Education

A framework for planning and critiquing medication compliance and persistence research using prospective study designs

Description
A framework for planning and critiquing medication compliance and persistence research using prospective study designs
Categories
Published
of 15
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  Clinical Therapeutics/Volume 31,Number 2,2009 AFrameworkforPlanning and CritiquingMedicationCompliance and PersistenceResearchUsingProspectiveStudyDesigns FemidaH.Gwadry-Sridhar,PhDl;ElizabethManias,RN,PhD2;YingZhang,PhD3;Anuja Roy, PhD4;KristinaYu-Isenberg,PhD5;Dyfrig A. Hughes,PhD6; and Michael B. Nichol,PhD7 1 Departments of Medicine and Physiology & Pharmacology, University of Western Ontario, London, Ontario, Canada;2S c hool of Nursing and Social Work, Faculty of Medicine, Dentistry & Health Sciences, University of Melbourne, Victoria, Australia; 3Department of Mathematics and Statistics,Acadia University, Wolfville, NovaScotia,Canada; 4Health Outcomes & Analytics, Walgreens Health Services, Deerfield, Illinois; SRegional Outcomes Research, Ortho-McNeilJanssen Scientific Affairs, Titusville, New Jersey; 6Centre for Economics and Policy in Health,Bangor University, Bangor, Wales, United Kingdom; and 7S c hool of Pharmacy, University of Southern California, Los Angeles, California ABSTRACT Background:Medicationcomplianceandpersistenceareimportantdeterminantsofclinicaloutcomes.Withtheapplicationofevidence-basedtherapy,it is increasinglyimportanttoensurethatstudiesthat use compliance or persistenceasaprimary or secondaryoutcomearedesignedsuitablyandemployappropriateanalysestosupporttheinferencesmade.Objective:Theaimofthisworkwastodescribethedesignsofmedicationcompliance/persistencestudiesandprovideguidanceonappropriateanalyses,withtheultimategoalofhelpinghealthprovidersandpayers of healthcareunderstandtheimpactofcomplianceandpersistenceonhealthoutcomes.Methods:MEDLINE,CINAHL,EMBASE,andall EBM Reviewsdatabasesweresearchedtolocatekeyresearcharticlesaboutprospectivemedicationcomplianceandpersistencestudies.Articlespublishedbetween1978and2008wereincludedinthesearch.Inclusioncriteriaincludedafocusonmedicationcomplianceandpersistence,andprospectiveresearchdesigns.Articlesthatlargelyfocusedonretrospectivestudydesignsorwerebasedonopinionratherthanevidencewereexcluded.Results:Asystematicframeworkwasdevelopedthatcomprisedaprospectivechecklistandaquantitativetooltoassessthequalityofstudies.Thekeyelementsofthechecklistincludedthefollowing:titleandabstract,introductionorbackground,objectives,methodsandstudydesign,statisticalanalysisandresults, February 2009 discussion,conclusions,anddisclosureofconflictsofinterest.Foreachelement,examplesareprovidedtohelpreadersmakeaninformeddecisionaboutthedesign,value,andquality of aparticularprospectivestudy.Conclusions:Thechecklistandquantitativetoolcan be usedtoprovideobjectivevalidationoftherigorofprospectiveresearchdesigns.  t is anticipatedthatfutureresearchwillfollowauniformapproachtopresentationandevaluationofdata,therebyfacilitatingaclearunderstandingoftheimpactofcomplianceandpersistenceonhealthoutcomes.  ClinTher.2009; 31:421-435)©2009ExcerptaMedicaInc. Key words:compliance,persistence,prospectivedesign. INTRODUCTION Therearemanyreasonsforresearcherstousecomplianceorpersistenceasprimaryorsecondaryoutcomesofprospectivestudies.Thesereasonsrangefromanassessmentoftheefficacyofacompliance-enhancinginterventiontoassessingtheassociationbetweencertainvariablesandpersistence.Commontomanyprospectivestudies is thattheyarelimitedintheir Accepted for publicationDecember 22, 2008. doi:1 0.1 016/j.c1inthera.2009.02.0210149-2918/$-seefront matter ©2009ExcerptaMedicaInc. All rightsreserved. 421  ClinicalTherapeutics abilitytoallowresearcherstomakeinferencesaboutheterogeneityinstudydesign;heterogeneityofmeasuresusedtoassesscompliance;andinadequaciesinthemethodsusedfordataanalysis. I- 3Together,theselimitationsmakeitdifficultfortheresultsoftrialsofcompliance-enhancinginterventions,forinstance,tobegeneralizedtoandimplementedinroutinepractice.Thispaperbuildsonapreviouschecklistthatwasdevelopedfortheuseofretrospectivedatabasesforcomplianceandpersistenceresearch. 4 Thepapercanbeusedtoassistresearchersinthedesignofprospectivestudiesandintheappraisalofpublishedarticles by providingaframeworkfortheassessmentoftheconsistencyandqualityofstudies of complianceandpersistence.Thefocus of thisdiscussion is ontheprospectiveassessmentofinterventionsaimed at promotingmedicationcompliance,buttheconceptsapplyequallytootherprospectivestudies of complianceandpersistence. NegativeConsequences of Noncompliance Noncompliancecontinuestobeanissueofimportancetoclinicaltherapeuticsinspiteofanumber of behavioraltheoriesandmodelsthathavebeendevelopedtounderstandwhypeopledonot comply.5-17 TheInternationalSocietyforPharmacoeconomicsandOutcomesResearch (ISPOR) DefinitionsWorkingGrouprecommendtheuseof compliance and persistence asthepreferredtermstodescribemedicationuse. 18 Onepragmaticwaytoconceptualize compliance is theextenttowhichaperson'sbehavior,relevanttolifestyleandmedicationrecommendations,coincideswithmedicalorhealthadvice. 18-21 On theotherhand,persistenceinvolvesthedurationoftimefrominitiationtodiscontinuationoftherapy.18Althoughtherearedifficultiesinmeasuringcompliance,poorcompliance(ie, :;;50 ) is typicallyobservedin 30% to 50% ofallpatients,irrespectiveofdisease,prognosis,orsetting. 22 Withsurprisingfrequency,patientseitherconsciouslyorunintentionallyfailtoadheretomedicaladviceandcomplywithdosinginstructions. 23 Approximatelyonethirdofallpatientsdonot fill theirprescription,or fill itbutdonottakethemedication.2 4-26 Inclinicaltrials,ithasbeenwellestablishedthatoutcomesarebetterforpatientswhocomplywiththerapycomparedwiththosewhodonot. 27  28 Arigorousmeta-analysisexaminedtherelationshipbetweencomplianceandoutcomesacrossavarietyofdisorders;itfoundthecombinedrelative   riskforadverseoutcomesinnoncompliantpatientswas 5.4(95% CI, 1.6-7.4) forhypertensivepatients, 2.8(95% CI, 1.7-4.7) fordyslipidemicpatients,and 1.5(95% CI, 0.9-2.4) forpatientswithheartdisease. 29  3o Our aimwastohelpthereadersanddesignersofprospectivestudiesthatincludecompliance or persistenceasaprimaryorsecondaryoutcomebyprovidingguidanceaboutthedesign,analysis,andcriticalappraisal of thesestudies,andaboutassessingtheeffectiveness of compliance-enhancinginterventions.Achecklistwasdevelopedasatoolfordesigningprospectiveanalysesandcriticallyevaluatingkeyelementsofcomplianceandpersistencestudiesaspart of thepeer-reviewprocess.Thisarticle is acollectiveeffort of themembers of theISPORMedicationComplianceandPersistenceSpecialInterestGroup(AnalyticMethodsWorkingGroup).Themembersofthisworkinggroupconsist of researchers,academics,andclinicalpractitionerswitharecord of publicationandinterestincomplianceandpersistenceresearch. METHODS TheEnglish-languageliteraturewassearchedviatheCINAHL,MEDLINE,EMBASE,andall EBM ReviewsdatabasesusingMeSHkeywordsrelevanttopatientcomplianceandinterventiontolocatearticleswithcomplianceasaprimaryorsecondaryoutcomethatwouldprovidekeycommentariesandresearchfindingsabouttheconduct of prospectivecomplianceandpersistencestudies.Searchtermsincluded compliance,adherence,patientcompliance,patientadherence,persistence, and patientpersistence. Yearscoveredbythesearchincluded 1978 through 2008 toensurethatallrelevantseminalpublicationsandcurrentresearchstrategieswerelocated.Inclusioncriteriaforpapersincludedthosethatfocusedonmedicationcomplianceandpersistenceandprospectiveresearchdesigns.Articlesthatlargelyfocusedonretrospectivestudydesignsorwerebasedonopinionratherthanevidencewereexcludedfromconsideration. RESULTS Key Components to Critical ppraisal of aManuscript We usedasystematicframeworkforthis review. Theframeworkincludesaprospectivechecklist(Table I) andaquantitativetooltoassessthequalityofpro- Volume   Number    Table I Checklist of key elements of prospectivestudies of medicationcompliance and persistence.ElementTitle/abstractIntroduction/backgroundObjectivesMethodsandstudydesignStatisticalanalysisandresultsDiscussion and conclusionsDisclosure of conflicts of interest spectivestudies(Table II).3 Bothtablesshouldbeusedtogethertoaidinthedesignandevaluationofprospectivecompliancestudies.Thekeyelementsforconsiderationarediscussedunderthefollowingheadings:titleandabstract,introduction or background,objectives,methodsandstudydesign,statisticalanaly sis andresults,discussion,conclusions,anddisclosureofconflicts of interest.Foreachelement,wereviewedtheliteratureandprovidedguidelinestohelpreadersmakeaninformeddecisionaboutthedesign,value,andquality of aparticularprospectivestudy. We alsoproposethattheresearcherrefertotheappendixintheretrospectivechecklistandtospecificarticlesrelevanttocriticallyappraisingstudiesbasedontheirdesign(randomizedcontrolledtrials [RCTs] orobservationalcohortstudies).32 Description of ChecklistItems TitleandAbstract Thetitle of thearticlemustprovideaccurateinformationaboutthetypeofstudyemployed(eg,RCT),thepopulationbeingexamined,andtheprimaryandsecondaryoutcomesbeingmeasured. An abstractcanbepresentedinastructuredformatorasanunstructureddescriptivesummary,dependingontherequirementsofparticularjournals.Regardless of thetypeofformatused,theabstractshoulddetailtheaimofthestudy,thestudydesign,statistical or otherformsofanalysesundertaken,andcomplianceorpersistenceresultsusingnumericaldata.Theconclusionorimplications of theresearchshouldmirrorthestudyresultsandnotoveremphasizethestudy'simportance.Forexample,acleartitlewouldbe EducationalInterven- February 2009 F.H. Gwadry-Sridhar et at. tionstoImproveMedicationComplianceinPatientswithHeartFailure. Avaguetitlewouldbe PharmaceuticalCareResearchandEducationProject:PatientOutcomes. 33Althoughthecontentofthepapermaybeexcellent,thereaderhaslittleinformationabouttheoutcomesorthedisease. Introductionor Background Theintroductionorbackgroundofaprospectivestudyshouldpresentareviewoftheliterature.Particularemphasisshouldbeplacedongapsinpastworkandhowthecurrentstudyintendstoaddressthesegaps.Thescope,severity,andnatureofthecomplianceorpersistenceproblembeingconsidered,preferablyinrelationtoaparticulardisease,shouldbeexaminedwithappropriatereferencetoclinical,healtheconomic,andpolicyliterature. An exampleofanarticlewithaclearintroduction,objective,andmethodssection is one by Vrijensetal. 34 Anotherexcellentexample is anarticle by Cramer. 35 We alsorefertoourpaperinthe AmericanHeartJournal, inwhichwefollowedtheConsolidatedStandardsofReportingTrials(CONSORT)guidelinesandusedanorganizedframeworktopresentthecontext,methodology,andresults. 36 37 Objectives Objectivesrefertothestatementsthattheresearchwillattempttoaddress.Forinstance,theymayrelatetotheefficacyofaparticularinterventionpertainingtocomplianceorpersistence.Objectivesshouldprovideclearlystatedinformationwithidentificationoftheprimaryoutcomeandanysecondaryoutcomesofinterest.Ahypothesis is amorespecificstatementthananobjectiveand is amenabletostatisticalevaluationandtesting.InthecaseofanRCT,anullhypothesisshouldbegiven,statingthattheintervention of interestwillnotproduceastatisticaldifferenceintheprimaryoutcomemeasure.Inaddition,theresearchermustgive an explicitdefinition of compliance or persistence asusedinthestudy.38Apoorlywordedobjective is onethat is vague,suchas Toevaluatetheeffectivenessofinterventions to improveadherence to antihypertensives. Abetterwaytoframetheobjective is achievedbyaddingsomedetail,suchas Toevaluatetheeffectofbehavioralinterventions on medicationcomplianceamongpatientstreatedwith 3 differentclassesofantihypertensives. 423  ClinicalTherapeuticsTable II. Scoringtoolforassessing the quality of studydesign and data analysis in prospectivestudies of medicationcompliance and persistence.*RandomizedControlledQualityCriterionTrialWasapowercalculationperformedapriori to determinesamplesize?Weretheinclusion/exclusioncriteriaforpatientsclearlydefined?Wasthemethod of randomization or allocationreported? If randomizationwasreported,wasanappropriatemethod of randomizationused?Werethepatientsblinded to therandomizationgroup?Werethe outcome assessorsblinded to thetreatmentreceived?Werethereinterventionandusual-caregroupsassigned?Wasthetherapeuticregimenfortheusual-caregroupexplicitlydefined?Wasthetherapeuticregimenfortheinterventiongroupexplicitlydefined?Wasthereidentification of whethercompliancewasaprimary or secondaryoutcome?Wasthereanexplicitly stated compliancemeasure?Wasthechoice of compliancemeasurejustified?Werethecriteriausedformeasuringoutcomesobjective?Werethecomplianceresultsreportedaccording to themeasure that wasselected?Werethereexplicitly statedoutcome measures and reportedresults?Wasthefollow-upperiod of sufficientlengthforthediseasegroupbeingstudied?Wasanappropriatestatistical data analysiscarriedout? If thiswasalongitudinalstudy,weretemporalrelationsconsidered? If thestudyproducedanegativeresult,wereconfidenceintervals or posthocpowercalculationsperformed?Doyouknowhowmanypatientswereexcludedfromthetrial(notenrolledforlogisticalreasons,refusedconsent,noteligible)?Wasthereadescription of withdrawalsanddropouts?Wasatableprovidedforthewithdrawalsanddropouts?Scoreobtained, in relation to maximumpossiblescore /22 Observational Cohort Study /18 Quasi-ExperimentalStudy /19 *A score of 1for yes or 0for no is assigned to eachcriterionforaparticularresearchstudy.Shaded areas indicatequestions that are not applicable tothat studydesign.AdaptedfromElliottetal.3 1 424 Volume 31 Number 2  MethodsandStudyDesign ThemaincategoriesofprospectivestudydesignscoveredinthispaperincludeRCTs,descriptivedesigns,quasi-experimentaldesigns,andobservationaldesigns.Regardlessofthedesignofastudy,themethodsshould be clearlyidentifiedandalignedwiththeobjectives. ReT An experimentaldesign is consideredthegoldstandardforevaluatingtheeffectiveness of interventionsorfortreatmentoutcomesresearch.Usually,thearmsareparallelandpatientsarefollowedforapredeterminedperiodoftime. 36 RCTs canhaveacrossoverdesigninwhichpatientsaremovedtotheotherarm,usuallyfollowingawashoutperiod; RCTs mayalsohavearepeated-measuresdesign,inwhichtime is regardedasafactoronmultiplelevelsinahierarchicalapproach,withunitsassubplots. 39 Compliancecan be aprimary or secondaryoutcome,dependingonhowthestudy is powered.GuidelinesthatassessthequalityofRCTdesignandreportingofresultsareavailablethroughtheCONSORTguidelines. 37 ,40-42  f theintervention is nonpharmacologicornonmedicinalinnature,theinterventionshould be reliableandvalidatedwheneverpossible.Forexample,ifaneducationalapproach is used,itshould be standardizedsothatitsdelivery is reproducible by otherresearchers.Unfortunately,manyarticlesrefertopatienteducationasaninterventiontoimprovecompliance,persistence,orboth,butdonotprovideanydetailaboutthetypeofeducationprovidedinthestudy,suchaswhetheritwasstandardized(eg,scripted).Thislackofdetailmakesitverydifficultforotherresearcherstoreproduceinterventionsthatmaywork.Suchalackoftransparencyhampersprogressintranslatingresearchintopractice. 43 RandomizationMethodsandRestrictions.Variousmethodscontrolforunequaldistributionbetween2armsandconfoundingstratification. 3,44,45 Forexample,astudyprovidinganeducationalinterventiontohospitalizedpatientswithheartfailure(HF)stratifiedtheparticipantsinto2groups;inthefirstgroup,HFwastheprimaryreasonforadmission,andinthesecond,HFwasacoincidentalreasonforadmission. 36 ThisstratificationwasdonetocontrolfordifferentialrisksassociatedwiththerelativeriskofpatientswithHFadmissionsthatneededto be controlledbeforerandomization. February2009F.H.Gwadry-Sridhar et a1. InterventionsAssignedtoEachGroup.Themethodologyforrandomizationandthedesignofthestudyshouldfollowevidence-baseduserguides. 1,46 Theinterventionorusualcaredeliveredineachgroupmust be clearlydescribed. In particular,thebackgroundofthestaff(eg,nurse,specialtyofphysician)andthetypeofeducationdeliveredinbothgroups(eg,booklets,videos,scriptedverbalinstruction)must be discussed.Thedescription of botharms is importantbecauseanonsignificantresultcould be theresultofactivitiesbetweentheinterventionandcontrolgroupsbeingequallyeffective.Cluster-RandomizedTrial.Thisdesignlendsitselftothedeliveryofcomplexinterventionsthatmay be deliveredonmultipleoccasions.Acluster-randomizedtrial 46 is defined by theunit,orcluster, by whichpatientsarerandomized.Often,inpopulation-basedclustertrials,wholecommunitiesmay be randomized;however,incomplianceresearch, we tendtorandom ize atthe level ofintervention,suchasindividualpharmacies,hospitals, or physicianclinics. An example of aclustertrial is oneinwhichpharmaciesarerandomizedtodeliveraninterventionorusualcare.Patientsreceivetheinterventionorusualcare,dependingonthepharmacytheyvisit. 36  f thepharmacyhasbeenrandomizedtoaninterventionarm,thenthepatientsreceiveaninterventionthroughtheirpharmacy.Theclusteringapproachminimizescontamination,becausepharmacistsmayotherwise be inclinedtodeliveraninterventiontoallpatientsregardlessofwhetherthepatient is ininterventionorcontrol.Thedesignmayalso be morecost-effectivefromatrainingperspectivebecauseonlyhalfofpharmaciesneedtrainingforinterventiondelivery.Researchersneedtoaddresssamplesizeissuesverycarefullybecausethepowercalculationsareaffected by thenumberofpatientsexpectedineachenvironmentalcluster;theexpecteddifferenceincomplianceandotheroutcomesmust be considered. 47 Descriptive Design Adescriptivedesignprovidesgeneralinformationaboutactivitiesrelatingtocompliance.However,itdoesnotprovidetherigorousevidenceneededtoevaluatetheefficacyofinterventions. 48,49 Adescriptivedesignusuallyinvolvesareportofacase or a se ries of cases.Examplesmayincludeinterviewswithconsumers or healthcareprofessionalsinrelationto 425
Search
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks