuroHeartfailuresurveyII(EHFSII):asurveyonhospitalizedacute heartfailurepatients:descriptionofpopulation,Eur.HeartJ.27(2006) 2725–2736.
[32]A.E.deVries,M.H.vanderWal,M.M.Nieuwenhuis,R.M.deJong,R.B.vanDijk, T.Jaarsma,etal.,Perceivedbarriersofheartfailurenursesandcardiologistsin usingclinicaldecisionsupportsystemsinthetreatmentofheartfailure patients,BMCMed.Inform.Decis.Mak.13(2013)54–6947.
[33]D.Short,M.Frischer,J.Bashford,Barrierstotheadoptionofcomputerised decisionsupportsystemsingeneralpracticeconsultations:aqualitative studyofGPs’perspectives,Int.J.Med.Inform.73(2004)357–362. [34]S.J.Leslie,M.A.Denvir,Clinicaldecisionsupportsoftwareforchronicheart
failure,Crit.Pathw.Cardiol.6(2007)121–126.
[35]J.G.Cleland,A.A.Louis,A.S.Rigby,U.Janssens,A.H.Balk,TEN-HMS Investigators:noninvasivehometelemonitoringforpatientswithheart failureathighriskofrecurrentadmissionanddeath:theTrans-European Network-Home-CareManagementSystem(TEN-HMS)study,J.Am.Coll. Cardiol.45(2005)1654–1664.