CMOs and Implementation Science Researchers: A productive partnership for clinical improvement Sponsored by CMOG and ROCC

CMOs and Implementation Science Researchers: A productive partnership for clinical improvement

Sponsored by CMOG and ROCCAndreas Theodorou, M.D.Russell Howerton, M.D.Laura Peterson, M.D.

Hosted by Alexander Ommaya, D.Sc. and David Longnecker, M.D.1Russell Howerton, M.D.,F.A.C.S., CMO of Wake Forest University Medical Center

Andreas Theodorou, M.D., CMO of the University of Arizona Medical Center

Laura A. Petersen, M.D., M.P.H., F.A.C.P., Associate Chief of Staff for Research, Houston VA Medical Center, Director VA HSR&D Center of Excellence, and Professor of Department of Medicine, Baylor College of Medicine

Guest Speakers 2During the presentation, your telephone line will be muted

To submit questions for the speakers, please use the chat box on the lower right side of your screen

Q& A CommunicationAgenda

12:30 Introduction David Longnecker, MD Alexander Ommaya, D.Sc.12:35 Russell Howerton, MD12:45 Questions12:50 Andy Theodorou, MD1:00 Questions1:05 Laura Petersen, MD, MPH1:15 Questions1:20 Next Steps1:30 AdjournAAMC Implementation Science/CMO WebinarMonday March 25th, 20135Observation Unit Stress Imaging to Manage Patients with Intermediate to High Risk Acute Chest Pain6ObjectiveTo reorganize existing resources to deliver care in a way that reduces hospital readmissions among patients with intermediate to high-risk chest pain.Wake Forest Baptist HealthVashi et al. JAMA 2013

Wake Forest Baptist HealthAccumulation of ED visits and hospital readmissions dont stop at 30 days; (see our data)8A possible solution: Observation Unit Stress Imaging Care PathwayHighly efficient units driven by care algorithms, staffed by midlevel providersEfficient and cost effective in low risk patients

ACC/AHA: Class I recommendation endorses OU careACC / AHA NSTE ACS guidelines: Anderson et al. Circulation 2007; Institute of Medicine:clinical decision units reduce boarding and diversion, avoid expensive hospitalization, and appear to contribute to improved management ...Wake Forest Baptist HealthThe data and recommendations are driven off of data from low risk patients; performance in intermediate to high-risk patients is only known in single center settings.9Very low riskLow riskHigh riskIntermediate riskLower complexity:Care easily integrated into care algorithmProven efficacy of OU care: Low event rates High patient satisfaction Widely adoptedHigher complexity = higher readmission rate 25-40% with pre-existing CAD (1-3)Perceived complexity inhibits development of care algorithmsIs OU care an alternative to inpatient care?Cost?Event rates after discharge?Readmission rates?Higher risk = high complexity1. Tatum et al. Ann Emerg Med, 19972. Stowers et al. Ann Emerg Med 20003. Gomberg-Maitland et al. AHJ 2005

In these higher complexity patients, our proposal cuts against the grain of natural instincts. First, one would anticipate that these patients do better with inpatient admission, being cared for by a cardiologist. Our data suggest that by integrating technology into a process, we can conduct a more consistent evaluation than currently occurs in the inpatient setting. Recall that in the inpatient arena, patients are receiving a highly tailored evaluation, whereby some undergo invasive angiography, stress testing, or no testing. We now must test in a multicenter setting whether using our OU strategy can achieve the substantial reductions in readmissions that we observed.10MethodsDesign for 2 RCTs conducted at Wake ForestPatients at intermediate to high-probability for ACSED evalRandomizeObservation unitInpatient careStress imagingSerial biomarkersCare per individual providersFollow up through 1 yearWake Forest Baptist HealthEnlarge text11Cost of OU care versus Inpatient care

(Miller et al. Ann Emerg Med 2010)

(Miller et al. JACC:Imaging 2011)Analyses based on intent to treatWake Forest Baptist Health

38% vs 13%, P=0.004OU CareInpatientPComposite7 (13%)20 (38%).004 Revascularization1 ( 2%) 8 (15%).031 Hospital readmission4 ( 8%)12 (23%).033 Recurrent cardiac testing2 ( 4%) 9 (17%).028InpatientOU CareTrial 2 Primary outcome: CompositeReadmit, Revasc, Recurrent testingWake Forest Baptist HealthCombined events, Trials 1 and 2Observation UnitInpatientAdverse Events Death1/104 (1.0%)0/110 (0%) MI (after randomization)6/104 (5.8%)5/110 (4.6%)Wake Forest Baptist HealthGoes against the perception that inpatient care is more comprehensive; so how does this work? ~75% get index visit testing, many receive cath as the initial test. This contributes to revascularization procedures.Literature shows that patients with persistence of sx after revasc (i.e. it was not their heart) have worse perception of health and are more likely to be readmitted. More accurate selection for revasc, and a more definitive diagnostic information near the point of care delivery, affords a transition away from the diagnostic mindset, to the treatment mindset. Also avoids the ongoing outpatient referral for testing that further prolongs the diagnostic process.

Notes about participant death: One subject in the OU group expired due to PEA arrest during the study follow-up period. This participant was initially evaluated with the study intervention and was felt safe for discharge home. Over the next 6 months, the patient had multiple admissions to the hospital with progressive deterioration in their health status. During the final admission, the patient suffered an in-hospital cardiac arrest and could not be resuscitated. The patient experienced a non-ST elevation myocardial infarction during the follow up period, more than six months after his initial visit. The cause of death listed on the death certificate was severe acidemia, ventricular fibrillation arrest, cardiomyopathy, and coronary artery disease. It was felt that this participants death most likely represented a progression of his underlying medical conditions present at the time of enrollment rather than a study-related event. This event was graded by the PI as unexpected and unlikely to be related to the study. 14SummaryOU care with perfusion imaging at Wake Forest:reduces cost, readmissions, and revascularization proceduresDeath and MI Very low rates with either strategyLeverages and reorganizes existing resources to achieve these benefitsWake Forest Baptist HealthFuture directions and opportunitiesImplementation:Can benefits observed in single center trials be achieved in a multi-center setting?Implementation study with outcome surveillanceAre results dependent on using cardiac MRI as the imaging modality?How can we remove barriers so we can organize EMRs to automate data capture and outcome surveillance across medical centers?Wake Forest Baptist HealthMusculoskeletal Emergency CenterDevelopment of an integrated practice unit (IPU).Replace physician-centric processes with patient-centric onesDecrease the distance between patient and final decision-maker.Prospective database to monitor clinical outcomes.Maximize Value. Optimize Education.

Wake Forest Baptist HealthAreas for Multicenter CollaborationDevelop evidence-based clinical practice guidelinesDevelop competencies and standardized education for the new field of Musculoskeletal Emergency Medicine Enhance understanding of operational efficiency and time-driven activity based costingMulticenter prospective databaseCreate best practices in this field

Wake Forest Baptist HealthQuestions? 3AmdHealthy Together Care PartnershipHealthcare Dream Implementation Nightmare

20

Andreas A. Theodorou, MD, FAAP, FCCMChief Medical OfficerUniversity of Arizona Medical CenterProfessor and Associate Chair, Pediatrics

20UAHN Quick Overview:Includes two hospital campuses -The University of Arizona Medical Center (University Campus, including Diamond Childrens) & (South Campus)40 clinics, a health plan division and practice plan for physicians from the University of Arizona College of MedicineOnly Level 1 Trauma Center in Southern Arizona (University Campus)Comprehensive Transplant Program(University Campus) NCI Designation

University & South Campus Patient Statistics 2012-13:21University Campus & South CampusAdmissions20,489Emergency Visits74,632Total Net Revenue533193Patient beds592

Common CMO Priority IssuesHospital/Healthcare-associated InfectionsCore MeasuresProcedural ComplicationsFallsPatient SatisfactionAccess to CarePatient Through-PutSafe Medication UseReadmission RatesHospitalization AvoidanceContinuity of Care

Dual-Eligible The Healthy Together Care Model

Focuses on approximately 345 dual eligible Special Needs Plan members within our University of Arizona Health Network

Designed to improve quality of care for this high risk/high cost population living in the home and community

23

All UAHN Health Plan Duals (n9,000)UAHN Health Plans Dual Eligibles in Pima County (n4,000)UAHN Health Plan Duals in Pima County assigned to UAHN primary care provider (n=345) Early data indicate that sub-population health risk and cost profile is representative of all UAHN Health Plans dual-eligible SNP population and the national dual-eligible population

Healthy Together Population24n=345

15%5%30%50%48%37%14%1%Within our sub-population, the costliest 5% of enrollees account for 48% of total cost of care, while the costliest 20% account for 85% of total cost*Based on retrospective chart review and analysis of 307 dual eligibles with UAHN Health Plan coverage and assigned to primary care with a UA Health Network provider (Goel, et al, 2011)Cost of Care 25The Care Model Uses Multiple Evidence Supported StrategiesInterprofessional team-based careHome-based Primary Care for ~ 45 most complex and homebound (with telehealth)Case management and telehealth for rest of cohort, in collaboration with primary care providersMedication ReconciliationIntegrated behavioral health/physical health care managementPatient Engagement and Shared Decision Making26 Healthy Together Care Partnership Delivery System/Health Plan Partnership designed to reduce utilization in a high risk/high utilizing populationTargeted utilization reduction in populationDecreased ED UtilizationDecreased Cost of Admissions Decreased Readmission RateDecreased Med Cost by Pharmacy Review Net savings if targets achieved: $1.5 Million

Good NewsPrimary goals of the program include cost savings, improved quality and satisfaction with care, blended physical and behavioral health, development of individualized strategies to manage at-risk patients, and development of best practices for dual eligible patients in SNP community care settings.Bad News18 months later the project had still not started!Overcoming Academic Medical Center Inertia: Building an Innovative Dual Eligible Service Line Great News!Project now fully launched and first 4 patients enrolled last week!2828Reasons for delayed ImplementationOrganizational ComplexityClinical Cost StructuresTraining ExpectationsCredentialing and Privileging of NPsEmployment/supervision of NPs, SWs and RNsWho Provides Space and Infrastructure Non-integrated Information SystemCoding and Compliance IssuesLong Term Plans30 Eventual development of a stand-alone product that would provide coverage to the entire population of dual-eligible SNP patients resulting in reduced costs and better outcomes (e.g., lower hospitalization rates, better medication compliance, improved morbidity and mortality)

Year 1: Dual-eligible patients in UAHN care (n=345), starting with highest-cost stratum, and rolling additional services out to remaining patients in lower-cost stratum.Year 2: Expand to include SNP patients in Pima CountyYear 3: Expand to include remainder of SNP

Savings realized through better care of high-cost stratum would be the basis of funding novel programs for the entire SNP population

Current stateStand-alone product for all UAHN SNP patientsHealthy Together Pilot30Healthy Together development teamJane Mohler, NP-C, MPH, PhD1,2;Nancy Wexler, MPH1; Richard Slaughter3; James Stover3; Patricia Harrison-Monroe, PhD1; Tom Ball, MD1,2; Mindy Fain, MD1,2

UA, College of Medicine1; Arizona Center on Aging2; UA Health Network, Health Plans3

313131Questions? CMOs and Implementation Science Researchers: A Productive Partnership for Clinical ImprovementLaura A. Petersen, MD, MPH, FACPProfessor of Medicine and Chief, Section of Health Services Research, Baylor College of MedicineDirector, Houston VA HSR&D Center of Excellence, Associate Chief of Staff for Research, Michael E. DeBakey VA Medical CenterMarch 25, 2013Petersen AAMC Webinar33

Iraq, March 2003 Embedded journalist Chip Reid, right,travels through southern Iraq with soldiers from the 3rd Battalion, 5th Marine Regiment.

Petersen AAMC WebinarPartnerships Between Researchers and the VA Health Care SystemPartners include national program offices, regional CMOs and facility leadership, clinical leaders and managers, and individual cliniciansIncrease impact of research on Veteran health by:Ensuring appropriate input into research priorities from a variety of VA stakeholdersEncouraging ongoing communication between research and operationsEnabling more timely response from the research community to emerging health system issuesFacilitating effective communication of research results and uptake of research into practice (implementation)

Petersen AAMC WebinarTranslation/Implementation HighwaysJAMA. 2007;297(4):403-406. doi:10.1001/jama.297.4.403

From: Practice-Based ResearchBlue Highways on the NIH RoadmapHold for cartoonPetersen AAMC Webinar

Problems with Linear TranslationT1 T217 years from basic discovery to clear evidence from clinical trials Contopoulos-Ioannidis et al. Science 321:1298-99

T2 T310 years for widespread guideline implementation

Linear approach to translation creates excessive lag in evidence implementationModels for Linking Research to ActionAdapted from Lavis et al, 2006

Petersen AAMC Webinar

39Technical Problems vs. Adaptive Challenges (from Ronald Heifetz and Marty Linsky, Leadership on the Line)Easy to identifyOften lend themselves to quick and easy (cut-and-dried) solutionsOften can be solved by an authority or expertRequire change in just one or a few places; often contained within organizational boundariesPeople are generally receptive to technical solutionsSolutions can often be implemented quickly-even by edictPetersen AAMC WebinarDifficult to identify (easy to deny)Require changes in values, beliefs, roles, relationships, & approaches to workPeople with the problem do the work of solving itRequire change in numerous places; usually cross organizational boundariesPeople often resist even acknowledging adaptive challengesSolutions require experiments and new discoveries; they can take a long time to implement and cannot be implemented by edictThe single biggest failure of leadership is to treat adaptive challenges like technical problemsTechnical ProblemsAdaptive ChallengesExamples from Health CareImplement electronic ordering and dispensing of medications in hospitals to reduce errors and drug interactionsImprove availability of hand sanitizerCreate workflow and structure to deal with low risk chest pain patients in the EDWhat are the appropriate peer facilities for quality and efficiency comparisons (Partnership project)Petersen AAMC WebinarEncourage nurses and pharmacists to question and even challenge illegible or incorrect prescriptions by physiciansGet health care providers to improve hand washing ratesChange primary care team roles to adopt a patient centered medical home model (Partnership project)Design and test new model of provider payment to reward quality (Partnership project)Technical ProblemsAdaptive ChallengesResearchers Can Help with Evidence to Overcome Both Technical and Adaptive ChallengesWhat are the social, cognitive, workflow barriers to handwashing?Petersen AAMC WebinarExamples of Partnership Projects at the Houston Health Services Research and Development Center of ExcellenceLongitudinal measures of quality (Petersen and Woodard, PIs) 10 publications (Circulation, Medical Care, JAGS, HSR, Diabetes Care)Resource efficiency (Petersen, PI) 7 publications (HSR, Medical Care)Hospital and community living center peer facilities (Petersen, PI) 2 publications (HSR, American Journal of Managed Care)Evaluation of primary care re-design (Hysong, PI)RCT of pay for performance (Petersen, PI) 3 publications

Petersen AAMC WebinarUnderstanding DifferencesNetwork Needs:Value fast turnaround, practical projects rather than publications (implementation/external validity)HSR&D cheaper, more knowledgeable than external consultantsIt doesnt really have to be perfectRapidly changing needs, prioritiesResearch Needs:Academic productsInternal validity focusSlower pace (IRB, funding cycles, data use agreements, HR, credentialing, contracting)Business model

ContractingHRCredentialingVA ResearchPetersen AAMC WebinarChallenges to PartnershipsRegulations! Data Use AgreementsResearch training for non-researchersBusiness modelPetersen AAMC WebinarAdvantages of PartnershipAligning research with specific health system partners to increase the impact on VHAAccelerating the timetable for research in areas critical to the health systemFocus upon implementation early in the research processPetersen AAMC WebinarEmbedded researchers are cheaper and more knowledgeable about the delivery systemMay have access to data that other consultants dont haveDiverse skill set to tackle problemsFor ResearchersFor PartnersBuilding and Maintaining PartnershipNeed champions within partnering organizationSome face to face meetings, especially early onContinued mutual recognition of needs of partnering organizationContinued attention to sustainabilityAppreciation of differencesPetersen AAMC WebinarPartnership Research is a Team SportDoing partnership work requires an excellent teamAbility to respond to questions and requestsRelationships, relationships, relationships!Petersen AAMC WebinarPetersen AAMC Webinar

Questions & Next Steps text

One Group of Research Users

Producers or Purveyors of Research