BMJ2013;346:e8501doi:10.1136/bmj.e8501(Published9January2013) Page1of14 Research RESEARCH Effectiveness of screening and brief alcohol intervention in primary care (SIPS trial): pragmatic cluster randomised controlled trial OPENACCESS EileenKanerprofessorofpublichealthresearch1,MartinBlandprofessorofhealthstatistics2,Paul Cassidygeneralpractitioner3,SimonCoultonprofessorofhealthservicesresearch4,VeronicaDale trial statistician2, Paolo Deluca trial manager5, Eilish Gilvarry consultant psychiatrist and honorary professorofaddictionpsychiatry6,ChristineGodfreyemeritusprofessorofhealtheconomics2,Nick Heather emeritus professor of drug and alcohol studies7, Judy Myles consultant psychiatrist8, DorothyNewbury-Birchlecturerinpublichealth1,AdenekanOyefesohonoraryreaderandconsultant clinical psychologist8, Steve Parrott health economics2, Katherine Perryman PhD student9, Tom Phillips clinical doctoral research fellow10, Jonathan Shepherd professor of oral and maxillofacial surgery11, Colin Drummond professor of addiction psychiatry5 1InstituteofHealthandSociety,NewcastleUniversity,NewcastleNE24AX,UK;2DepartmentofHealthSciences,UniversityofYork,UK;3Teams FamilyPractice,Gateshead,UK;4CentreforHealthServiceStudies,UniversityofKent,Canterbury,UK;5NationalAddictionCentre,Instituteof Psychiatry,King’sCollegeLondon,UK;6Northumberland,TyneandWearNHSFoundationTrust,Newcastle,UK;7DepartmentofPsychology, FacultyofHealthandLifeSciences,NorthumbriaUniversity,UK;8DivisionofPublicHealthScienceandEducation,StGeorge’sUniversityofLondon, UK;9GreaterManchesterCLAHRCPractitionertheme,UniversityofManchester,UK;10HumberMentalHealthandTeachingNHSTrust,Willerby, UK;11ViolenceResearchGroup,CardiffUniversity,UK Abstract lifestylecounselling.Deliveryofthepatientleafletandbriefadvice ObjectiveToevaluatetheeffectivenessofdifferentbriefintervention occurreddirectlyafterscreeningandbrieflifestylecounsellingina strategiesatreducinghazardousorharmfuldrinkinginprimarycare. subsequentconsultation. Thehypothesiswasthatmoreintensiveinterventionwouldresultina MainoutcomemeasuresTheprimaryoutcomewaspatients’self greaterreductioninhazardousorharmfuldrinking. reportedhazardousorharmfuldrinkingstatusasmeasuredbythealcohol DesignPragmaticclusterrandomisedcontrolledtrial. usedisordersidentificationtest(AUDIT)atsixmonths.AnegativeAUDIT result(score<8)indicatednon-hazardousornon-harmfuldrinking. SettingPrimarycarepracticesinthenortheastandsoutheastof SecondaryoutcomeswereanegativeAUDITresultat12months, EnglandandinLondon. experienceofalcoholrelatedproblems(alcoholproblemsquestionnaire), Participants3562patientsaged18ormoreroutinelypresentingin healthutility(EQ-5D),serviceutilisation,andpatients’motivationto primarycare,ofwhom2991(84.0%)wereeligibletoenterthetrial:900 changedrinkingbehaviour(readinesstochange)asmeasuredbya (30.1%)screenedpositiveforhazardousorharmfuldrinkingand756 modifiedreadinessruler. (84.0%)receivedabriefintervention.Thesamplewaspredominantly ResultsPatientfollow-uprateswere83%atsixmonths(n=644)and male(62%)andwhite(92%),and34%werecurrentsmokers. 79%at12months(n=617).Atbothtimepointsanintentiontotreat InterventionsPracticeswererandomisedtothreeinterventions,each analysisfoundnosignificantdifferencesinAUDITnegativestatus ofwhichbuiltonthepreviousone:apatientinformationleafletcontrol betweenthethreeinterventions.Comparedwiththepatientinformation group,fiveminutesofstructuredbriefadvice,and20minutesofbrief Correspondence to: E Kaner [email protected] Extra material supplied by the author (see http://www.bmj.com/content/346/bmj.e8501?tab=related#webextra) PaT plot of procedures Results of interaction tests Mean alcohol problems questionnaire (APQ) score by intervention group No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe:http://www.bmj.com/subscribe BMJ2013;346:e8501doi:10.1136/bmj.e8501(Published9January2013) Page2of14 RESEARCH leafletgroup,theoddsratioofhavinganegativeAUDITresultforbrief screeningtest19oramodifiedsinglealcoholscreening advicewas0.85(95%confidenceinterval0.52to1.39)andforbrief questionnaire),20andthreeinterventions(patientinformation lifestylecounsellingwas0.78(0.48to1.25).Aperprotocolanalysis leaflet,fiveminutesofbriefadvice,and20minutesofbrief confirmedthesefindings. lifestylecounselling).Targetedscreeningfocusedon ConclusionsAllpatientsreceivedsimplefeedbackontheirscreening presentationslinkedtomentalhealth,gastrointestinalproblems, hypertension,minorinjuries,andnewpatientregistrations.The outcome.Beyondthisinput,however,evidencethatbriefadviceorbrief screeningresultsarereportedindetailelsewhere.21Thispaper lifestylecounsellingprovidedimportantadditionalbenefitinreducing focusesonpatientleveloutcomesafterbriefalcoholintervention hazardousorharmfuldrinkingcomparedwiththepatientinformation ratherthantrialprocessmeasures. leafletwaslacking. TrialregistrationCurrentControlledTrialsISRCTN06145674. Settings Introduction Weinitiallyrecruited24practicesacrossLondonandthesouth eastandnortheastofEngland,plusfivestandbysitesincase Internationalstudieshaveshownthat20-30%ofpatientswho ofdropouts.Allpracticesdeliveredthefullrangeofgeneral routinelypresentinprimarycarearehazardousorharmful drinkers.1Hazardousdrinkingisarepeatedpatternofdrinking medicalservicesandcoveredarangeofurbanandruralareas, thatincreasestheriskofphysicalorpsychologicalproblems,2 sociallydeprivedandaffluentcommunities,andculturally diversepopulations.Recruitmentspanned15monthsfromMay whereasharmfuldrinkingisdefinedbythepresenceofthese problems.3Severalmeta-analyseshaveshownthatscreening 2008toJuly2009.Fifteenpracticescompletedtheirtarget recruitmentof31patientsbutninepractices(sixinLondonand usingshortquestionnairesfollowedbybriefintervention threeinthenortheast)onlypartiallyrecruitedthistarget(three (comprisingsimpleadviceorpsychologicalcounselling) patientinformationleaflet,onebriefadvice,andfivebrief significantlyreducesalcoholconsumptioninprimarycare populations.4-8ACochraneCollaborationsystematicreviewof lifestylecounselling).Thuswesubsequentlyutilisedthefive standbypractices.Inaddition,fivepracticesthathadcompleted 29primarycaretrialsreportedthatbriefinterventioninpatients thefullrecruitmentagreedtobereallocatedtoamoreintensive wasassociatedwithastatisticallysignificantreducedalcohol interventionthantheiroriginalallocation.Thusweincluded34 consumptionof38gaweek(95%confidenceinterval23gto practiceclustersinthefinalanalysis. 54g)atoneyearcomparedwithcontrolstypicallyreceiving assessmentonly,treatmentasusual,orwritteninformation.8 Inclusion and exclusion criteria GiventhatastandarddrinkunitintheUnitedKingdomcontains 8gofethanol,9thisisequivalenttoareductionof4or5units Weincludedpatientswhoscreenedpositiveforanalcoholuse aweek.Nevertheless,briefalcoholinterventionisrarely disorderandwhowerealertandoriented,aged18ormore, deliveredinpractice10andthisgapinimplementationhasbeen residentwithin20milesofthepractice,andabletounderstand ascribedtoseverallimitationsinthecurrentevidence. Englishsufficientlytocompletestudyquestionnaires.We Briefinterventiontrialshavebeencriticisedforbeingefficacy excludedpatientsalreadyinvolvedinanalcoholresearchstudy studies(optimisinginternalvalidity)ratherthanpragmatic orwhowerespecificallyseekinghelpforalcoholproblemsand trials.11Mosttrialshavealsofocusedonmiddleagedmale thosewhowereseverelyinjuredorunwell,hadaseriousmental drinkers,withothergroupsunder-represented.812Moreover,the healthproblem,weregrosslyintoxicated,orhadnofixedabode. optimalintensityofbriefinterventioniscurrentlyunclear.One study8reportednosignificantadditionalbenefitoflonger Randomisation interventionscomparedwithbriefones.However,although Asecureremoterandomisationservicecarriedout threesystematicreviewsfoundabenefitofmotivational randomisation.Twentyfourallocationswereinitiallygenerated interviewingcomparedwithnoinput(controls)inreducing foreachofthepossiblefactorialcombinationsofscreening alcoholconsumption,theirconclusionsonitsimpactcompared approach,screeningtool,andintervention.Randomisationwas withotheractiveinterventions,includinggivingadvice, stratifiedbygeographicalarea(northversussouth).Thestandby differed.13-15Thusthepresenttrialoccurredinacontextwhere andreallocatedpracticesweresubsequentlyrandomlyallocated briefinterventionwaswidelyregardedasbeingeffectiveat inasimilarmanner. reducinghazardousandharmfuldrinking.Nevertheless,the necessarylengthormodalityinputofbriefinterventionhasyet Consent tobedetermined.Wecarriedoutapragmatictrialofthe effectivenessofthreedifferentbriefinterventionstrategiesat Primarycarestaffinitiallyestablishedverbalconsentwith reducinghazardousandharmfuldrinkinginprimarycare.This patientstocheckeligibilityforthetrial.Atthisstagethey studywasoneofthreetrialsintheScreeningandIntervention collectedbasicpersonalinformationandscreenedthepatient ProgrammeforSensibledrinking(SIPS)study.Thetwolinked usingthefastalcoholscreeningtest19ormodifiedsinglealcohol trialswerebasedinemergencydepartments16andprobation screeningquestionnaire.20Patientswhoscreenedpositivewere offices.17 theninvitedtoprovidewrittenconsentforthetrial.All consentingpatientsenteredtheinterventionpartofthetrial. Methods Interventions Detailsofthistrialprotocolhavebeenpublished.18Thetrial Immediatelyafterthescreeningprocess(inthesame incorporatedclusterrandomisationofpracticestoavoidtherisk consultation),trialparticipantsreceivedashortassessmentof ofcontaminationbetweenthetrialarms.Theoveralltrialhad theirdrinkingbehaviour,andbriefinterventionwasdelivered a2×2×2×3nestedfactorialdesignencompassinggeographical accordingtotheinterventiontowhichthepracticehadbeen area(northeastandsoutheastofEnglandandLondon), randomised.Allinterventionmaterialsareavailablefromthe screeningapproach(universalscreeningofallpatientsversus studywebsite(www.sips.iop.kcl.ac.uk/pil.php). targetedscreeningfocusedonpresentationswherealcoholwas mostlikelytoberelevant),screeningtool(eitherthefastalcohol No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe:http://www.bmj.com/subscribe BMJ2013;346:e8501doi:10.1136/bmj.e8501(Published9January2013) Page3of14 RESEARCH Patientinformationleaflet—Patientsreceivedsimplefeedback providedtoclarifythesequenceandtimingofscreening, ontheirscreeningoutcomeandapatientinformationleafletthat assessment,briefintervention,andfollow-upactivity(see hadbeendevelopedbytheDepartmentofHealthinEngland supplementarydiagram1,PaTplot).Seventypercentoftarget calledHowmuchistoomuch?(www.sips.iop.kcl.ac.uk/pil.php). patientrecruitmentoccurredwithinsixmonths,withthe This16pageleafletdescribestheeffectsofalcoholonhealth remaining30%overninemonths.Owingtothisslow andwellbeingandshowsthenumberofunitscontainedin recruitment,researchstaffwhohaddeliveredtraininginstudy popularalcoholicdrinkstohelpthereaderunderstandhowmuch proceduressupportedscreeningandbriefinterventiondelivery theyaredrinking.Thebackpageliststwointernethelpsites in10practicesandrecruited152patients,whichwas5%ofthe plusdetailsofanationalDrinklinenumber.Theonly totalnumberoftrialparticipants. modificationtothisleafletwastheadditionofanadhesivelabel onthebackpagewithcontactdetailsoflocalalcoholtreatment Measures agenciesrelevanttothepracticesetting. Baseline Briefadvice—Patientsreceivedfeedbackonscreeningandthe Theprimaryoutcomewasdrinkingstatusatsixmonthsas patientinformationleafletplusfiveminutesofstructuredbrief measuredbythealcoholusedisordersidentificationtest advicefrompracticestaffbasedontheHowmuchistoomuch? (AUDIT,scorerange0-40),25whichhasbeenvalidatedforuse briefinterventionprogramme.22Inadditiontoprovidingspecific asanoutcomemeasureinprimarycare.26Ascoreof≥8onthis detailsaboutthehealthandsocialconsequencesofhazardous testindicateshazardousorharmfuldrinkingorthelikelihood andharmfuldrinking,patientswereshownasexspecificgraph, ofdependentdrinking,withasensitivityof92%andspecificity whichindicatedthattheirdrinkingexceededthatofmostofthe of94%.25Sincebriefinterventionsaimtoreducehazardousor population,andalistofbenefitsthatwouldresultfromreduced harmfuldrinking,theprimaryoutcomemeasurewasnegative drinking.Thereafterpatientsweretakenthroughamenuof statusonthetest(proportionofparticipantsscoring<8)atsix techniquestohelpreducedrinkingandaskedtoconsidera months. personaltargetforanachievablereductionindrinking. ParticipantsalsocompletedtheEuroQoltomeasurequalityof Brieflifestylecounselling—Patientsreceivedfeedbackon life,27ashortserviceusequestionnaire,28andamodified screeningplusthepatientinformationleafletandstructured readinessrulerwithonequestion(andfourresponsecategories) briefadviceintheinitialconsultation.Theywerethenaskedto tomeasurepatients’motivationor“readinesstochange”their makeanappointmentforafollow-upconsultationwithintwo drinkingbehaviour.29Theuseofsuchreadinesstochange weeksfora20minutesessionofbrieflifestylecounselling.22 measureshasbeenrecommendedinclinicalpracticetohelp Thecounsellingwasbasedonacondensedformofmotivational tailorbriefinterventionsandalsopredictsubsequentdrinking interviewingcalledhealthbehaviourchange.23Thepatientsfirst outcomes.30 describedtheirtypicaldrinkingdayandthenratedthe importanceofchangingtheirdrinkingandtheirconfidence Follow-up aboutchangingtheirdrinkingona10pointscale(whereahigher numberindicatedgreaterimportanceorconfidenceandvice Atsixand12monthsafterrandomisationresearcherswhowere versa).Thepractitionerthenworkedwiththeseratingsto blindedtotheallocatedinterventioncontactedtheparticipants establishwhytheywereatthecurrentlevelandhowtheymight bytelephoneorpost.Researchersadministeredthesame beincreasedtoahigherpoint,beforeelicitingbothprosand instrumentsasatbaselineplusthealcoholproblems consofdrinkingandfinallyworkingthroughasixstepplanto questionnaire.31Ashortpatientsatisfactionquestionnaire32was helpreducedrinkinglevels. alsoadministeredat12months. Training and support Financial incentives Implementationoftheprotocolinthispragmatictrialvaried Eachpracticereceived£3000tocoverstafftimeanddisruption. acrosspracticesaccordingtotheirsize,levelofinterest,and Paymentswerestagedandoccurredaftertraining,oncompletion availableresources.Practicebasedtrainingwasdeliveredto ofparticipantrecruitment,andattheendofdatacollection. 195people(includingallparticipatingclinicians)indesignated ModelledonsmokingcessationpaymentsintheQualityand teammeetings.Thetrainingincludedgeneralinformationon OutcomesFramework,33screeningandbriefinterventionwas alcoholepidemiologyandUKstandarddrinks,anintroduction incentivisedas£1perpatientscreened(€1.26or$1.85,all tothetrialprotocol,demonstrationoftherelevantscreening conversionsat2008exchangerates),£8perbriefadvice,and andbriefinterventionapproaches,androleplaytopractise £32perbrieflifestylecounselling.Eachpatientparticipant delivery.Researchassociatesdeliveredtraininginthepatient receiveda£10vouchershortlyaftercompletingthebaseline informationleafletandbriefadvicegroup.Analcoholhealth assessmentandateachfollow-upinterview. workerwhowasexperiencedinalcoholcounsellingdelivered thetraininginbrieflifestylecounselling.Inthisgroupbrief Sample size calculation lifestylecounsellingwaspractisedwithtrainedactorsandtape Acomprehensivemeta-analysis4suggestedaclinicallyimportant recorded.Thecompetencyofcounsellingwascheckedusing differenceinnegativestatusonAUDITbetweenbrief thebehaviourchangecounsellingindex.24Onlypractitioners interventionandcontrolsof13%(5%reductionincontrolsand whoreachedarequiredstandard(agreedbyinter-raterconsensus 18%inbriefinterventionrecipients).Detectingthisdifference betweenthreeindependentclinicalassessors)wereapproved atthe5%significancelevelwith80%power(withatwosided todeliverbrieflifestylecounsellinginthetrial. test),required109patientspergroup(total327).Assuminga losstofollow-upof25%,thesamplesizewasinflatedto145 Implementation pergroup(total435).Toaccountforpotentialclustereffects Amixtureofgeneralpractitionersandpracticecarenurses weusedanintraclasscorrelationcoefficientof0.04basedon delivered95%ofscreeningandbriefinterventionactivityin ourexperienceofprimarycaretrials.34Assumingaclustersize thistrial.Agraphicalsummaryofthetrialprocedureshasbeen of31patientsperpractice,thisinflatedthecalculationbya No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe:http://www.bmj.com/subscribe BMJ2013;346:e8501doi:10.1136/bmj.e8501(Published9January2013) Page4of14 RESEARCH factorof1.7.Hencewerequired248patientspergroup(total patientsreportedneveroronlysometimesthinkingabout 744),withanexpectationthatatleast558wouldbefollowed drinkingless. upatsixmonths. Primary outcome Statistical analysis TheproportionsofpatientswithanegativeAUDITstatus Analysiswasbyintentiontotreat,withpatientsanalysedinthe increasedatsixmonthsinallthreeinterventions(fig2⇓).The grouptowhichtheyhadbeenrandomised.Weusedlogistic differencesbetweentheinterventionswerenot,however, regressiontoanalysetheprimaryoutcome,negativestatuson significant(table2⇓).Noneoftheinteractionstestedwere AUDITatsixmonths.Includedinthemodelwerescreening significant(seesupplementarytableS1)sothemodelwithout approach,screeningtool,andintervention,andthemodelwas interactionswasusedtoestimatethedifferencesbetween adjustedforage,sex,andbaselineAUDITscore.Totake interventions.TheoddsratiosofhavinganegativeAUDIT accountoftheclusterstudydesignweusedthe statusforbriefadvicecomparedwiththepatientinformation Huber-White-sandwichestimator,robuststandarderrorsmethod. leafletwas0.85(95%confidenceinterval0.52to1.39)andfor Asafactorialstudy,interactioneffectswereconsidered.Tothe brieflifestylecounsellingcomparedwiththepatientinformation modelweindividuallyaddedthreetwo-wayinteractions, leafletwas0.78(0.48to1.25).Theprimaryoutcomewasnot screeningapproach*tool,screeningapproach*intervention,and affectedbymissingdata(table3⇓). screeningtool*interventionandcomparedtheresultwitha modelwithoutinteractions.Iftheinteractionwasfoundtobe Secondary analyses significantweincludeditinthefinalmodel. At12monthstherewerenostatisticallysignificantdifferences Datawerepresentedasoddsratiosandcorresponding95% betweenthethreeinterventionsintheproportionsofpatients confidenceintervals.Wecarriedoutsecondaryanalysesusing withanegativeAUDITresult(table2).Comparedwiththe appropriatemethodsfortheoutcomes(linear,logistic,orordinal patientinformationleafletintervention,at12monthstheodds regression)controllingforthesamecovariatesastheprimary ratioofhavinganegativeAUDITresultwas0.91(0.53to1.56) outcomeandadjustingfortheclusterednatureofthestudy. forbriefadviceand0.99(0.60to1.62)forbrieflifestyle Interventionefficacywasalsoexploredinaperprotocol counselling.Aperprotocolanalysis,includingjustthosewho analysis. receivedtheirallocatedtreatment,andananalysiscombining Toassesstheimpactofmissingdataontheprimaryoutcome, themoreintensiveinterventions(briefadviceplusbrieflifestyle wecarriedoutmultipleimputationusingtheICEprocedurein counsellingversuspatientinformationleaflet)alsoindicated STATA.Wetestedseveralmodelsfortheprediction;thefinal nosignificantdifferencesbetweentheinterventionsatsixor12 modelusedAUDITnegativestatusatsixmonthsasthe months. dependentvariable,baselineAUDITscore,screeningapproach, Inaddition,therewerenostatisticallysignificantdifferencesin screeningtool,intervention,sex,andage.Aseriesof10 meanAUDITscorebyinterventionorovertime(table4⇓).At imputationswasdone.WecombinedtheseusingtheMIcombine sixmonths,themeandifferencebetweenbriefadviceandthe commandinSTATA.Theimputationconfirmedtheinitial patientinformationleafletwas0.06(−0.70to0.83)andbetween analysiswhereonlythoseactuallyfollowedupwereincluded. brieflifestylecounsellingandthepatientinformationleaflet was−0.38(−1.51to0.75).At12monthsthesemeandifferences Results werelargerbutnotstatisticallysignificant.Comparedwiththe patientinformationleaflet,themeandifferenceforbriefadvice Of3562presentingpatients,2991(84.0%)wereeligiblefor was−0.20(−0.83to0.43)andforbrieflifestylecounselling screening;900(30.1%)ofthesepatientswereidentifiedas was−0.25(−1.19to0.68).Theestimateswerederivedfrom hazardousorharmfuldrinkers.Overall,756(84.0%)consented modelswithoutinteractions. toparticipateinthetrial;consentratesweresimilarbetween Atsixmonthsthereweredifferencesinreportedreadinessto thethreeinterventions(fig1⇓).Allparticipantsreceiveda change(table5⇓),with32%(n=65)ofpatientsinthepatient patientinformationleaflet,whereas99%(n=250)ofthose informationleafletgroupreporting“tryingtocutdown” allocatedtotheothertwointerventionsreceivedbriefadvice. comparedwith34%(n=69)receivingbriefadviceand45% However,just57%(n=143)ofrelevantpatientsreturnedand (n=93)receivingbrieflifestylecounselling.Theexpected receivedthebrieflifestylecounsellingintervention. orderedoddsforbrieflifestylecounsellingcomparedwiththe Atsixmonthsthefollow-uprateswere85%(patientinformation patientinformationleafletincreasedby1.74(95%confidence leaflet85%(n=212),briefadvice86%(n=215),andbrief interval1.27to2.39,P=0.001)withashifttothenexthigher lifestylecounselling85%(n=217))andat12months82% category—thatis,agreaterreadinesstochange.Forbriefadvice (patientinformationleaflet79%(n=197),briefadvice83% comparedwiththepatientinformationleaflet,theexpected (n=209),brieflifestylecounselling83%(n=211)).Follow-up orderedoddsincreasedby1.37(0.95to1.98,P=0.095).A ratesbetweentheinterventionsdidnotdiffersignificantly. similarfindingoccurredat12months,with32%(n=61)ofthose However,thosefollowedupatsixmonthshadlowermean inthepatientinformationleafletgrouptryingtocutdown baselineAUDITscoresthanthosenotfollowedup:12.4(SE comparedwith37%(n=74)receivingbriefadviceand48% 0.25)v14.3(SE0.66). (n=95)receivingbrieflifestylecounselling.Forbrieflifestyle Theaverageageofparticipantswas45years,62%(n=756)of counsellingcomparedwiththepatientinformationleaflet,the participantsweremen,92%(n=755)werewhite,34%(n=253) expectedorderedoddsincreasedby1.86(1.31to2.65,P=0.001). hadattainedhigherdegreelevel,and34%(n=258)weresmokers Forbriefadvicecomparedwiththepatientinformationleaflet, (table1⇓).Atbaseline82%(n=611)ofparticipantswere theexpectedorderedoddsincreasedby1.24(0.83to1.87, identifiedashazardousorharmfuldrinkersbytheAUDIT,with P=0.293). anaveragescoreof12.7(SD6.4).Reportedreadinesstochange Participantswhoreceivedbrieflifestylecounsellingalso variedacrossthethreeinterventions,although62%(n=465)of reportedgreatersatisfactionthanthosewhoreceivedthepatient informationleaflet(table6⇓)basedongeneralcommunication No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe:http://www.bmj.com/subscribe BMJ2013;346:e8501doi:10.1136/bmj.e8501(Published9January2013) Page5of14 RESEARCH (meandifference0.13,95%confidenceinterval0.01to0.26) Intermsofpotentialweaknesses,sincedirectmonitoringofthe andtheinterpersonalmannerofthecliniciandeliveringthe consultationswouldhaveunderminedthepragmaticnatureof intervention(meandifference0.10,0.002to0.19).These thestudy,itwasdifficulttoascertaintheextenttowhichthe differenceswerenotobservedbetweenbriefadviceandthe interventionsweredeliveredasintended.Moreover,trialstaff patientinformationleafletinterventions. deliveredscreeningandbriefinterventiontoasmallproportion ofpatients(5%)owingtoflaggingrecruitment.Itispossible Interaction with earlier screening activity thatthelackofinterventiondifferencesmayhavebeendueto unsuccessfulimplementationofthebriefinterventionprotocols Atsixmonthstherewasasignificantinteractionbetweenbrief bytheprimarycareclinicians.Thesecliniciansoftengiveadvice interventionandearlierscreeningapproach,thereforetheresults andlifestylecounsellinginotherareasofpracticeandthere arepresentedassixseparategroups(seesupplementarytable mayhavebeenunconscioususeoftheseskillswithpatients 2).Eachgroupwascomparedwiththereferencegroupofpatient whowerenotintendedtoreceivetheminthistrial.Theissue informationleaflet/universalscreening.Attheinitialfollow-up ofinterventionfidelitywillbeexploredinanindepthqualitative point,patientsinthebrieflifestylecounselling/universal (interviewbased)processstudywithcliniciansfromthistrial, screeninggroup(meandifference−0.78,95%confidence whichoccurredafterpatientfollow-upwascompleted. interval−1.53to−0.03)andthepatientinformation leaflet/targetedscreeninggroup(meandifference−0.77,−1.42 Only57%ofpatientsinthebrieflifestylecounsellinggroup to−0.12)hadsignificantlylowerscoresonthealcoholproblems actuallyreceivedtheintervention,whichcouldhavereduced questionnaire.Buttheotherfourcombinationsofbrief itspotentialimpact.But,asthiswasapragmaticstudythislevel interventionandscreeningapproachdidnotdiffersignificantly. ofreturntoasubsequentcounsellingsessionprobablyreflects Furthermore,outcomesmeasuredbythealcoholproblems whatwouldhappeninusualpractice.Afurtherissueof questionnaireat12monthsdidnotdiffersignificantly(see pragmatismwasourattempttomakebaselineassessmentas supplementarytableS2). shortaspossible.Wewereawarethatextensivequestioning aboutalcoholcouldhaveactedasaninterventioninitsown Discussion right.Henceweusedthe10itemAUDIT(alcoholusedisorders identificationtest)questionnairetoprovideourprimaryoutcome Allpatientsinthistrialreceivedfeedbackontheirhazardous measure.26Althoughthistooldidnotallowdetailedmeasurement orharmfuldrinkingstatusimmediatelyafterthescreening ofactualconsumptionlevels(intermsofstandarddrinkunits process.Attwofollow-uppoints,however,briefadviceand dailyorweekly),itcapturedwideraspectsofalcoholrelated brieflifestylecounsellingdidnotprovideastatistically riskandharmthroughitemsondrinkingquantity,frequency, significantadditionalbenefitinreducinghazardousorharmful intensity,andnegativebehaviouralconsequences. drinkingcomparedwiththeprovisionofapatientinformation leaflet.Brieflifestylecounsellingsignificantlyincreased Relevance of the findings to the discipline patients’motivationtoreducetheirdrinkingthroughapositive Numeroussystematicreviewsandmeta-analyseshavenow shiftinreadinesstochangecomparedwiththosereceivingthe shownbenefitsofscreeningandbriefalcoholinterventionin patientinformationleaflet.Moreover,thesepatientsalso primarycarecomparedwitharangeofcontrolconditions reportedgreatersatisfactionwiththebriefinterventionprocess (typicallyassessmentonly,treatmentasusual,orwritten thanthoseinthepatientinformationleafletgroup.However, information).However,aconsistentfindinginthisdiscipline nosignificantdifferencesbetweenthebriefinterventionswere issignificantlyreducedalcoholconsumptioninbothintervention foundforalcoholrelatedproblemsorhealthrelatedqualityof andcontrolsgroups.36-38Oneexplanationforreduceddrinking life.Thisstudythereforedoesnotsupporttheadditionaldelivery inbothinterventionandcontrolgroupsisregressiontothemean, offiveminutesofbriefadviceor20minutesofbrieflifestyle inwhichextrememeasuresofbehaviourtendtoshifttoless counsellingoverandabovethedeliveryoffeedbackon extremepositionsfromonetimepointtoanother.39Wecannot screeningplusapatientinformationleaflet. fullydiscountthisexplanationforourfindings.However,we thinkitunlikelysincelevelsofdrinkinginthistrial,asshown Strengths and weaknesses of the study bymeanAUDITscores,werenotparticularlyextreme.Indeed, Thistrialwasalargepragmaticmulticentreevaluationof somepatientswereclosetothethresholdforhazardousdrinking. screeningandbriefinterventionintypicalprimarycare Inaddition,thechangesovertimeinourcontrolconditionwere conditions.Clusterrandomisationavoidedthepotentialproblems similartomeaneffectsizesreportedforbriefintervention.837 ofcontaminationbetweenthetrialarmsandthussubversionof Analternativeexplanationisthatourcontrolcondition, thestudyprotocol.Fivepracticesdeliveredmorethanonebrief consistingofsimplefeedbackandwritteninformationabout interventionbutonlyaftertheyhadsuccessfullycompleted alcohol,mayhavecontainedactivefactorsofbehaviour recruitmentfortheinterventiontowhichtheyhadbeen change.40Indeedtwocontrolledtrialshavereportedsignificant originallyallocated.Thesereallocatedpracticessubsequently clinicaleffectsofscreening41andassessmentalone42ondrinking deliveredamoreintensivebriefinterventionapproachafter behaviour.Consequently,itislikelythatthecumulativeimpact traininginthenewprocedure. ofscreening,assessment,simplefeedback,andthedeliveryof Othermethodologicalstrengthsofthestudyweretheuseof writteninformationmayhaveoverwhelmedtheadditionalinput remoterandomisationproceduresandvalidatedoutcome offiveminutesofstructuredbriefadvicebothintermsof measuresofclinicalrelevance.Ratesofeligibilityandpatient elapsedtimeandthenumberofdistinctbehaviourchange consenttothetrial(84%)werehigherthaninmostprevious techniquesused.43 similarstudies,35whichaddsweighttothegeneralisabilityof Toconclusivelydemonstratethatprovidingfeedbackplusa theresearch.Furthermore,thestudyachieveditsexpected patientinformationleafletledtoreducedratesofhazardousand samplesize,andpatientfollow-uprateswerehigherthan harmfuldrinking,ourtrialwouldideallyhavehadanadditional anticipated(>80%)withnodifferentiallosstofollow-up nointerventionarm.However,inviewoftheextensiveevidence betweenthethreeinterventions. supportingbriefinterventioneffectivenessatthetimeofthe No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe:http://www.bmj.com/subscribe BMJ2013;346:e8501doi:10.1136/bmj.e8501(Published9January2013) Page6of14 RESEARCH trial,ourcontrolconditionwasconsideredtobetheminimal Contributors:Alloftheauthorscontributedtothedesignand ethicallyacceptableinputafterscreeningactivity. developmentofthistrialprotocol.CDwasthechiefinvestigatorofSIPS andEKwasdeputychiefinvestigatorandleadfortheprimarycaretrial. Implications for clinicians, service Expertiseonclinicalaspectsoftheresearchwasprovidedforprimary commissioners, and policymakers carebyPCandJM,fornursingpracticebyTPandforpsychiatryCD andEG.StatisticalinputwasprovidedbySC,VDandMB.Health Thehighlevelsofconsenttothistrialandthehighratesof economicsinputwasprovidedbyCGandSP.Trialconductanddelivery screeningandimmediatedeliveryofbriefinterventionindicate expertisewasprovidedbyPD,DNBandKP.Alcoholandpolicyexpertise thatroutinelypresentingpatientsinprimarycarearewillingto wasprovidedbyAOandDS.Briefinterventionexpertisewasprovided receivefeedback,writteninformation,andadviceabouttheir byCD,EK,NHandJS.EKwrotethefirstdraftofthepaperandall drinkingbehaviour.Inaddition,thehighlevelsofpatient authorscontributedtosuccessivedrafts.Allauthorsreadandapproved satisfactionafterbriefinterventioninallthreeconditionssupport thefinalmanuscript. theacceptabilityofthistypeofinput.However,thesignificant patientattritioninthebrieflifestylecounsellinggroupsuggests Funding:EKwaspartfundedbyFusetheCentreforTranslational thattheremaybeabenefitofdeliveringbriefintervention ResearchinPublicHealth.FuseisaUKCRCPublicHealthResearch directlyafterscreeningratherthandelayinguntilasubsequent CentreofExcellence,andfundingcomesfromtheBritishHeart occasion.Inemergencycare,ithasbeenrecommendedthatthe Foundation,CancerResearchUK,EconomicandSocialResearch timebetweenscreeningandbriefinterventionisminimisedand Council,MedicalResearchCouncil,andtheNationalInstituteforHealth ideallyoccursonthesameday.44Indeeditislikelythatthe Research.ThisstudywasfundedbytheDepartmentofHealth.The processofidentifyingandquantifyingalcoholrelatedriskor viewsexpressedhereindonotnecessarilyreflectthoseofthe harmmaybea“teachablemoment”wherepatientshavea DepartmentofHealthortheNationalHealthServiceinEnglandand heightenedreceptivitytotheideaofreducingtheirdrinking.44 Wales. Competinginterests:AllauthorshavecompletedtheICMJEuniform Nevertheless,inpatientswhoreturnedtoasubsequent disclosureformatwww.icmje.org/coi_disclosure.pdf(availableon consultationforbrieflifestylecounselling,thereweresignificant requestfromthecorrespondingauthor)anddeclare:allauthorshad positivechangesinmotivationtoreducedrinkingandinpatient financialsupportfromtheDepartmentofHealthinEngland(Alcohol satisfactionlevels.Consequently,itispossiblethatwhereas PolicyUnit)forthesubmittedwork;nofinancialrelationshipswithany mosthazardousandharmfuldrinkersinprimarycarerequire organisationsthatmighthaveaninterestinthesubmittedworkinthe minimalinputafterscreening,theremaybeagroupofpatients previousthreeyears;nootherrelationshipsoractivitiesthatcouldappear whowouldvalueandbenefitfromadditionalsupport.Hencea tohaveinfluencedthesubmittedwork. steppedcareapproachmightbehelpfulinthisarea.Herethe leastintensive(lesscostly)interventionisusedwithmost Ethicalapproval:Thisstudyreceivedmulticentreethicalapproval patientswhopresentwithalcoholrelatedriskorharm,and (06/MRE02/90)pluslocalagreementfromallrelevantlocalresearch furtherinterventionisreservedforpatientswhodonotrespond ethicscommittees.Researchgovernanceapprovalwasgrantedbyall orwhoaskformoresupporttohelpreducetheirdrinking relevantprimarycaretrusts.Theresearchwasdoneinaccordancewith behaviour.Apreviousstudyfoundthatasteppedcarealcohol theHelsinkideclaration. interventionwasmorecosteffectivethanaminimalintervention Datasharing:Noadditionaldataavailable. inprimarycare.34Thisapproachwasalsorecommendedinrecent NationalInstituteforHealthandClinicalExcellenceguidelines 1 FunkM,WutzkeS,KanerE,AndersonP,PasL,McCormickR,etal.Amulticountry controlledtrialofstrategiestopromotedisseminationandimplementationofbriefalcohol onthepreventionofalcoholproblemsinadultsandyoung interventioninprimaryhealthcare:findingsofaWorldHealthOrganizationcollaborative people.45 study.JStudAlcohol2005;66:379-88. 2 SaundersJB,LeeNK.Hazardousalcoholuse:itsdelineationasasubthresholddisorder, Finally,giventheextensivepublishedevidenceonthe andapproachestoitsdiagnosisandmanagement.ComprPsychiatry2000;41:95-103. effectivenessofscreeningandbriefinterventioninreducing 3 WorldHealthOrganization.Internationalclassificationofdiseases.10threvision.WHO, 1992. hazardousandharmfuldrinking,thecaseforitswider 4 MoyerA,FinneyJW,SwearingenCE,VergunP.Briefinterventionsforalcoholproblems: implementationinprimarycareisstrong.Regardingthe ameta-analyticreviewofcontrolledinvestigationsintreatment-seekingand non-treatment-seekingpopulations.Addiction2002;97:279-92. necessarylevelofinput,ourfindingsconfirmtheconclusion 5 BallesterosJA,DuffyJC,QuerejetaI,ArinoJ,Gonzalez-PintoA.Efficacyofbrief ofaCochraneCollaborationsystematicreviewthatlonger(more interventionsforhazardousdrinkersinprimarycare:systematicreviewandmeta-analysis. intensive)briefinterventionsaddnosignificantadditional Alcohol,ClinExpRes2004;28:608-18. 6 BertholetN,DaeppenJ-B,WietlisbachV,FlemingM,BurnandB.Briefalcoholintervention benefitovershorterinputinprimarycare.8However,thisreview inprimarycare:systematicreviewandmeta-analysis.ArchInternMed2005;165:986-95. containedjustonetrial,basedinFinland,whichdirectly 7 WhitlockEP,PolenMR,GreenCA,OrleansT,KleinJ.Behavioralcounselinginterventions inprimarycaretoreducerisky/harmfulalcoholusebyadults:asummaryoftheevidence comparedthreedifferingintensitiesofbriefintervention. fortheUSPreventiveServicesTaskForce.AnnInternMed2004;140:557-68. Equivalentoutcomeswerereportedforwomen46andmen47who 8 KanerE,BeyerF,DickinsonH,PienaarE,CampbellF,SchlesingerC,etal.Effectiveness ofbriefalcoholinterventionsinprimarycarepopulations.CochraneDatabaseSystRev receivedsimplefeedbackcomparedwiththreeorsevenbrief 2007;(2):CD004148. interventionsessions.However,thisfindingwasattributedto 9 MillerW,HeatherN,HallW.Calculatingstandarddrinkunits:internationalcomparisons. BrJAddict1991;86:43-7. afailuretosuccessfullyimplementthemoreintensivebrief 10 KanerE.Briefalcoholintervention:timefortranslationalresearch.Addiction interventionsinroutinepracticeratherthanthemeritofsimple 2010;105:960-1. feedbackinitself.Thecurrentstudystronglysuggeststhat 11 BaborTF,Higgins-BiddleJC,DauserD,BurlesonJA,ZarkinGA,BrayJ.Briefinterventions forat-riskdrinking:patientoutcomesandcost-effectivenessinmanagedcareorganizations. screeningfollowedbysimplefeedbackandwritteninformation AlcoholAlcohol2006;41:624-31. maybethemostappropriatestrategytoreducehazardousand 12 EdwardsAGK,RollnickS.Outcomestudiesofbriefalcoholinterventioningeneralpractice: theproblemoflostsubjects.Addiction1997;92:1699-704. harmfuldrinkinginprimarycare. 13 BurkeBL,ArkowitzH,MencholaM.Theefficacyofmotivationalinterviewing:a meta-analysisofcontrolledclinicaltrials.JConsultClinPsychol2003;71:843-61. 14 LundahlB,KunzC,BrownellC,TollefsonD,BurkeB.Ameta-analysisofmotivational ThispaperispublishedonbehalfoftheSIPSprogrammeresearch interviewing:twentyfiveyearsofempiricalstudies.ResSocWorkPract2010;20:137-60. group.Afulllistoftheresearchgroupmembersisavailableathttp:// 15 VasilakiE,HosierS,CoxW.Theefficacyofmotivationalinterviewingasabriefintervention forexcessivedrinking:ameta-analyticreview.AlcoholAlcohol2006;41:328-35. sips.iop.kcl.ac.uk/contactus.php.WethankRuthMcGovernandRobert 16 CoultonS,BlandM,CassidyP,DelucaP,DrummondC,GilvarryE,etal.Screeningand Pattonforrefiningthestudyinterventionsandsupportingthe briefinterventionforhazardousandharmfulalcoholuseinaccidentandemergency departments:arandomisedcontrolledtrialprotocol.BMCHealthServRes2009;9(14). implementationofthistrial. No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe:http://www.bmj.com/subscribe BMJ2013;346:e8501doi:10.1136/bmj.e8501(Published9January2013) Page7of14 RESEARCH Whatisalreadyknownonthistopic Around20-30%ofpatientswhoroutinelypresentinprimarycarearehazardousorharmfuldrinkers Briefalcoholinterventioninprimarycarecansignificantlyreducehazardousandharmfuldrinking However,theoptimalintensityofbriefinterventioninputiscurrentlynotclear Whatthisstudyadds Briefadviceandbrieflifestylecounsellingdidnotprovideastatisticallysignificantbenefitinreducinghazardousorharmfuldrinking comparedwithapatientinformationleaflet Screeningfollowedbysimplefeedbackandwritteninformationmaybethemostappropriatestrategytoreducehazardousandharmful drinkinginprimarycare 17 Newbury-BirchD,BlandM,CassidyP,CoultonS,DelucaP,DrummondC,etal.Screening 35 BeichA,ThorsenT,RollnickS.Screeninginbriefinterventiontrialstargetingexcessive andbriefinterventionsforhazardousandharmfulalcoholuseinprobationservices:a drinkersingeneralpractice:systematicreviewandmeta-analysis.BMJ2003;327:536-40. clusterrandomisedcontrolledtrialprotocol.BMCPublicHealth2009;9:418. 36 BernsteinJ,BernsteinE,HeerenT.Mechanismsofchangeincontrolgroupdrinkingin 18 KanerE,BlandM,CassidyP,CoultonS,DelucaP,DrummondC,etal.Screeningand clinicaltrialsofbriefalcoholintervention:implicationsforbiastowardsthenull.Drug briefinterventionsforhazardousandharmfulalcoholuseinprimarycare:acluster AlcoholRev2010;29:498-507. randomisedcontrolledtrialprotocolBMCPublicHealth2009;9:287. 37 JenkinsRJ,McAlaneyJ,McCambridgeJ.Changeovertimeinalcoholconsumptionin 19 HodgsonR,AlwynT,JohnB,ThomB,SmithA.TheFASTalcoholscreeningtest.Alcohol controlgroupsinbriefinterventionstudies:systematicreviewandmeta-regressionstudy. Alcohol2002;37:61-6. DrugAlcoholDepend2009;100:107-14. 20 CanagasabyA,VinsonDC.Screeningforhazardousorharmfuldrinkingusingoneor 38 McCambridgeJ,KypriK.Cansimplyansweringresearchquestionschangebehaviour? twoquantity-frequencyquestions.AlcoholAlcohol2005;40:208-13. systematicreviewandmetaanalysesofbriefalcoholinterventiontrials.PLOSOne 21 CoultonS,DrummondC,DelucaP,KanerE,Newbury-BirchD,PerrymanK,etal.The 2011;6:e23748.doi:10.1371/journal.pone.0023748. utilityofdifferentscreeningmethodstodetecthazardousdrinkingandalcoholusedisorders 39 BlandJ,AltmanD.Regressiontowardsthemean.BMJ1994;308:1499. inthescreeningandinterventionprogrammeforSensibledrinking(SIPS)programme. 40 MichieS,PrestwichA,deBruinM.Importanceofthenatureofcomparisonconditionsfor AddictSciClinPract2012;7(Suppl1):A83. testingtheory-basedinterventions:reply.HealthPsychol2010;29:468-70. 22 HeatherN,GirvanM,KanerE,CassidyP.Implementingscreeningandbriefalcohol 41 McCambridgeJ,DayM.Randomizedcontrolledtrialoftheeffectsofcompletingthe interventioninpilotgeneralpracticesintheTyneandWearHealthActionZone(final AlcoholUseDisordersIdentificationTestquestionnaireonself-reportedhazardousdrinking. report).AfinalreportonresearchfundedbytheTyneandWearHealthActionZone. Addiction2007;103:241-8. NorthumbriaUniversity,2008. 42 KypriK,LangleyJD,SaundersJB,Cashell-SmithML.Assessmentmayconcealtherapeutic 23 RollnickS,MasonP,ButlerC.Healthbehaviourchange:aguideforpractitioners.Churchill benefit:findingsfromarandomisedcontrolledtrialforhazardousdrinking.Addiction Livingstone,1999. 2007;102:62-70. 24 LaneC.TheBehaviourChangeCounsellingIndex(BECCI)Scale.UniversityofWales, 43 AbrahamC,MichieS.Ataxonomyofbehavourchangetechniquesusedininterventions. 2002. HealthPsychol2008;27:379-87. 25 SaundersJB,AaslandOG,BaborTF,DeLaFuenteJR,GrantM.Developmentofthe 44 WilliamsS,BrownA,PattonB,CrawfordM,TouquetR.Thehalf-lifeofthe‘teachable AlcoholUseDisordersIdentificationTest(AUDIT):WHOCollaborativeProjectonEarly moment’foralcoholmisusingpatientsintheemergencydepartment.DrugAlcoholDepend DetectionofPersonswithHarmfulAlcoholConsumption.Addiction1993;88:791-804. 2005;77:205-8. 26 BradleyKA,McDonellMB,BushK,KivlahanDR,DiehrP,FihnSD.TheAUDITalcohol 45 NationalInstituteforHealthandClinicalExcellence.Alcohol-usedisorders—preventing consumptionquestions:reliability,validity,andresponsivenesstochangeinoldermale thedevelopmentofhazardousandharmfuldrinking.2010.http://guidance.nice.org.uk/ primarycarepatients.AlcoholClinExpRes1998;22:1842-9. PH24. 27 RabinR,CharroF.EQ-5D:ameasureofhealthstatusfromtheeuroqolgroup.AnnMed 46 AaltoM,SaksanenR,LaineP,ForsstromR,RaikaaM,KivilutoM,etal.Briefintervention 2001;33:337-43. forfemaleheavydrinkersinroutinegeneralpractice:a3-yearrandomizedcontrolled 28 UKATTResearchTeam.Costeffectivenessoftreatmentforalcoholproblems:findings study.AlcoholClinExpRes2000;24:1680-6. oftherandomisedUKalcoholtreatmenttrial(UKATT).BMJ2005;331:544. 47 AaltoM,SeppaK,MattilaP,MustonenH,RuuthK,HyvarinenH,etal.Briefintervention 29 HeatherN,SmailesD,CassidyP.DevelopmentofaReadinessRulerforusewithalcohol formaleheavydrinkersinroutinegeneralpractice:athreeyearrandomizedcontrolled briefinterventions.DrugAlcoholDepend2008;98:235-40. study.AlcoholAlcohol2001;36:224-30. 30 WilliamsE,HortonN,SametJ,SaitzR.Dobriefmeasuresofreadinesstochangepredict Accepted:7November2012 alcoholconsumptionandconsequencesinprimarycarepatientswithunhealthyalcohol use?AlcoholClinExpRes2007;31:428-35. 31 DrummondC.Therelationshipbetweenalcoholdependenceandalcohol-relatedproblems Citethisas:BMJ2013;346:e8501 inaclinicalpopulation.BrJAddict1990;85:357-66. 32 WareJ,SnyderM,WrightW,eds.Developmentandvalidationofscalestomeasure Thisisanopen-accessarticledistributedunderthetermsoftheCreativeCommons patientsatisfactionwithmedicalcareservices.NationalTechnicalInformationService, AttributionNon-commercialLicense,whichpermitsuse,distribution,andreproductionin 1976. 33 NHSEmployers.ClinicalDirectedEnhancedService(DES)GuidanceforGMScontract anymedium,providedtheoriginalworkisproperlycited,theuseisnoncommercialand 2008/9.NHSEmployersandBMA,2008. isotherwiseincompliancewiththelicense.See:http://creativecommons.org/licenses/by- 34 DrummondC,CoultonS,JamesD,GodfreyC,ParrottS,BaxterJ,etal.Effectiveness nc/2.0/andhttp://creativecommons.org/licenses/by-nc/2.0/legalcode. andcost-effectivenessofasteppedcareinterventionforalcoholusedisordersinprimary care:pilotstudy.BrJPsychiatry2009;195:448-56. No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe:http://www.bmj.com/subscribe BMJ2013;346:e8501doi:10.1136/bmj.e8501(Published9January2013) Page8of14 RESEARCH Tables Table1|Personalandbaselinevariablesbyinterventionallocation.Valuesarenumbers(percentages)unlessstatedotherwise Characteristics Patientinformationleaflet Briefadvice Brieflifestylecounselling Total Percluster: n=14 n=11 n=19 n=44 Median(interquartilerange)Noofpatients 13.5(2-32) 31(3-34) 6(1-47) 12(1-47) Median(interquartilerange)age(years) 47.4(36.0-60.8) 40.2(31.7-53.6) 42.5(32.0-67.7) 43.4(31.7-67.7) Median(interquartilerange)%male 61.8(0-100) 70.0(33-75) 55.3(33-100) 61.8(0-100) Median(interquartilerange)AUDITscore 11.8(8.3-16.1) 13.0(9.4-15.5) 13.0(8.5-24.0) 12.5(8.3-24.0) Patients: n=251 n=251 n=254 n=756 Mean(SD)age(years) n=251,48.2(17.0) n=251,40.4(16.4) n=253,44.9(14.8) n=755,44.5(16.4) Mean(SD)EQ-5D* n=243,0.78(0.27) n=247,0.84(0.24) n=246,0.81(0.26) n=736,0.81(0.26) Mean(SD)AUDITscore n=245,12.3(6.4) n=245,12.6(5.9) n=244,13.1(6.9) n=734,12.7(6.4) Men 163/251(64.9) 155/251(61.8) 152/254(59.8) 470/756(62.2) White 215/251(86) 239/251(95) 238/253(94) 697/755(91.7) Educationafterage16 118/250(47) 159/251(63) 137/251(55) 414/752(55.1) Degreeorequivalentprofessionalqualification 63/249(25) 111/250(44) 79/249(32) 253/748(33.8) Single 60/251(24) 98/250(39) 69/252(27) 227/753(30.1) Smokers 83/251(33) 93/251(37) 82/253(32) 258/755(34.2) Readinesstochange: n=250 n=248 n=247 n=745 Neverthinkaboutdrinkingless 66(26.4) 67(27.0) 64(25.9) 197(26.4) Sometimesthinkaboutdrinkingless 89(35.6) 100(40.3) 79(32.0) 268(36.0) Havedecidedtodrinkless 26(10.4) 34(13.7) 36(14.6) 96(12.9) Alreadytryingtocutdown 69(27.6) 47(19.0) 68(27.5) 184(24.7) AUDIT=alcoholusedisordersidentificationtest. *Measureofhealthutility. No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe:http://www.bmj.com/subscribe BMJ2013;346:e8501doi:10.1136/bmj.e8501(Published9January2013) Page9of14 RESEARCH Table2|Proportionsofparticipantswithnegativealcoholusedisordersidentificationtestresultatbaselineandsixand12monthfollow-up. Valuesarenumbers(percentages)unlessstatedotherwise Oddsratio*(95%CI),Pvalue Brieflifestyle Patient Brieflifestyle Briefadvice/patient counselling/patient Timepoint informationleaflet Briefadvice counselling informationleaflet informationleaflet ICC(SE) Baseline 50/247(20) 51/249(20) 37/249(15) — — 0.02(0.02) 6months 72/202(36) 61/208(29) 59/205(29) 0.85(0.52to1.39),0.51 0.78(0.48to1.25),0.30 0.03(0.02) 12months 74/190(39) 72/205(35) 72/203(36) 0.91(0.53to1.56),0.73 0.99(0.60to1.62),0.96 0.04(0.02) ICC=intraclasscorrelationcoefficient. *Oddsratiofromlogisticregressionmodelsadjustingforscreeningapproach,screeningtool,age,sex,andbaselinealcoholusedisordersidentificationtestscore. No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe:http://www.bmj.com/subscribe BMJ2013;346:e8501doi:10.1136/bmj.e8501(Published9January2013) Page10of14 RESEARCH Table3|Summaryofsensitivityofprimaryoutcomeresultstomissingdata(statusfromalcoholusedisordersidentificationtestatsix months) Oddsratio(95%CI),Pvalue Analysis Briefadvice/patientinformationleaflet Brieflifestylecounselling/patientinformationleaflet Completecase 0.85(0.52to1.39),0.51 0.78(0.48to1.25),0.30 Multipleimputationestimate 0.89(0.53to1.50),0.66 0.78(0.50to1.22),0.28 Perprotocolanalysis 0.85(0.52to1.38),0.51 0.93(0.50to1.72),0.81 No commercial reuse: See rights and reprints http://www.bmj.com/permissions Subscribe:http://www.bmj.com/subscribe