CochraneDatabaseofSystematicReviews Computer-based diabetes self-management interventions for adults with type 2 diabetes mellitus (Review) PalK,EastwoodSV,MichieS,FarmerAJ,BarnardML,PeacockR,WoodB,InnissJD,MurrayE PalK,EastwoodSV,MichieS,FarmerAJ,BarnardML,PeacockR,WoodB,InnissJD,MurrayE. Computer-baseddiabetesself-managementinterventionsforadultswithtype2diabetesmellitus. CochraneDatabaseofSystematicReviews2013,Issue3.Art.No.:CD008776. DOI:10.1002/14651858.CD008776.pub2. www.cochranelibrary.com Computer-baseddiabetesself-managementinterventionsforadultswithtype2diabetesmellitus(Review) Copyright©2013TheCochraneCollaboration.PublishedbyJohnWiley&Sons,Ltd. TABLE OF CONTENTS HEADER . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 PLAINLANGUAGESUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 SUMMARYOFFINDINGSFORTHEMAINCOMPARISON . . . . . . . . . . . . . . . . . . . 3 BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 OBJECTIVES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 METHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 Figure1. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 Figure2. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Figure3. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 Figure4. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Figure5. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18 DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 Figure6. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 AUTHORS’CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 ACKNOWLEDGEMENTS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26 REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 CHARACTERISTICSOFSTUDIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 DATAANDANALYSES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 Analysis1.1.Comparison1HbA1c,Outcome1HbA1c. . . . . . . . . . . . . . . . . . . . . . 68 Analysis1.2.Comparison1HbA1c,Outcome2Sensitivityanalysis-removingChristian2008. . . . . . . . 69 Analysis1.3.Comparison1HbA1c,Outcome3Sensitivityanalysis-removingLeu2005. . . . . . . . . . 70 Analysis1.4.Comparison1HbA1c,Outcome4Sensitivityanalysis-removingclusterrandomisedtrials. . . . . 71 Analysis1.5.Comparison1HbA1c,Outcome5Sensitivityanalysis-removeGlasgow2003. . . . . . . . . . 72 Analysis1.6.Comparison1HbA1c,Outcome6Subgroupanalysis-outcomesatlessthan6months. . . . . . . 73 Analysis1.7.Comparison1HbA1c,Outcome7Subgroupanalysis-outcomesatlaterthan6months. . . . . . 74 Analysis1.8.Comparison1HbA1c,Outcome8Subgroupanalysis-mobilephonebasedinterventions. . . . . . 75 Analysis1.9.Comparison1HbA1c,Outcome9Subgroupanalysis-interventionsbasedathome. . . . . . . . 76 Analysis2.1.Comparison2Dietarychange,Outcome1Fruitandvegetablescreenerscore. . . . . . . . . . 76 Analysis2.2.Comparison2Dietarychange,Outcome2Estimateddailyfatintake. . . . . . . . . . . . . 77 Analysis2.3.Comparison2Dietarychange,Outcome3Changeincalorificintake. . . . . . . . . . . . . 78 Analysis2.4.Comparison2Dietarychange,Outcome4Pooledeffectondiet. . . . . . . . . . . . . . . 78 Analysis3.1.Comparison3ImpactonweightorBMI,Outcome1PooledeffectonweightorBMI. . . . . . . 79 Analysis4.1.Comparison4Lipids,Outcome1Totalcholesterol. . . . . . . . . . . . . . . . . . . 80 Analysis4.2.Comparison4Lipids,Outcome2Changeintotalcholesterol. . . . . . . . . . . . . . . 81 Analysis4.3.Comparison4Lipids,Outcome3Highdensitylipoprotein(HDL). . . . . . . . . . . . . 81 Analysis4.4.Comparison4Lipids,Outcome4ChangeinHDL. . . . . . . . . . . . . . . . . . . 82 Analysis4.5.Comparison4Lipids,Outcome5Lowdensitylipoprotein(LDL). . . . . . . . . . . . . . 82 Analysis4.6.Comparison4Lipids,Outcome6ChangeinLDL. . . . . . . . . . . . . . . . . . . 83 Analysis4.7.Comparison4Lipids,Outcome7TC:HDLratio. . . . . . . . . . . . . . . . . . . . 83 Analysis4.8.Comparison4Lipids,Outcome8Changeintriglycerides. . . . . . . . . . . . . . . . . 84 Analysis4.9.Comparison4Lipids,Outcome9Pooledeffectoncholesterol. . . . . . . . . . . . . . . 84 ADDITIONALTABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85 APPENDICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90 CONTRIBUTIONSOFAUTHORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146 DECLARATIONSOFINTEREST . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 SOURCESOFSUPPORT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 DIFFERENCESBETWEENPROTOCOLANDREVIEW . . . . . . . . . . . . . . . . . . . . . 147 INDEXTERMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148 Computer-baseddiabetesself-managementinterventionsforadultswithtype2diabetesmellitus(Review) i Copyright©2013TheCochraneCollaboration.PublishedbyJohnWiley&Sons,Ltd. [InterventionReview] Computer-based diabetes self-management interventions for adults with type 2 diabetes mellitus KingshukPal1,SophieVEastwood1,SusanMichie1,AndrewJFarmer2,MariaLBarnard3,RichardPeacock4,BindieWood1,JoniD Inniss1,ElizabethMurray1 1ResearchDepartmentofPrimaryCareandPopulationHealth,UniversityCollegeLondon,London,UK.2DepartmentofPrimary CareHealthSciences,UniversityofOxford,Oxford,UK.3DepartmentofDiabetes,TheWhittingtonHospitalNHSTrust,London, UK.4ArchwayHealthcareLibrary,London,UK Contact address: Kingshuk Pal,ResearchDepartmentof Primary Care and Population Health,University College London, Upper Floor3,RoyalFreeHospital,RowlandHillStreet,London,NW3PF,[email protected]@gmail.com. Editorialgroup:CochraneMetabolicandEndocrineDisordersGroup. Publicationstatusanddate:New,publishedinIssue3,2013. Reviewcontentassessedasup-to-date: 14November2011. Citation: PalK,EastwoodSV,MichieS,FarmerAJ,BarnardML,PeacockR,WoodB,InnissJD,MurrayE.Computer-baseddiabetes self-managementinterventions foradultswithtype2diabetesmellitus.CochraneDatabaseofSystematicReviews2013, Issue3.Art. No.:CD008776.DOI:10.1002/14651858.CD008776.pub2. Copyright©2013TheCochraneCollaboration.PublishedbyJohnWiley&Sons,Ltd. ABSTRACT Background Diabetes is one of the commonest chronic medical conditions, affecting around 347 million adults worldwide. Structured patient educationprogrammesreducetheriskofdiabetes-relatedcomplicationsfour-fold.Internet-basedself-managementprogrammeshave beenshowntobeeffectiveforanumberoflong-termconditions,butitisunclearwhat aretheessentialoreffectivecomponentsofsuch programmes.Ifcomputer-basedself-managementinterventionsimproveoutcomesintype2diabetes,theycouldpotentiallyprovidea cost-effectiveoptionforreducingtheburdensplacedonpatientsandhealthcaresystemsbythislong-termcondition. Objectives Toassesstheeffectsonhealthstatusandhealth-relatedqualityoflifeofcomputer-baseddiabetesself-managementinterventionsfor adultswithtype2diabetesmellitus. Searchmethods Wesearchedsixelectronicbibliographic databases forpublishedarticlesandconferenceproceedingsandthreeonline databases for theses(alluptoNovember2011).Referencelistsofrelevantreportsandreviewswerealsoscreened. Selectioncriteria Randomisedcontrolledtrialsofcomputer-basedself-managementinterventionsforadultswithtype2diabetes,i.e.computer-based softwareapplicationsthatrespondtouserinputandaimtogeneratetailoredcontenttoimproveoneormoreself-managementdomains throughfeedback,tailoredadvice,reinforcementandrewards,patientdecisionsupport,goalsettingorreminders. Datacollectionandanalysis Tworeviewauthorsindependentlyscreenedtheabstractsandextracteddata.Ataxonomyforbehaviourchangetechniqueswasusedto describetheactiveingredientsoftheintervention. Computer-baseddiabetesself-managementinterventionsforadultswithtype2diabetesmellitus(Review) 1 Copyright©2013TheCochraneCollaboration.PublishedbyJohnWiley&Sons,Ltd. Mainresults Weidentified16randomisedcontrolledtrialswith3578participantsthatfittedourinclusioncriteria.Thesestudiesincludedawide spectrumofinterventionscoveringclinic-basedbriefinterventions, Internet-basedinterventionsthatcouldbeusedfromhomeand mobilephone-basedinterventions.Themeanageofparticipantswasbetween46to67yearsoldandmeantimesincediagnosiswas6to 13years.Thedurationoftheinterventionsvariedbetween1to12months.Therewerethreereporteddeathsoutof3578participants. Computer-baseddiabetesself-managementinterventionscurrentlyhavelimitedeffectiveness.Theyappeartohavesmallbenefitson glycaemiccontrol(pooledeffectonglycosylatedhaemoglobinA1c(HbA1c):-2.3mmol/molor-0.2%(95%confidenceinterval(CI) -0.4to-0.1;P=0.009;2637participants;11trials).TheeffectsizeonHbA1cwaslargerinthemobilephonesubgroup(subgroup analysis: mean difference in HbA1c -5.5 mmol/mol or -0.5% (95% CI -0.7 to -0.3); P < 0.00001; 280 participants; three trials). Currentinterventionsdonotshowadequateevidenceforimprovingdepression,health-relatedqualityoflifeorweight.Four(outof 10)interventionsshowedbeneficialeffectsonlipidprofile. Oneparticipantwithdrewbecauseofanxietybuttherewerenootherdocumentedadverseeffects.Twostudiesprovidedlimitedcost- effectivenessdata-withonestudysuggestingcostsperpatientoflessthan$140(in1997)or105EUROandanotherstudyshowed nochangeinhealthbehaviourandresourceutilisation. Authors’conclusions Computer-baseddiabetesself-managementinterventionstomanagetype2diabetesappeartohaveasmallbeneficialeffectonblood glucose control and the effectwas larger in themobile phone subgroup. There is no evidence to show benefits inother biological outcomesoranycognitive,behaviouraloremotionaloutcomes. PLAIN LANGUAGE SUMMARY Usingcomputerstoself-managetype2diabetes Diabetesisoneofthecommonestlong-termmedicalconditions,affectingaround347millionadultsworldwide.Around90%ofthem havetype2diabetesandareatsignificant riskofdevelopingdiabetesrelatedcomplicationssuchasstrokesorheartattacks.Patient educationprogrammescanreducetheriskofdiabetes-relatedcomplications,butmanypeoplewithtype2diabeteshaveneverattended structurededucationprogrammestolearnhowtolookafterthemselves(self-management).Betteruseofcomputersmightbeoneway ofhelpingmorepeoplelearnaboutself-management. We identified 16 trialsinvolving 3578 adults thatmetour criteria. These studies included differenttypesof interventions usedin differentplacesliketouchscreencomputersinhospitalclinics,computersconnectedtotheInternetathomeandprogrammesthat communicatedwithmobilephones.Theaverageageofpeopletakingpartwasbetween46to67yearsoldandmostofthosepeople hadlivedwithdiabetesfor6to13years.Participantsweregivenaccesstotheinterventions for1to12months,dependingonthe intervention.Threeoutofthe3578participantsdiedbutthesedeathsdidnotappeartobelinkedtothetrials. Overall,thereisevidencethatcomputerprogrammeshaveasmallbeneficialeffectonbloodsugarcontrol-theestimatedimprovement inglycosylatedhaemoglobinA1c(HbA1c-along-termmeasurementofmetaboliccontrol)was2.3mmol/molor0.2%.Thiswas slightlyhigherwhenwelookedatstudiesthatusedmobilephonestodelivertheirintervention-theestimatedimprovementinHbA1c was5.5mmol/molor0.5%inthestudiesthatusedmobilephones.Someoftheprogrammesloweredcholesterolslightly.Noneofthe programmeshelpedwithweightlossorcopingwithdepression. Oneparticipantwithdrewbecauseofanxietybuttherewerenoobvioussideeffectsandhypoglycaemicepisodeswerenotreportedin anyofthestudies.Therewasverylittleinformationaboutcostsorvalueformoney. Insummary,existingcomputerprogrammestohelpadultsself-managetype2diabetesappeartohaveasmallpositiveeffectonblood sugarcontrolandthemobilephoneinterventionsappearedtohavelargereffects.Thereisnoevidencetoshowthatcurrentprogrammes canhelpwithweightloss,depressionorimprovinghealth-relatedqualityoflifebuttheydoappeartobesafe. Computer-baseddiabetesself-managementinterventionsforadultswithtype2diabetesmellitus(Review) 2 Copyright©2013TheCochraneCollaboration.PublishedbyJohnWiley&Sons,Ltd. SUMMARY OF FINDINGS FOR THE MAIN COMPARISON [Explanation] Computer-baseddiabetesself-managementinterventionsforadultswithtype2diabetesmellitus Patientorpopulation:participantswithtype2diabetesmellitus Interventions settings: clinic-based (touch screen or other clinic computer), home computer-based and mobile phone-based interventions Intervention:computer-basedsoftwareapplicationsthatrespondtouserinputandaimtogeneratetailoredcontenttoimproveoneor moreofthecognitive,behaviourandskillsandemotionalself-managementdomainsthroughfeedback,tailoredadvice,reinforcement andrewards,patientdecisionsupport,goalsettingorreminders Comparison:standarddiabetescare,non-interactivecomputer-basedprogrammes,papereducationalmaterial,delayedstart/waiting list,face-to-facediabetesself-managementeducation Outcomes Relativeeffect Noofparticipants Qualityoftheevidence Comments (95%CI) (studies) (GRADE) Health-relatedqualityof Seecomment 2113 ⊕⊕⊕(cid:13) No study showed statis- life (5) moderatea tically significant differ- [follow-up: 2 to 18 ences between interven- months] tionandcontrolgroups Deathfromanycause Seecomment 3578 ⊕⊕⊕⊕ Atotal ofthreedeathsin [follow-up: 2 to 18 (16) high the16studies. Twopar- months] ticipantsdiedinonestudy (Lorig2010)andonepar- ticipant died in another studyfromcomplications of a cerebrovascular at- tack (Leu2005). No fur- therdetailswereprovided inthestudyreports. Depression Seecomment 2273 ⊕⊕⊕(cid:13) No study showed sta- [follow-up: 2 to 18 (6) moderateb tistically significant dif- months] ferences in depression scores or incidence of depression between in- tervention and control groups Adverseeffects Seecomment 3578 ⊕⊕⊕⊕ Onestudyreportedapar- [follow-up: 2 to 12 (16) high ticipant withdrawing due months] to anxiety related to the study HbA1c[%] 1.-0.2(-0.4to-0.1) 1.2673 1.⊕⊕⊕(cid:13) 1.Computer-basedinter- [follow-up: 2.-0.5(-0.7to-0.3) (11) moderatec ventions resulted in a 0. 1.2to12months 2.280 2.⊕⊕(cid:13)(cid:13) 2%greaterHbA1creduc- 2.3to12months] (3) lowd tion than control groups (differenceinchangeand Computer-baseddiabetesself-managementinterventionsforadultswithtype2diabetesmellitus(Review) 3 Copyright©2013TheCochraneCollaboration.PublishedbyJohnWiley&Sons,Ltd. finalvalues) 2. Subgroup mobile phone interventions re- sulted in a 0.5% greater HbA1c reduction than control groups (differ- enceinfinalvalues) Economicdata Seecomment 761 ⊕⊕(cid:13)(cid:13) One study looked at [follow-up:18months] (1) lowe health behaviour and re- source utilisation but found no significant dif- ferences between inter- ventionorcontrolgroups GRADEWorkingGroupgradesofevidence Highquality:Furtherresearchisveryunlikelytochangeourconfidenceintheestimateofeffect. Moderatequality:Furtherresearchislikelytohaveanimportantimpactonourconfidenceintheestimateofeffectandmaychange theestimate. Lowquality: Further research isverylikelytohaveanimportant impact onourconfidence intheestimate ofeffect and islikelyto changetheestimate. Verylowquality:Weareveryuncertainabouttheestimate. aSeriousriskofbias bSeriousriskofbias cInconsistency,indirectness dSubgroupanalysis,lownumberofparticipants,indirectness eOnestudyonly,seriousriskofbias mentsinmanagingdiabetesisevidentlyimportant. BACKGROUND Improving bloodsugar control inpatientswithdiabetescanre- Theburdenofdiabetesisgrowingwith347millionpeoplecur- ducetheriskofdeathandmicrovascularcomplications(DCCT rentlyaffectedworldwide(Danaei2011)andnumbersprojectedto 1993;UKPDS1998);however,achieving significant reductions increaseto552millionby2030(InternationalDiabetesFederation inblood glucose levelscanbe difficultinpractice (Peters1996; 2011).IntheUK,thecosttotheNationalhealthService(NHS) Saaddine2002).Cardiovascularriskfactorssuchasraisedblood related to diabetes in 2002 was estimated to be around “£1.3 pressure and lipids are also important and targeting these indi- billionayear,withmostofthiscostarising fromthelong-term viduallyortogethercanbeeffectiveinreducingmortality(Gaede complications resultingfromdiabetesnotbeing managedprop- 2003). There is a growing body of evidence that supports the erly”(Wanless2002),whiletheInternationalDiabetesFederation notion thatimproving self-care improvesthebiological compli- (IDF)suggeststhatinthedevelopedworldthecostofcaringfor cationsofdiabetes,aswellascognitiveandemotionaloutcomes patientswithdiabetesisdoublethatofthebackgroundpopula- (Campbell2003). tion. Complications of diabetes range fromanincreased risk of heart attacks, strokes and amputations to blindness and kidney damage:a60-yearoldmalenewlydiagnosedwithtype2diabetes Diabetesandself-management (withoutpre-existingcardiovasculardisease)canexpecttolose8 to 10 years of life if his diabetes is poorly controlled (National Corbin and Strauss (Corbin 1988) described three distinct ele- CollaboratingCentre2008).Theneedforcost-effectiveimprove- mentsofcopingwithachronicillness. Computer-baseddiabetesself-managementinterventionsforadultswithtype2diabetesmellitus(Review) 4 Copyright©2013TheCochraneCollaboration.PublishedbyJohnWiley&Sons,Ltd. 1. Medicalandbehaviouralmanagement:e.g.taking Newtechnologyofferssomeexcitingnewopportunitiestoexpand medication,attendingfollow-up. onthesuccessaboveandcountersomeofthedifficulties(Griffiths 2. Rolemanagement:e.g.takingonthe’patient’role,the 2006;Tate 2004). Desktop, laptopor handheldcomputersand effectonrelationships. mobilephoneshavetheprocessingpowerandconnectivitytoal- 3. Emotionalmanagement:dealingwiththefear,anger,guilt lowremoteaccesstoinformationandalgorithmsthatmaybeable etc.thatoftenaccompanylivingwithachronicillness. totargetmostofthecomponentsofexistingface-to-faceDSME LorigandHolmanusedtheworkofCorbinandStraussasabasis programmes.Theyalsohavethepotentialtoberelativelycheap, todescribesixskillsrequiredforself-management(Lorig2003): easilydistributable,deliveredatmultiplelocations(clinical,com- 1)problemsolving, 2)decisionmaking,3)resourceutilisation,4) munity-based,athomeoronthemove)attimesconvenientfor theformationofapatient-providerpartnership,5)actionplanning patients,offerpatientsasmanyinterventionsastheyneedorwant andbehaviourchange,and6)patientstailoringmanagementplans andoffercontinuingsupport,sendoutautomaticremindersand tosuittheirneeds. present information in an attractive, tailored format to suit pa- Theseskillsdescribe themedical,behavioural androle-manage- tients’needs.Connectivitymentionedabovealsoallowseasyfor- mentelementsofself-management,buttheabilitytocopewith mationofsocialnetworkingandpeersupportgroupsbeyondtra- theemotionalburdenassociatedwiththeillnessisalsoneeded.Di- ditionalclinicalsettings.Diabetesself-managementinterventions abetesself-managementeducation(DSME)isaformaltermused often show evidence of short-term benefits that may fade over todescribetheongoingprocessoffacilitatingtheknowledge,skill time(Minet2010).Computer-basedinterventionshavethepo- andabilitiesnecessaryfordiabetesself-care(Funnell2009),and tentialtoprovideongoingself-managementsupporttore-enforce newtechnologyhasthepotentialtoimprovepatientoutcomesby thebenefitsovertime. helpingpatientsimprovetheirabilitiesinallofthesedomains. Descriptionofthecondition Theevidencefordiabetesself-management Diabetesmellitusisametabolicdisorderresultingfromadefectin education insulinsecretion,insulinaction,orboth.Aconsequenceofthisis chronichyperglycaemia(thatiselevatedlevelsofplasmaglucose) Anumberofexistingevidence-basedprogrammestoimproveself- with disturbances of carbohydrate, fat and protein metabolism. carearealreadywidelyusedandexamplesofthesediabetesself- Long-term complications of diabetes mellitus include retinopa- management education programmes include: the diabetes edu- thy,nephropathyandneuropathy.Theriskofcardiovasculardis- cation and self-management for ongoing and newly diagnosed ease is increased. For a detailed overview of diabetes mellitus, (DESMOND)programmeforpeoplewithnewlydiagnosedtype pleaseseeunder’Additional information’ intheinformation on 2diabetes(Davies2008),the’RethinkOrganizationtoiMprove theMetabolicandEndocrineDisordersGroupinTheCochrane EducationandOutcomes’(ROMEO)forpeoplewithtype2dia- Library(see’About’,’CochraneReviewGroups(CRGs)’).Foran betes(Trento2010),the’DiabetesX-PERTProgramme’(Deakin explanationofmethodologicalterms,seethemainglossaryinThe 2006)for peoplewith type 2diabetes and the dose adjustment CochraneLibrary. for normal eating (DAFNE) for people with type 1 diabetes (DAFNE 2002). Examples of general patient self-management programmes include the chronic disease self-management pro- gramme (CDSMP) (Lorig 2001) and the expert patient pro- Descriptionoftheintervention gramme (EPP) (Department of Health 2001; Kennedy 2007). Thetermcomputer-baseddiabetesself-managementintervention Group-basedtrainingforself-managementinpeoplewithtype2 includesanyapplicationthattakesinputfromapatientanduses diabetesappearstoimprovediabetescontrol(glycatedhaemoglo- communicationorprocessingtechnologytoprovideatailoredre- binreducedby1.4%atsixmonths)andknowledgeofdiabetesin sponsethatfacilitatesoneormoreaspectofdiabetesself-manage- theshort-andlonger-termwithweakerevidencetoshoweffectson ment,i.e.technologythatpromotes1)problemsolving, 2)deci- bloodpressure,weightandhealth-relatedqualityoflife(Deakin sionmaking,3)resourceutilisation,4)theformationofapatient- 2005). In contrast, theredoes not currently appear to be much providerpartnership,5)actionplanning,emotionalmanagement evidence to show thatindividual patienteducation significantly orbehaviourchangeor6)self-tailoring,withoutneedingcontin- improvesglycaemiccontrol, body mass index or blood pressure uousprofessionalinput. (Duke2009). Adverseeffectsoftheintervention Althoughdifficultieswiththeuptakeandreachofsuchinterven- Thepotentialfornewtechnology tions have beendocumentedin theliterature (Glasgow 2010a), Computer-baseddiabetesself-managementinterventionsforadultswithtype2diabetesmellitus(Review) 5 Copyright©2013TheCochraneCollaboration.PublishedbyJohnWiley&Sons,Ltd. thereiscurrentlynogoodevidencedocumentingadverseeffects behaviourchange;2)theyonlypartiallyexplaintheobservedvari- oftheseinterventions. Possible adverseeffectscouldincludethe anceseeninbehaviouraloutcomes;3)theyareheavilyfocusedon following. motivatedorintentionalbehaviourwhilethebehavioursthatneed • Patientsreceivingincorrectadviceormisinterpretingself- tobechangedtoimprovehealthareoftenautomatic,habitualpat- managementguidance. ternsofbehaviour;and4)theydonotmodelmultiplebehaviour • Patientsmakingdecisionsthatclinicianswoulddeem change(Munro2007). ’inappropriate’. Aninterventionbasedontheoryismorelikelytobeeffectivethan • Frustrationatabsenceofservicestheinterventionsuggests onenotbasedontheory(Noar2008),andalsohastheadvantages wouldbeuseful. ofageneralisableframework,thepossibilityofunderstandingwhy • Senseoffailure,lossofself-esteemorself-worthamongst theinterventionsmightwork,andfacilitatingtheaccumulationof patientswhostopusingitordonotfindithelpful. knowledge(Michie2008).However,thereisnoclearconsensuson • Exclusionarisingfromdigitaldivideorinabilitytouse howtochooseoneormoretheoriesofhealthbehaviourinorder technology. tocreateanintervention.Oneapproachistoassessthebehaviours • Riskofhealthservicesystemonlyprovidinge-health thatneedtobechangedintermsoftheoreticaldomainsexplain- intervention,leavingthoseunabletousesuchinterventions ing why current behaviours exist. Using a mapping framework unserved.Breakdownorstrainonexistingdoctor-patient (Michie2008),thesedomainscanthenbeusedtoselectappro- relationshipsifthereisadifferenceinadvicefromthe priatebehaviourchangetechniques(examplesshowninTable1). interventionandhealthcareproviders. Theuseofintegrativetheoreticaldomainsallowsacomprehensive • Clinicianinformationoverloadfromdatageneratedbyself- theoreticalassessmentratherthanstartingbyapplyingonlyoneor managementrecording. twotheoriesandpotentiallymissingimportantexplanations.Part • Increaseduseandstrainonhealthservicesfrommore oftheaimofthisreviewistodescribetheinterventionsinterms engagedpatients. ofbehaviourchangetechniquesandtheories(aswellasmodeof deliveryortechnologyused)therebyallowingatheory-basedra- tionaleforgroupingorcombininginterventioncomponents. Howtheinterventionmightwork Computer-based interventions to improve diabetes self-care are complexinterventions(MedicalResearchCouncil2008),andthey canbejudgedontheirabilitytoimprovebiological,cognitive,be- haviouralandemotionaloutcomes.Forthistohappen,interven- Whyitisimportanttodothisreview tionsneedtohelppatientsimprovetheirknowledgeandunder- standingofdiabetesandchangetheirpatternsofeating,physical Thereisevidencethatlow-intensitybriefinterventionsforsimple activityandadherencetotreatmentregimens.Thetheorybehind behaviourchange(e.g.smokingcessation)areeffectiveevenwhen theeducationalcomponentofinterventionscanbebasedonprin- deliveredbycomputer-basedapplications(Portnoy2008),while ciples of adult learning and education (Collins 2004) although higher-intensityface-to-faceinterventions(grouporone-to-one) thetheoreticalbasisofdiabeteseducationalinterventionsisoften setthecurrentstandardforself-managementtraining.Thefunda- poorly described (Brown 1999). Knowledge and understanding mentalquestioniswhetherthereisacost-effectivenicheforcom- arethoughttobeimportantcognitionsthatcaninfluencehealth puter-based‘intermediate’interventions(i.e.interventionsthatare behaviourandtheyarecomponentsofsomeofthetheoriesmen- morecostlyandtimeconsumingthanthebriefinterventionmodel tionedbelow. butcheapertoimplementthanface-to-facecontact)forthemore Therearecurrentlyanumber ofdifferenttheoriesthatareused complex(multiple)behaviourchangerequiredinchronicdisease tomodelhealthbehaviourwhichoftenoverlapandmayusedif- self-management,inthisparticularcase,type2diabetesmellitus. ferenttermstodescribe similar concepts (Noar 2005). Some of Therearesomeimportantquestionsaboutcomputer-basedinter- themostcommonlycitedmodelsforhealthbehaviourfocuson ventions for diabetes self-managementthatneedanswering and cognitiveconstructssuchasattitudes,beliefsandexpectations(re- havenot beenfullyanswered by previous reviews in theareaas latedtooutcomes,self-belieforwhatotherpeoplemightthink) therehasbeeninsufficientevidenceinthepast. andexamplesof suchmodelsinclude the’HealthBeliefModel’ 1. Unknownefficacy-dotheyreallywork? (Rosenstock1966),’TheoryofReasonedActionandPlannedBe- 2. Uncertaintyaboutactivecomponents-howdotheywork? haviour’(Ajzen2001),’SocialCognitiveTheory’(Bandura1986) 3. Whatistheclinicalsignificanceofanyreportedbenefitsof and’ProtectionMotivation Theory’(Rogers1975).Usingthese theseinterventions? theoriesasabasisfordesigningself-managementprogrammeshas 4. Aretheycosteffective? a number of challenges: 1) the theories focus on predicting be- 5. Whatharmcancomefromcomputer-basedinterventions? haviour and were not primarily created as theories for enabling 6. Whichpopulationsandsub-populationsdotheybenefit? Computer-baseddiabetesself-managementinterventionsforadultswithtype2diabetesmellitus(Review) 6 Copyright©2013TheCochraneCollaboration.PublishedbyJohnWiley&Sons,Ltd. OBJECTIVES Behaviourandskills • Incorporatingappropriatenutritionalmanagement. Toassesstheeffectsonhealthstatusandhealth-relatedqualityof • Incorporatingphysicalactivityintolifestyle. lifeofcomputer-baseddiabetesself-managementinterventionsfor • Utilisingmedications(ifapplicable)fortherapeutic adultswithtype2diabetesmellitus. effectiveness. • Monitoringbloodglucose,urineketones(when appropriate),andusingtheresultstoimprovecontrol. METHODS • Accessingservicesandpreventing,detectingandtreating acutecomplications. • Preventing(throughriskreductionbehaviour),detecting, andtreatingchroniccomplications. Criteriaforconsideringstudiesforthisreview Emotional Typesofstudies • Integratingpsychosocialadjustmenttodailylife. • Managinganxiety,depressionandstress. Randomisedcontrolledclinicaltrials. • Providingsocialsupportforpatients. Typesofparticipants Control Adultpatientswithtype2diabetesmellitus.Adultpatientswere • Standarddiabetescare. definedaspatientsaged18andover. • Non-interactivecomputer-basedprogramme. • Papereducationalmaterial. • Delayedstart/waitinglist. Diagnosticcriteria • Face-to-facediabetesself-managementeducation. Tobeconsistentwithchangesinclassificationanddiagnosticcri- Thereisno equivalentof aplacebo-controlledtrial asallof the teriaofdiabetesmellitusthroughtheyears,thediagnosisshould controlshavepatientswithachronicillnesswhowillhaveongo- havebeenestablishedusingthestandardcriteriavalidatthetime ingclinicalinputthatcouldhaveaneffectonthetargetpopula- ofthebeginningofthetrial(forexampleADA1999;ADA2008; tion.Somewouldarguethatanyeffectofstandardcareisdesir- WHO 1998). Ideally, diagnostic criteria should have been de- able asitcounteracts theeffectof theselectionbias inherentin scribed.Wherenecessary,authors’definitionofdiabetesmellitus choosingvolunteersfortrialswhoaremorelikelytobemotivated wereused. andconcernedabouttheirhealthanditprovidesamorerealistic estimateofanyadvantage ofthetreatmentoverexistingclinical care.However,inthecaseofbehaviouralinterventions,’standard Typesofinterventions care’mayinvolveanumberofthebehaviouraltechniquesthatare beingtestedandtheirpresenceinthecontrolgroupcouldmake theresultsdifficulttointerpret.Wehavethereforetriedtoanalyse Intervention any ’standard care’ providedto comparison groups and implicit interventionortechniquethatmightbepartofthecomparison Computer-basedsoftwareapplicationsthatrespondtouserinput group(deBruin2009). andaimtogeneratetailoredcontenttoimproveoneormoreofthe following self-management domains through feedback, tailored advice,reinforcementandrewards,patientdecisionsupport,goal Setting settingorreminders. Therewerenorestrictionsbasedonsettingortechnologyusedto delivertheintervention. Cognitive • Knowledgeaboutthediabetesdiseaseprocess, Exclusions complicationsandtreatmentoptions. Anyprogram,websiteorapplication. • Goalsettingtopromotehealth. • Targetedonlyatpatientswithtype1diabetes • Self-efficacyandconfidenceinownabilitytomanage • Involvingparticipantsagedundertheageof18(including diabetes. studiesonmixedpopulationsofadultsandchildren) Computer-baseddiabetesself-managementinterventionsforadultswithtype2diabetesmellitus(Review) 7 Copyright©2013TheCochraneCollaboration.PublishedbyJohnWiley&Sons,Ltd. • Usedonlyforcommunicationbetweenpatientsand Typesofoutcomemeasures professionals • Targetedexclusivelyathealthprofessionals Studiescarriedoutonmixedpopulationsofpatientswithtype1 Integrating all the factors contributing to diabetes self-manage- andtype2diabeteswereincludedinthereviewaslongasmore ment into a unified model to describe how they might affect than50%ofthepatientshadtype2diabetes.Wherepossible,data outcomesischallenging.Adeliberatelysimplifiedschematicthat forpatientswithtype2diabeteswereextractedandthedataforpa- couldaidthisprocessisshowninFigure1.Asmanyofthehealth tientswithtype1diabetesdiscarded.Whenthatwasnotpossible, outcomes take many years to develop, it is not practical to use dataforthemixedpopulationwereused.Four studiesincluded themasprimaryoutcomemeasuresforthisreviewasfollow-upin inthereviewhadmixedpopulations(Leu2005;Lo1996;Smith thestudieswouldnotbelongenoughtodemonstratedifferences 2000;Wise1986).Oneofthesestudies(Leu2005)providedsuf- inthese.However,moreproximalvariablessuchasglycosylated ficientdatatoincludeitinthemeta-analysis.Asubgroupanalysis haemoglobin A1c (HbA1c, representing an average measure of was undertaken to examine the impact of removing studies on bloodsugars),bodymassindex(BMI),depressionoranxietymay mixedpopulations. showchangesoversuitabletimescales. Figure1. Amodeltodemonstratehowself-managementinterventionsmightaffectoutcomesintype2 diabetes Primaryoutcomes Cognitions • Health-relatedqualityoflife • Changeinknowledgeandunderstanding • Deathfromanycause • Self-efficacy • HbA1c Behaviours Secondaryoutcomes • Physicalactivity Computer-baseddiabetesself-managementinterventionsforadultswithtype2diabetesmellitus(Review) 8 Copyright©2013TheCochraneCollaboration.PublishedbyJohnWiley&Sons,Ltd.