Determining)and)Addressing)Adherence)to)the)NCCN)Guidelines)for)Chronic)Phase)CML!

Transcription

1 Brewer, B Page 1 Determining)and)Addressing)Adherence)to)the)NCCN)Guidelines)for)Chronic)Phase)CML Overall) Goal: Test a novel behavior change and educa4on interven4on to improve physician adherence to the NCCN guidelines and best prac4ce recommenda4ons for chronic;phase chronicmyeloidleukemia(cp;cml)atfouroncologyprac4cesacrosscolorado. Theliteratureontheclinicalprac4ceofphysicianstrea4ngCP;CMLsuggestsseveralmajorgaps between iden4fied best prac4ces and the real world prac4ces of physicians. The most significantgapiden4fiedintheliteratureisthelackofappropriatemonitoringofthebcr;abl1 transloca4on by qpcr in CP;CML. A second major concern iden4fied in the literature is the problemofpa4entadherencetotyrosinekinaseinhibitor(tki)medica4onsover4me.athird concern is a lack of physician knowledge about how to interpret data from BCR;ABL1 when monitoringisconductedcorrectlyandhowtomakethecorrectchoiceoftkimedica4onbased onthisinforma4on. This study will address these three concerns using a novel interac4ve behavior change interven4onforphysiciansandmidlevelprovidersthatwillbeofferedin;persontoeachofthe four prac4ces in the study. The primary outcome measure will be the change in the rate of monitoringbcr;abl1byqpcrinperipheralblood. Key)Objec@ve)1:Establishaone;yearbaselinerateofadherencetoNCCNguidelinesforqPCR monitoring of BCR;ABL1 in the peripheral blood of CP;CML pa4ents treated at four oncology prac4cesacrossthestateofcolorado. Itishypothesizedthatthebaselinerateofcommunityoncologistsandhematologistscorrectly monitoringbcr;abl1byqpcrintheircp;cmlpa4entswillbesubop4mal.thishypothesisis based on the results of several studies that have indicated a very low rate of appropriate assessmentofthisvitalmarkeroftreatmentresponse. Key) Objec@ve) 2: Implement and assess an interac4ve behavior modifica4on and educa4on interven4ondesignedtoimproveprovideradherencetonccnguidelinesforqpcrmonitoring of BCR;ABL1 and increase provider knowledge of treatment guidelines about TKI choice and waysproviderscanaddresstheproblemofpa%entadherencetotkimedica4ons. Asecondhypothesisisthattheinterac4veeduca4onalinterven4on,thatincludesrou4neaudit andfeedback,willsignificantlyimprovephysicianbehaviorinmonitoringbcr;abl1intheircp; CMLpa4ents.TherateofBCR;ABL1tes4ngwillbetrackedoveraninterven4onperiodofone yearandcomparedtothebaselinerateestablishedinkeyobjec4ve1. This interven4on is also hypothesized to create improvement in provider knowledge about monitoring guidelines, evidence based techniques that promote pa4ent adherence to TKI medica4ons,andappropriatetkichoicebasedonmuta4onalanalysisandclinicalsitua4on.this hypothesis will be assessed using a pre;post test at the 4me of the physician training. An addi4onalfollowuppost;testwillbeconductedattheendof6monthstodeterminethelas4ng effectsoftheinterven4onover4me.inaddi4ontothequan4ta4vedatacollected,quan4ta4ve

2 Brewer, B Page 2 datawillbecollectedaboutthebarriersthatprovidersencounterwhenasemp4ngtomeetthe best prac4ce standards and NCCN guidelines described above and will be used to inform problemsolvingstrategiesaspossible. TechnicalApproach2 Assessment)of)Need:) Anincreasingpopula4on ThesuccessofTKImedica4onsintrea4ngCP;CMLnowmeansthattheoverallprevalenceofthe diseaseinthepopula4onwillriseover4me,asthevastmajorityofpeoplewiththediseaseare abletoliveoutanearlynormallifespan.intheunitedstates,althoughtheannualincidenceof CP;CMLremainsstableatapproximately1/100,000people,theprevalenceofCMLises4mated toincreasefromapproximately70,000peoplein2010toaplateauofapproximately181,000by theyear ThisongoingincreaseinthenumberofpeoplelivingwithCP;CMLhighlights the importance of correctly addressing the behavioral factors that op4mize the long term treatmentoutcomesforpa4ents.thetreatmentofcp;cmlwithtki srepresentsthebeginning ofaneraoftargetedmoleculartreatmentsthatrequireappropriatemonitoringofgenemarkers byphysiciansandwillalsorequirepa4entstoadheretoadailymedica4onregimenandmedical follow;up appointments for monitoring. Behavioral modifica4on and educa4on have been showntoeffec4velyaddresstheseconcerns. Summary) The three sec4ons below are summarized briefly here and represent the most salient gaps in prac4cevs.recommendedprac4ceforthecareofpa4entswithcp;cmliden4fiedinana4onal needs assessment and the NCCN guidelines. 1,12 Gap 1: The most significant prac4ce deficit iden4fiedna4onallyforthetreatmentofcp;cmlisthelowrateofpropermonitoringofbcr; ABL1bycommunityoncologists(31%).Monitoringofthismarkerbyperipheralbloodisvitalto assess treatment response and guide treatment decisions. Gap 2: The second major concern iden4fiedisthelackofphysicianandmidlevelprovider sunderstandingofhowtoassessand addresspa4entadherencetotheirtkimedica4ons.gap3:athirdmajordeficitisthephysicians appropriate choice of first;line TKI and a lack of knowledge about when to switch TKI s and whichtkirepresentsthebestchoiceintheeventoftkiresistanceorlackofdiseaseresponse.) Gap)1:Physiciannon;adherencetoNCCNguidelinesformonitoringofBCR;ABL1 TheNCCNguidelinescurrentlyrecommendtes4ngforBCR;ABL1withqPCRinCP;CMLpa4ents everythreemonthsaberini4a4ngtherapy,regardlessoftreatmentresponse. Arecentreport bytheannenbergcenterforhealthsciencesfoundthatonly31%ofcommunityphysiciansand 52% of academic medicine physicians in the U.S. were correctly tracking this vital marker of treatment response in peripheral blood. 1 Most were not adequately using this molecular analysistotracktheirpa4ent sresponsetotkitherapyandmanywereperformingunnecessary bone marrow biopsies to conduct monitoring, usually on a subop4mal 4meline. 1 CP;CML monitoring was also found to be subop4mal in a recent study with 1,200 CML pa4ents. This studyfoundthat41%ofpa4entsonatkididnotreceiveqpcrmonitoringofbcr;abl1within oneyearoftreatmentini4a4on,while31.9%had1;2testsinthatyearand27%had3;4tests. Thisstudyalsocomparedpa4entsinthe notests grouptopa4entsinthe 3;4tests group,

3 Brewer, B Page 3 and found that the laser group had 37% fewer inpa4ent admissions for CP;CML related concerns, sugges4ng that monitoring in accordance with NCCN guidelines for qpcr tes4ng is economicallyandmedicallyuseful. 2 Gap)2:Physicianslackknowledgeofhowtoeffec4velyassessandthenpromotetheadherence oftheircp;cmlpa4entstakingtkimedica4ons.2 Pa4entadherencetooralTKImedica4onisstronglyassociatedwithoveralltreatmentresponse andlikelyremainsoneoftheprimaryfactorseffec4ngthelossofmajormolecularremission (MMR) or lack of response to treatment. 3, 4 The ADAGIO study examined the adherence of pa4ents to ima4nib and compared their reported level of adherence to their actual pill consump4on. 64% of pa4ents reported perfect adherence to their medica4on, however only 14%ofthisgroupactuallyachievedperfectadherence.71%ofpa4entsinthestudywerefound tobetakinglessthantheprescribeddosage. 3 Pooradherencehasasubstan4alimpactontreatmentresponse.Inonestudy,pa4entswithless thanorequalto90%adherencetomedica4onwerefoundtohaveonlya28.4%rateofmmrto treatmentincomparisonwitha94.5%rateforthosewithgreaterthan90%adherence. 4 90% adherenceistheequivalentoftaking27/30dosesina30daymonth. 5 Gap)3:Communityphysicianshavebeenshowntolackinforma4onaboutfirst;lineTKIchoice, the importance of early and deep molecular response when star4ng treatment and when to switchtkimedica4on. CommunityoncologyphysiciansmayencounteronlyafewCP;CMLpa4entsperyearandthus keeping up with current recommenda4ons for TKI selec4on may be challenging. In a recent needs assessment, 62% of oncologists con4nued to use ima4nib as a first line treatment for their CML pa4ents when second;genera4on TKI s such as dasa4nib and nilo4nib have been shown to produce an earlier and deeper molecular molecular response with data and recommenda4ons that show that may be beser tolerated. 1, 6, 7, 8 Providers were also not sufficientlyawarethatdasa4nibandnilo4nibwererecommendedbythenccnguidelinesfor pa4ents with high risk disease and were associated with earlier and deeper molecular remission. 9Thereisalsoevidencethatphysicianslacktheabilitytodifferen4atetheappropriate clinical ac4ons when pa4ents become resistant to a first;line TKI. Many were unaware of the need to conduct a muta4onal analysis and the TKI recommenda4ons for specific muta4ons. Difficulty interpre4ng and appropriately determining the meaning of increases on qpcr monitoringusingtheinterna4onalscalehasbeeniden4fiedasanotherconcern. 1 Primary)Audience)of)the)Interven@on:) Thisinterven4onwillbeprovideddirectlytooncologists,hematologists,midlevelprovidersand othermembersofcareteamsdirectlyinvolvedinpa4entcareortreatmentdecision;makingat four prac4ces across the state of Colorado. The interven4on will be offered for Con4nuing MedicalEduca4on(CME)credit.Thesefourprac4cesrepresentauniqueopportunitytostudy adherencetothenccnrecommenda4onsforcp;cmlastheytreatalargenumberofpa4ents inacommunitysejng,shareacommonemr(epic),andhaveallrecentlybecomeconnected

4 Brewer, B Page 4 totheuniversityofcoloradohealthsystem.eachofthefourprac4ceshaveexpressedsupport andenthusiasmforthisstudy. Direct)beneficiaries)of)this)program:) Thedirectbeneficiariesofthisinterven4onincludethefourprac4cesinvolvedinthisstudyand the CP;CML pa4ents of these prac4ces. The physicians in these prac4ces will benefit from improved knowledge about treatment guidelines and updates about recent developments in the state of science and clinical care for pa4ents with CP;CML. This informa4on will be presented in an interac4ve fashion that makes use of baseline data from the prac4ces actual monitoring adherence to the NCCN guidelines for BCR;ABL1 by qpcr over the past year. This prac4cespecificdataandinterac4veformat(describedbelow)willengagetheirasen4onand servetomo4vatefuturebehaviorchangeintheirprac4ce. Midlevel providers, social workers and nurses at these prac4ces will benefit from knowledge about how to appropriately assess and address the adherence of their pa4ents to TKI medica4ons. Pa4ents with CP;CML who are treated at the prac4ces involved will benefit substan4ally from any improvements in physician knowledge and changes in appropriate monitoringofbcr;abl1givenhowvitaltherou4nemonitoringofthismarkerisforsuccessful treatmentofcp;cml. Interven@on)Design)and)Methods:) Interven4onFactors:Crea4ngLas4ngChangeinProviderBehavior Tradi4onalCMEpresenta4onsdeliveredinanon;interac4velectureformatmaynotbeeffec4ve in changing physician prac4ce behaviors. 10 The literature on changing provider behavior suggests several strategies that have been demonstrated to change the prac4ce behavior of medicalproviders.thisinterven4onhasbeendesignedfromthegrounduptoincorporatethe most effec4ve strategies for changing physician and provider behavior. These factors are summarized below and are integrated into each of the interven4on components described in theparagraphsthatfollow. The most studied effec4ve strategy to change provider behavior is the use of audit and feedback. 11 Audit and feedback involves establishing a feedback loop between physician behaviorandthemetricofdesiredchange.inthisstudy,eachprovider srateofappropriately assessingbcr;abl1byqpcrwillbeprovidedatquarterlyrepor4ngintervals.theresearchon interven4ons to change the behavior of prac4cing physicians and providers also suggests a strong correla4on between presenta4ons that are interac4ve and those that create desired change in provider behavior. 12 The use of interven4ons directed at local prac4ce experts and opinion leaders has also been demonstrated to be an effec4ve strategy of changing provider behavioracrossaprac4ce. 13Theiden4fica4onofbarrierstodesiredchangehasshownmixed resultsintheverylimitedliteraturebutappearstheore4callyjus4fiedandmayprovideuseful informa4on to inform the current study. All of these strategies have been integrated into the interven4on described in detail below. The interven4on addresses the three areas of deficit iden4fiedintheneedsassessmentsec4onabove. TrainingOncologyProviders: Part)1:ChangingproviderbehaviortomatchNCCNguidelinesforBCR;ABL1tes4ngbyqPCR.

5 Brewer, B Page 5 An in;person training will be provided to each of the four prac4ce groups in this study. The training will make use of prac4ce specific informa4on about that prac4ces rate of tes4ng for BCR;ABL1incomparisonwiththeidealrateperNCCNguidelines.Eachphysicianintheprac4ce willbeprovidedwithacardcontainingtheirpersonalrateoftes4ngduringthebaselineperiod incomparisonwiththeidealrateofassessment.therankingoftheprac4ceincomparisonwith theotherprac4cesinthestudywillalsobeshownwitheachprac4ceknowingonlytheirown rateandtheotherprac4ces ratesde;iden4fied:seefigure1. This presenta4on of data about the performance of the individual physician (privately on an indexcard)andtheprac4ceinrela4ontootherprac4cesinthestudyandthena4onalaverage, will promote physician engagement with the presenta4onandmo4vatefuturebehaviorchange. 80% 60% 40% 20% 0% Practice 1 Practice 2 Your Practice Practice 4 National Average: Academic Centers National Average: Community Practices Fig. 1: BCR-ABL1 Testing Rate by Practice - Baseline Baseline Furthermo4va4onforchangewillbesupportedby anitera4veprocessofupdatesontherateofbcr; ABL1 assessment during the interven4on period following the presenta4on. This will be accomplishedbyaquarterly toeachprovider in the study, which will contain both provider; specificandprac4ce;specifictes4ngrates.thistype ofauditandfeedbackinterven4onhasbeenshown to be an effec4ve strategy for changing physician prac4ce. 11 The training will present informa4on about the appropriate interpreta4on of BCR;ABL1 data by qpcr using the Interna4onal Standard. Changes in knowledgebaseofthephysicians(forthisandmany other aspects of informa4on contained in the presenta4on)willbeevaluatedusingapre;posttest with a follow;up post;test at 6;months aber the ini4al presenta4on. Also at this 4me point, a semi; structured qualita4ve interview to iden4fy and addressproviderbarrierstoappropriateassessment will be implemented, along with another update given to the physician and prac4ces about theirrateofbcr;abl1assessment. Interac4vePresenta4on:u4lizingacellphone;basedservicetoallowlivepolling Because interac4ve presenta4ons have been correlated with desired change in physician behaviors,thispresenta4onwillusethepresenta4onsobwaretophatorasimilarservicethat allowsthepresentertopolltheaudiencewithspecificques4ons.audiencemembersrespond using a text;based cell phone message (no smart phone required) and responses are then displayedliveonthescreen.thistypeofinterac4vepresenta4onwillengagetheaudienceand also allow the presenta4on of clinical examples relevant to the informa4on being taught. An example of a ques4on used to poll the audience in this manner would be a mul4ple choice

6 Brewer, B Page 6 ques4on regarding the appropriate frequency of BCR;ABL1 assessment by qpcr that contains answer choices that approximate the correct choice and may also involve the rate of assessmentusingbonemarrowbiopsy(ademonstratedgapinprac4ce). Theseques4onswill be targeted across the presenta4on to address the knowledge and prac4ce gaps iden4fied in theneedsassessment. Part)2:ChangingProvidersInterac4onswiththeirPa4entstoPromotePa4entAdherencetoTKI Medica4on Themosteffec4vestrategyphysicianscanusetopromotetheadherenceoftheirpa4entstoTKI medica4ons is the establishment of an open dialogue and a discussion about adherence to medica4on regimens at2 every2 visit. 14 Research and clinical experience demonstrates that oncologists, mid;level providers and nurses oben lack the needed training and support to implement these sorts of interven4ons. This interac4ve presenta4on u4lizes the same live audiencepollingofques4onsdescribedabove.anexample ofaques4onaskedinalivepollingformatforthispartof thepresenta4oncanbefoundinfigure2. Examples of specific evidence6based techniques taught to theprac4ceswillinclude: The difference between an open and a leading ques4on aboutadherence. The use of simple behavioral recommenda4ons with pa4ents such as: tracking, sejng a 4me to take the medica4on, pujng the medica4on in a standard place andintegra4ngitintoarou4ne The importance of normalizing common difficul4es with adherencesopa4entsareopenabouttheiradherence Adirectdiscussionaboutthefinancialcostsofmedica4on andaddressingfinancialbarriersinthepa4entvisit(oben withhelpfromsocialwork) Fig. 2 Whatisthebestexampleofanopen ques4onaboutadherenceduringa followupvisit? A. So you re taking your imatinib appropriately, right? B. How many pills are you taking each day? C. Are you using a pill box to track your medication? D. So, you re taking two pills a day, how has that been going for you? Ongoingeduca4onanddiscussionabouttheconceptofresistance,thepossibilityofdisease progression and tolerance/repor4ng about side effects along with the idea that TKI medica4onsarenotacure,theysimplymaintainthepa4entwhentheyaretakencorrectly. The importance of direct verifica4on of dose and behavior of pa4ents (pa4ent brings pill bosle to provider visit so that amount of medica4on can be verified as being taken appropriately;withapillcountifneeded) Discussionsabouttheplanningofrefillstoaccommodateout;of;towntravel Involvementandeduca4onoffamilymembersintheadherenceofthepa4ent This training component will also u4lize clinical examples that highlight the complexity of pa4ent adherence to TKI medica4on based on psychological factors associated with the medica4onbeingtheonlyvisiblereminderofdisease.exampleswillbede;iden4fiedbutwillbe based on clinical situa4ons experienced by Dr. Brewer who, as a clinical health psychologist, workswithcp;cmlpa4entstopromotemedica4onadherence. Thisexamplewillbegivento

7 Brewer, B Page 7 promote appropriate referral by providers or nurses to social work or psychology services to addressadherenceincomplexsitua4ons. Part)3:Changingproviderbehaviorinregardtoini4alTKIchoice,theimportanceofearlyand deep molecular remission and when to switch TKI therapy based on the results of BCR;ABL1 monitoringbyqpcr. Content will be matched to that iden4fied in the NCCN Guidelines and the Annenberg report citedintheneedsassessment.specificareasofcontentwillbepresentedandthenwillu4lize interac4veliveaudiencepollingwithspecificclinicalexamplestoassureinterestandasen4on. Specificareasofcontentthatwereiden4fiedintheneedsassessmentinclude: Approvedop4onsandrecommenda4onsforfirst;lineTKIchoicebasedondiseasestatus Theimportanceofearlyanddeepmolecularremission(MMR) HowtointerpretBCR;ABL1byqPCRwhenconduc4ngappropriatethreemonthmonitoring WhentoswitchTKImedica4on Iden4fyingTKIresistanceandwhentoconductamuta4onalanalysisofABL What muta4ons in ABL may mean and which TKI medica4ons may be most beneficial given specificmuta4ons Aswiththeabovesec4ons,knowledgechangesinthepar4cipa4ngprac4ceswillbetrackedat pre;postand6;monthfollowup,andacomponentthatservestoiden4fyandaddressbarriers torecommendedprac4cewillbeimplemented. Followup stoproviderseverythreemonths During the one year period aber the presenta4on the providers will receive quarterly update swiththeirrateofbcr;abl1assessmentbyqpcr,theirprac4cesrate,andanyupdated NCCNguidelineorotherrelevantproviderrecommenda4onsthatmayhavechangedduringthis 4me. Iden4fica4onofBarriersandAc4onItemsat6;months During the 6;month follow;up visit with each prac4ce, a semi;structured interview to iden4fy barriers to implemen4ng the content of the presenta4on will be conducted. Qualita4ve data about these barriers will be recorded and ac4on items to resolve barriers may be iden4fied. Ac4on items may include anything from problem;solving pa4ent adherence concerns to implemen4ngspecificstructuresintheemracrossallsites.thisrepresentsauniqueaspectof thisprojectandwillprovideaniden4fica4onoffactorspreven4ngop4maltreatmentofpa4ents withcp;cmlatprac4cesaroundthestate.asummaryreportofthesebarriersandanyprac4ce changeswillbeproducedasadeliverable. Ifproveneffec4ve,thisinterven4oncanbeeasilybeexpandedtoothercentersandalsoapplied tomorecomplexcasesinthetreatmentofacutelymphocy4cleukemia(all),acutemyeloid Leukemia (AML) and other cancers that rely on ac4ve monitoring of molecular markers and requireadjustmentsofcorrespondingtreatmentwithtkimedica4ons.

8 Brewer, B Page 8 Evalua@on)Design) BaselineDataCollec4on OneyearofbaselinedataforBCR;ABL1assessmentbyqPCRwillbepulledfromtheEMRofall four medical centers in this study. One of the unique aspects of this study, which allows consistent and accurate data collec4on, is that four hospitals across the state now have the sameemrwhichiscalledepic.manyofthesehospitalshaverecentlyimplementedepicand each medical center in the study had a different go live date for EPIC. The retrospec4ve baseline data collec4on can begin on the date that the last hospital in the study switched to usingepicintheironcologyclinics.thisdateisnovember)2nd,)2013formemorialhospitalin Colorado Springs Colorado. Of note this date is approximately six months from the date of poten4al funding of this project which is May of 2014; allowing six months of baseline to be collectedretrospec4velypriortothepoten4alfundingoftheproject. One year of baseline data is needed to ensure sufficient power and accurately document the baseline rate of BCR;ABL1 assessment in comparison to the tes4ng rate recommended in the NCCNguidelines(everythreemonthsor44mesperyear).Theprevalence2(nottheincidence)of CMLinthepopula4onisrela4velylowandes4matedbyroughcalcula4onstobeabout1,333 pa4entslivingwiththediseaseinthestateofcoloradofor2014. Sixmonthsofretrospec4ve data(beforethedateoffunding)andsixmonthsofprospec4vedata(aberthedatethisgrant maybefunded)willbecombinedtomakethisoneyearbaseline.thisbaselinewillbecomplete sixmonthsfromthedateoffunding. DataExtrac4onfromEPIC(EMR) Baselinedatawillconsistofdiagnos4candbillingdata,whichwillbeobtainedusingICD;9code for CML (205.10) and data from the EMR of the four medical centers in this study. This informa4on will be de;iden4fied at the pa4ent level, but will iden4fy the prac4ce where the pa4entwastreatedandtheoncologistorphysiciantrea4ngthepa4ent.dataaboutbcr;abl1 monitoring by qpcr will consist of billing data using the CPT codes and for this test. BCR;ABL1assessmentbyqPCRfromperipheralbloodwillbeevaluatedagainsttheideal four4mepointsperyearrecommendedbythenccn,todeterminethebaselinerateofeach prac4cesadherencetothenccnguidelines. ComparisonofBaselineRatetoPost;Interven4onRate Diagnos4c and billing data will also be collected following the one;year interven4on period usingthesameprocessandwillbeanalyzedtoassesschangeintes4ngratesbyphysicianand byprac4ce.thedataanalystwillberesponsiblefordatamanagementandanalysisandwillbe supportedbyaddi4onalconsulta4onfromthebiosta4s4cscorefortheuniversityofcolorado CancerCenter. PreliminaryPowerAnalysis Apreliminarypoweranalysiswasconductedusingeachofthefourpoten4alBCR;ABL1tes4ng data points per year. Table 1. highlights the percentage change in adherence to BCR;ABL1 tes4ngandthecorrespondingminimumnumberofpossibletes4ngeventsperyearneededto detectthatchangeatthe.05leveland80%power.thesecondcolumnispresentedasarange toaccountforthepossibilityofdifferentbaselineadherencerates.

9 Brewer, B Page 9 Table)1)Sample)Size)Required)for)80%)Power,)α)=)0.05)or)beWer) PercentChangeinAdherencetoBCR;ABL1 Tes4ng ; ; ; ; ;27 Evalua4onofChangesinProviderKnowledge Preandpost;testsimmediatelybeforeandaberthetrainingswilldocumentimmediatechanges inproviderknowledge.afollow;uppost;test,conductedat6months,willdocumentchangesin providerknowledgefromtheinterven4on.datafromthepre;postand6;monthposttestswill be entered into a separate rela4onal database to determine provider changes in knowledge aboutthecontentofthepresenta4onandthemaintenanceofanychangesinknowledgeat6; months. ExpectedChangeinProviderKnowledgeandBehavior Basedonpriorstudyofchangesinproviderknowledgeandthedifficultyofcrea4ngbehavior changeinmedicalproviders,changesinknowledgeareexpectedtobesignificantlygreaterthan thosefoundinbehavior. 12 Theamountofchangeinphysicianbehaviorregardingappropriate assessment of BCR;ABL1 will depend on the baseline assessment rate. The results of several recentna4onalassessmentsofmonitoringthatputtherateofappropriatemonitoringofthis marker at 31% for community and 51% for providers at academic sejngs, the poten4al for changes of up to 20;40% improvement in adherence to the NCCN guidelines for appears significant given the intensive use of interac4ve audit and feedback about physician s rate of assessmentofthisvitalmarkeroftreatmentresponse. Dissemina4onofProjectOutcomes Theoutcomesofthisprojectwillbepublishedinrelevantpeerreviewedscien4ficjournals.The informa4onabouttheefficacyofproviderbehaviorchangeandchangesinproviderknowledge as a result of the interven4on will also be shared directly with the prac4ces involved in the study.thedatafromthe6;monthprocessofiden4fica4onofbarrierstoeffec4veprac4cewill alsobesharedalongwithanyprac4ceimplementa4onsthatmaygeneralizetootherprac4ces. AnychangestotheprocessthephysiciansuseinEPICwillbedocumentedforintegra4oninto otherprac4cesthatalsou4lizeepic,orsimilarsystems,asanemr.ifeffec4ve,thisstudymay be con4nued or expanded with addi4onal funding. The treatment of CP;CML with TKI medica4ons represents the future cancer treatments in which highly toxic treatments are avoided but pa4ents and their medical providers must engage in appropriate long;term monitoringoftreatmentresponseandpa4entsmustengageinadherencetothetkimedica4on over 4me. If effec4ve, this interven4on may be tailored to other popula4ons that require ongoingvigilanceofmonitoringandadherencetoadailymedica4onregimen. RangeofPossibleTes4ngEventsperYear(4 perpa4ent)

10 Brewer, B Page 10 Detailed)Workplan)and)Deliverables)Schedule) Uponpoten4alfundingofthisprojectinMayof2014,thecrea4onofarela4onaldatabaseand collec4onofrelevantdiagnos4candbillingdataforthebcr;abl1assessmentrateinepicwill beini4ated.thisdatawillincluderetrospec4vedataforthepast6months(star4ngnovember 2013) for all prac4ces involved in the study. During the ini4al six months of the study the interac4ve training will be created along with the development of related pre and post;test content. The baseline data will be computed in aggregate, per prac4ce and per provider in Novemberof2014andreportedsothatitmaybeusedasdescribedaboveintheinterven4on sec4on. The ini4al trainings provided by the PI to the prac4ces involved in the study will occur at the fourprac4cesduringthemonthsofdecember2014andjanuary2015withpreandpost;tests givenbeforeandaberthepresenta4onateachprac4ce.pre;posttestdatawillbeanalyzedto assessthechangeinphysicianknowledgeasaresultofthetraining.datacollec4onofbcr;abl1 willcon4nuelatercomparisonbetweenbaselineandpostinterven4onratesofassessment. Thefirstquarterlyupdatewillbesenttoeachoftheprovidersincludingtheirpersonalrateand theirprac4cesrateofbcr;abl1assessmentincomparisonwiththeidealasdeterminedbythe NCCNguidelines.Thenextquarterlyupdateofassessmentrateswillbeprovidedin;personat6 months from the ini4al presenta4on (June and July 2015). During this 6;month follow up mee4ng an addi4onal post;test will be given to the par4cipants in the study and analyzed to assessphysicianknowledgereten4on.atthismee4ng,briefinterviewswiththethoughtleaders will be conducted in each prac4ce to determine any ongoing barriers that exist in any of the aspectscoveredbytheinterven4onwiththeresultsrecordedasqualita4vedatatoassistwith the interpreta4on of quan4ta4ve results recorded in this study. Problem solving interven4ons around these barriers at each prac4ce will be implemented if appropriate. A final quarterly updateabouteachphysicianandprac4cesrateofbcr;abl1assessmentwillbesentby in December2015. One year following the ini4al presenta4on (January 2016), the final follow;up dataset will be extracted from EPIC and analyzed for comparison with the baseline data. The results of this comparison will be reported back to the par4cipa4ng prac4ces involved in the study and summariesofthepre;postand6;monthpostknowledgeassessmentswillalsobeanalyzedfor changesinphysicianknowledgeover4me.finaldataanalysisandwriteupwillbecompletedby April2016. Fig.3TimelineofStudyEvents:Workplan

11 Brewer, B Page 11 SpecificDeliverables,DatesofDeliveryandEs4matedCostsperDeliverable Deliverable Description of Deliverable Date Delivered Estimated Cost One year baseline data summary One year baseline data summary for rate of appropriate assessment of BCR-ABL1 in CML patients at four practices across Colorado (6 months pulled from prior to funding start) 11/12/14 $10,000 Presentation Completed Pre and Post test questions developed Interactive presentation slides with live polling capability through software (Top Hat) 12/5/14 $12,000 Pre and Post test questions developed with biostatistics consultation 12/5/14 $3,000 Quarterly to Provider 1 Quarterly to Provider 2 Summary of 6 month follow-up to identify barriers Quarterly to Provider 3 Quarterly to Provider 4 Final summary of rate of BCR- ABL1 assessment Final summary of changes in physician/ provider knowledge Contains individualized feedback about current rate of BCR- ABL1 assessment over the last three months and the practices rate overall. Rate must be calculated and sent individually to each provider. Contains individualized feedback about current rate of BCR- ABL1 assessment over the last three months and the practices rate overall. Rate must be calculated and sent individually to each provider. Qualitative results of 6 month interview with thought leaders about identification of barriers to implementation of what was learned in the presentation: includes structured interview questions. Contains individualized feedback about current rate of BCR- ABL1 assessment over the last three months and the practices rate overall. Rate must be calculated and sent individually to each provider. Contains individualized feedback about current rate of BCR- ABL1 assessment over the last three months and the practices rate overall. Rate must be calculated and sent individually to each provider. Final summary of rate of BCR-ABL1 assessment for each practice and comparison with baseline assessment. This will be distributed to physicians as the fourth quarterly along with their individual rate of change Final summary of changes in physician/provider knowledge at three time points (pre, post and 6-month follow up). Summary of qualitative data and problem solving changes implemented also provided. 3/5/15 $1,000 6/5/15 $1,000 7/14/15 $11,000 9/5/15 $ /5/15 $1000 4/30/16 $11,500 4/30/16 $7,000

12 Brewer, B Page 25 February 25th, 2014 Craig T. Jordan, PhD Nancy Carroll Allen Professor Chief, Division of Hematology University of Colorado Anschutz Medical Campus East 19 th Ave, Room 9122 Research Complex 2, Campus Box B170 Aurora, CO craig.jordan@ucdenver.edu Benjamin W. Brewer, Psy.D. Assistant Professor Director, Clinical Psychology and Counseling Services Division of Hematology University of Colorado Cancer Center 1665 Aurora, Court, Suite 4331 Aurora, Colorado Dear Ben, We are pleased to offer this letter of support for the project Determining and Addressing Adherence to the NCCN Guidelines for Chronic Phase CML based at the University of Colorado, Denver. As Chief and Associate Chief for the Division of Hematology at the University of Colorado, Denver Anschutz Medical Campus we are delighted to see this study go forward, and are very supportive of the project. We fully endorse your efforts and feel that, as a clinical health psychologist who works with CML patients routinely, you will be able to provide a unique approach to this grant that takes into account the behavioral aspects of both physician and patient behaviors that affect adherence to the best practice recommendations for CML. We are enthusiastic about this projects creation of a database of CML patients that will track BCR/ABL assessment by PCR as it will provide our program with our rate of assessment, which may lead to improvements in our ability to provide appropriate care to our patients. This project also creates a system wide database of CML patients that may prove useful for future research collaboration and an intervention that will likely improve the treatment of patients across Colorado. We are invested in continuing to build a strong relationship between our practice and the other practices involved in this study as we are now unified under common leadership and have recently become part of the same University Health System. Our practice is looking forward to participation in this project and we wholeheartedly support your application. Sincerely, Craig T. Jordan, Ph.D. Professor, Department of Medicine Chief, Division of Hematology University of Colorado, Denver Clay Smith, MD Professor, Department of Medicine Associate Chief, Division of Hematology University of Colorado, Denver

13 Brewer, B Page 26

14 Brewer, B Page 27

15 Brewer, B Page 28

16 Brewer, B Page 29 References 1 TheAnnenbergCenterforHealthSciencesatEisenhower,ClinicalCareOp4onsandAXDEV GroupInc.CML,ALL,andB;CellLymphomas:UnderstandingProfessionalPrac4ceGapsand Educa4onalNeedsamongHematologistsandMedicalOncologistsintheUnitedStates. AvailableathSp:// 2Guérin,A.,Chen,L.Wu,E.K.,Dea,K.,Goldberg, S.L.Economicbenefitsofadequatemolecular monitoringinpa4entswithchronicmyelogenousleukemia(cml).j2of2medical2economics, 17(2),89;98. 3 Noens,L.,VanLierde,M.A.,DeBock,R.,Verhoef,G.,Zachée,P.,Berneman,Z.,...&Abraham, I.(2009).Prevalence,determinants,andoutcomesofnonadherencetoima4nibtherapyin pa4entswithchronicmyeloidleukemia:theadagiostudy.blood,113(22),5401; Marin,D.,Bazeos,A.,Mahon,F.X.,Eliasson,L.,Milojkovic,D.,Bua,M.,...&Khorashad,J.S. (2010).Adherenceisthecri4calfactorforachievingmolecularresponsesinpa4entswith chronicmyeloidleukemiawhoachievecompletecytogene4cresponsesonima4nib.journal2of2 clinical2oncology,28(14),2381; Welch,M.A.,&MMSC,P.Ensuringop4maladherencetoBCR;ABL1tyrosinekinaseinhibitor therapyforchronicmyeloidleukemia.from2the2editor, Kantarjian,H.M.,Giles,F.,GaSermann,N.,Bhalla,K.,Alimena,G.,Palandri,F.,...&leCoutre, P.(2007).Nilo4nib(formerlyAMN107),ahighlyselec4veBCR;ABLtyrosinekinaseinhibitor,is effec4veinpa4entswithphiladelphiachromosome posi4vechronicmyelogenousleukemiain chronicphasefollowingima4nibresistanceandintolerance.blood,110(10),3540; HochhausA.,Hughes,T.P.SaglioG.,etal.Outcomeofpa4entswithchronicmyeloidleukemia inchronicphase(cml;cp)basedonearlymolecularresponseandfactorsassociatedwithearly response:4;yearfollow;updatafromenestnd(evalua4ngnilo4nibefficacyandsafetyinclinical TrialsNewlyDiagnosedPa4ents).Programandabstractsofthe54thAnnualMee4ngofthe AmericanSocietyofHematology;December8;11,2012;Atlanta,Georgia.Abstract Hochhaus,A.,Shah,N.P.,Cortes,J.E.,Baccarani,M.,Bradley;Garelik,M.B.,Dejardin,D.,& Kantarjian,H.(2012).Dasa4nibversusima4nib(IM)innewlydiagnosedchronicmyeloid leukemiainchronicphase(cml;cp):dasision3;yearfollow;up.j2clin2oncol,30(suppl15), Na4onalComprehensiveCancerNetwork.Clinicalprac4ceguidelinesinoncology:chronic myelogenousleukemia.v availableat:hsp:// 2013

17 Brewer, B Page SaSerlee,W.G.,Eggers,R.G.,&Grimes,D.A.(2008).Effec4vemedicaleduca4on:insights fromthecochranelibrary.obstetrical2&2gynecological2survey,63(5),329; Jamtvedt, G., Young, J. M., Kristoffersen, D. T., O'Brien, M. A., & Oxman, A. D. (2006). Audit and feedback: effects on professional practice and health care outcomes. Cochrane Database Syst Rev, 2(2). 12 Thomson,O.,Freemantle,N.,Oxman,A.D.,Wolf,F.,Davis,D.A.,&Herrin,J.(2002). Con4nuingeduca4onmee4ngsandworkshops:Effectsonprofessionalprac4ceandhealthcare outcomes.evidence6based2nursing,5(1), Flodgren,G.,Parmelli,E.,Doumit,G.,GaSellari,M.,O Brien,M.A.,Grimshaw,J.,&Eccles,M. P.(2011).Localopinionleaders:effectsonprofessionalprac4ceandhealthcareoutcomes. Cochrane2Database2Syst2Rev,8(8). 14 Ruddy,K.,Mayer,E.,&Partridge,A.(2009).Pa4entadherenceandpersistencewithoral an4cancertreatment.ca:2a2cancer2journal2for2clinicians,59(1),56; Huang,X.,Cortes,J.,&Kantarjian,H.(2012).Es4ma4onsoftheincreasingprevalenceand plateauprevalenceofchronicmyeloidleukemiaintheeraoftyrosinekinaseinhibitortherapy. Cancer,118(12),3123;3127.