ORIGINAL RESEARCH Primary Care Clinic Re-Design for ... · fallen disproportionately on the primary...
Transcript of ORIGINAL RESEARCH Primary Care Clinic Re-Design for ... · fallen disproportionately on the primary...
ORIGINAL RESEARCH
Primary Care Clinic Re-Design for PrescriptionOpioid ManagementMichael L. Parchman, MD, MPH, Michael Von Korff, PhD, Laura-Mae Baldwin, MD,Mark Stephens, BS, Brooke Ike, MPH, DeAnn Cromp, MPH, Clarissa Hsu, PhD, andEd H. Wagner, MD, MPH
Background: The challenge of responding to prescription opioid overuse within the United States hasfallen disproportionately on the primary care clinic setting. Here we describe a framework comprised of6 Building Blocks to guide efforts within this setting to address the use of opioids for chronic pain.
Methods: Investigators conducted site visits to thirty primary care clinics across the United Statesselected for their use of team-based workforce innovations. Site visits included interviews with leader-ship, clinic tours, observations of clinic processes and team meetings, and interviews with staff and cli-nicians. Data were reviewed to identify common attributes of clinic system changes around chronic opi-oid therapy (COT) management. These concepts were reviewed to develop narrative descriptions of keycomponents of changes made to improve COT use.
Results: Twenty of the thirty sites had addressed improvements in COT prescribing. Across thesesites a common set of 6 Building Blocks were identified: 1) providing leadership support; 2) revisingand aligning clinic policies, patient agreements (contracts) and workflows; 3) implementing a registrytracking system; 4) conducting planned, patient-centered visits; 5) identifying resources for complexpatients; and 6) measuring progress toward achieving clinic objectives. Common components of clinicpolicies, patient agreements and data tracked in registries to assess progress are described.
Conclusions: In response to prescription opioid overuse and the resulting epidemic of overdose andaddiction, primary care clinics are making improvements driven by a common set of best practices thataddress complex challenges of managing COT patients in primary care settings. (J Am Board Fam Med2017;30:44–51.)
Keywords: Ambulatory Care Facilities; Analgesics, Opioid; Chronic Pain; Drug Overdose; Leadership; Opioid-Related Disorders; Prescriptions; Primary Health Care; Registries; Research Personnel; Workflow
The current epidemic of prescription opioid over-dose and deaths may be the most significant iatro-genic epidemic in the recent history of medicine inthe United States.1–3 Although the responsibility
for initiation of opioids for chronic pain may beattributed to primary care clinicians to some de-gree, the sheer numbers of patients taking a pre-scription opioid medication for long-term chronicpain has placed an enormous burden on primarycare where the majority of opioid prescriptions arewritten.4 This burden may contribute to burnoutand stress in primary care settings where both pre-scribers and clinic support staff struggle daily tobalance risks and the potential for abuse and diver-sion with empathy for the suffering of chronic painpatients.5
As evidence accumulates to support more judi-cious use of COT, guidelines for prescribing opi-oids such as those recently released by the U.S.Center for Disease Control and Prevention havebeen released or updated.6 However, implementing
This article was externally peer reviewed.Submitted 6 June 2016; revised 1 September 2016; ac-
cepted 7 September 2016.From the MacColl Center for Health Care Innovation,
Group Health Research Institute, Seattle, WA (MLP,MVK, DC, CH, EHW); Department of Family Medicine,University of Washington, Seattle (L-MB, BI); ChangeManagement Consulting, Seattle (MS).
Funding: The Robert Wood Johnson Foundation pro-vided funding for the LEAP project. Additional funding wasprovided by the Agency for Healthcare Research & Quality(grant R18HS023750).
Conflict of interest: none declared.Corresponding author: Michael L. Parchman, MD, MPH,
Group Health Research Institute, 1730 Minor Ave, Suite1600, Seattle WA 98101 !E-mail: [email protected]).
44 JABFM January–February 2017 Vol. 30 No. 1 http://www.jabfm.org
these recommendations will require changes toclinic systems and workflows of health care teamsacross the entire clinic, not just changes by individ-ual prescribers. Systematic changes to make opioidprescribing safer have been evaluated in large inte-grated health care delivery systems such as Veter-an’s Administration clinics and other large grouppractice settings.7–9 Less is known about systemsapproaches to improve address the overuse of COTacross a diversity of primary care settings.
Here we describe a framework comprised of 6Building Blocks to guide smaller clinics in practiceredesign and improvement, derived from observa-tions of thirty primary care clinics selected for theirinnovations in team-based care delivery. The intentis to provide general principles and best practicesthat can be adapted or applied across diverse pri-mary care clinic settings.
MethodsData SourcesIn 2012 the Robert Wood Johnson Foundationfunded Primary Care Teams: Learning from Ex-emplar Ambulatory Practices (LEAP).10 Thegoal of LEAP was to study team-based workforceinnovations across diverse primary care clinicswithin the United States. A national advisorycommittee of experts and stakeholders in primarycare provided nominations. Sites were screenedin a telephone interview and reviewed by mem-bers of the LEAP study team who then maderecommendations to the advisory committee whomade the final selection of thirty clinics for sitevisits.
A team of 3– 4 LEAP investigators conducteda 3-day site visit to each clinic. Site visits in-cluded group interviews with the leadershipteam, comprehensive tours of each clinic focusedon understanding workflows and general clinicfunctioning, detailed observations of clinic pro-cesses and team meetings, interviews with indi-vidual staff and clinicians, and shadowing of pa-tients during their visits. Supporting documents,tools, and resources such as clinic policies, work-flows, patient-facing materials, and decision toolswere also collected.
Data Coding and AnalysisSeveral LEAP investigators initially coded the sitevisit data (DC, CH) using broad codes to capture
key primary care team member roles and functions.One broad code for “medication management” in-cluded experiences with managing COT. OneLEAP research team member (DC) then subcodedfor opioid-related data within the broad medicationmanagement code by searching the LEAP datasetfor key words such as “opioids,” “pain,” and “nar-cotics” to ensure that the opioid subcode would beapplied to all relevant data. The opioid subcodecontained 91 pages of data.
A small team of LEAP investigators (MLP, DC,CH) reviewed the coded data with the goal ofidentifying common components that informedclinic redesign efforts to improve COT manage-ment. These concepts were presented to the largerteam (MVK, MS, LMB, EHW), who verified themand developed descriptions of each component.Themes and concepts were also compared and con-trasted to a Group Health opioid improvementinitiative to identify common elements and strate-gies.8,9 As descriptive phrases and definitions weredeveloped, examples of activities used by the clinicsto make the necessary changes were identified fromthe LEAP data sources.
ResultsTwenty of the 30 LEAP clinics had made system-atic improvements in COT management. (Table 1)Each site had different priorities in developingtheir policies and processes related to COT. Theseincluded avoiding the issue (eg, several refused toprescribe opioids for chronic pain); concerns aboutstaff security (protecting staff from threats by drugseekers); enhancing safe management of opioids;and actively treating prescription opioid addiction.Concerns about drug seeking were a higher priorityin many sites than opioid dose reduction and theirCOT-prescribing programs often reflected theirpriorities.
Across the LEAP site where COT was used forchronic noncancer pain, a common set of changeprinciples were noted as playing an important rolein addressing their priorities. Here we describethese elements as 6 Building Blocks that summar-ize our findings. (Table 2)
Building Block 1: Providing Leadership Support“It was definitely another of those transformational ef-forts, because it required everyone in the organization tounderstand what we’re doing and so even the folks at the
doi: 10.3122/jabfm.2017.01.160183 Clinic Re-Design for Opioid Management 45
Table 1. Primary Care Clinic Characteristics and Examples of Opioid Improvement Efforts
State Clinic Type Location Payer Mix Example Opioid Improvement Building Block(s)
PA FQHC, nurse-led Urban 4% Medicare Chronic pain group therapy 4,558% Medicaid14% commercial23% uninsured
WV FQHC, AHEC Rural 17% Medicare Chronic pain group visits 1,2,325% Medicaid Pain registry22% commercial Chronic opioid prescribing policy and
pathway30% uninsuredSC FQHC Rural 32% Medicaid Standard care plans 2,3,4
12% Medicare Patient agreements15% Other In house physical therapy40% uninsured Suboxone
OR FQHC, residency Urban 20% Medicare Chronic pain group visit 2,3,450% Medicaid In-house CAM therapy0% commercial Revised policies30% uninsured Random urine drug tests
Patient agreementsSuboxone28-day refills
NH MMG Rural 45% Medicare Patient agreements 2,52% Medicaid Opioid QI team50% commercial Revised policies3% uninsured
WA MMG Rural, Suburban 20% Medicare Chronic pain re-design team 2,49% Medicaid Suboxone61.5% commercial Pain registry9.5% uninsured Patient agreements
Random urine drug screensWorkflow for refills
MA MMG Suburban 23% Medicare Patient agreements 2,3,45% Medicaid Revised clinic policies70% commercial Suboxone2% uninsured Chronic pain group visits
Random urine drug screensNM FQHC Frontier/ Rural 28% Medicaid Chronic pain group visits 3,4
30% commercial Behavioral health integration on teams19% Medicare Mental health “first aid” training for
staff17% Sliding Fee Suboxone6% self pay
CO FQHC Rural 47% Sliding Scale Revised policies 217% Medicaid Routine PDMP check with refills15% Medicare Patient agreements18% commercial dental No refills on Fridays
ME MMG Suburban 35% Medicare Registry with chronic pain manager 2,3,44.4% Medicaid45% commercial Chronic pain group visits5.4% uninsured Revised policies
Continued
46 JABFM January–February 2017 Vol. 30 No. 1 http://www.jabfm.org
Table 1. Continued
State Clinic Type Location Payer Mix Example Opioid Improvement Building Block(s)
CO MMG Urban, Suburban,Frontier/ Rural
20% Medicare Patient agreements 2,3,421% Medicaid Pre-visit preparation in daily huddle50% commercial Random urine drug test9% self-pay Track PEG scores and PHQ-9
SuboxoneOH FQHC Urban 50% Medicaid Random urine drug test 2,3,4
20% uninsured State PDMP check with refills20% commercial Clinic refill policies10% Medicare
PA PVT Suburban 90% commercial Patient agreements 28% Medicare Revised policies1% uninsured1% Medicaid
ME FQHC, residency Suburban 26% Medicare Provider support and learning group 2,425% Medicaid40% commercial Suboxone9% uninsured Revised policies
WA MMG, residency Suburban 10% Medicare Chronic pain registry with dedicated MAregistry manager
1,2,3,4
50% Medicaid Revised policies30% commercial Patient agreement10% uninsured Nurse intake for new patients on opioids
Random urine drug testState PDMP checkPEG scoresReferral for high risk
WI MMG Rural 17.7% Medicare Patient agreement 25.5% Medicare Revised refill policies73.8%
commercial3% uninsured
MA CHC (hospitalnetwork)
Urban 40% Public Physical therapy assistant 2,440% uninsured Chronic pain group visits led by social
worker20% privateDC FQHC Urban 63% Medicaid Chronic pain group visits 4
6% Medicare Massage therapy20% DC
Alliance6% commercial5% uninsured
NY AHC, residency Urban 10% Medicare Revised clinic policies 2,450% Medicaid Patient agreements30% commercial Behavioral Health Social Worker8% uninsured
CA FQHC Rural 50% Medicaid Chronic pain group visits 2,3,417% Medicare Revised clinic policies3% commercial Pre-visit planning in daily huddle28% uninsured Patient agreements
Abbreviations: AHC, Academic Health Center; AHEC, Area Health Education Center; CHC, community health center; FQHC, FederallyQualified Health Center; MMG, Multi-specialty Medical Group or part of large system; PVT, Private Practice; RHC, Rural Health CenterlTHC, Teaching Health Center.
doi: 10.3122/jabfm.2017.01.160183 Clinic Re-Design for Opioid Management 47
front desk had to understand we were not all a suddenchanging or getting tough, thinking that people weredrug seeking or abusing.” –Office Manager, LEAP siteLeadership played an essential role by both prior-itizing the work and facilitating a consensus-build-ing process to help providers and staff reach ashared understanding about standards of care forCOT patients. Consensus building often startedaround defining who qualified as a COT patientand discussions about the growing evidence ofharm for their patients. Many times this took theform of a story about a patient from their ownclinic who experienced harm from COT use.
Building Block 2: Revising Clinic Policies, PatientAgreement, and Workflows“We’ve become aware of the evidence and the researchindicating that providing opiates has a lot of risk asso-ciated with it and there’s better ways of dealing withchronic pain than just providing prescriptions…. So weput a lot of time into defining what kind of system shouldwe put in place to make sure we evaluate people appro-priately, that we monitor use, and we have a system andact when we discover an issue or a problem with potentialmisuse…. So that took some years, I guess, doing that.”–Medical Director, LEAP site The work often beganwith revising 2 documents, clinic policies, and pa-
Table 2. Six Building Blocks to Guide Management of Chronic Opioid Therapy
Building Block Description Examples of Action Steps
1. Provide LeadershipSupport
Leadership can build organization-wide consensus to prioritize safe,more selective, and more cautiousopioid prescribing
Identify clinical champions to spearhead COT practicechange initiatives.
Provide protected time and space for providers andstaff to discuss and agree upon short and long-termgoals for COT practice change initiatives
2. Revise Policies,Patient Agreementsand Workflows
Revise and implement clinic policiesand define standard work forhealth care team members toachieve safer opioid prescribingand COT management in eachclinical contact with COTpatients.
Convene a team from each area of the clinic to reviseexisting policies or write new ones
Review patient agreement and revise to ensurealignment with clinic policies.
Discuss with all staff and clinicians and modify roles,responsibilities and workflows accordingly
3. Implement a Registryfor PopulationManagement
Implement pro-active populationmanagement before, during, andbetween clinic visits of all COTpatients to ensure that care is safeand appropriate and providemeasure to track COTimprovement activities.
Enter all existing COT patients and their relevantenrollment data into a COT registry.
Assign each COT patient to a single providerresponsible for managing their opioid use and.
Assign a team member in each clinic with responsibilityand protected time for managing and updating theregistry.
Use the registry to track data for prescriptionmanagement (e.g., COT dose, PEG scores tomonitor function and pain, date of state prescriptiondatabase checks)
4. Conduct PlannedPatient-CenteredVisits
Conduct pre-visit planning andsupport patient-centered,empathic communication forCOT patient care.
Review COT registry reports prior to the visit toidentify care gaps
Monitor and adjust management based on function andquality of life rather than pain scores (the PEG scale)
Offer organizational support for clinic staff andproviders to preview charts and do team huddlesabout COT patients
Support staff training, to encourage the use ofempathic communication techniques that
5. Identify Resourcesfor Complex Patients
Develop resources to ensure thatpatients who develop complexopioid dependence, are identifiedand provided with appropriatecare
Identify addiction referral resources and other mental/behavioral health resources, and ensure they arereadily available, setting-up referral protocols oragreements as necessary.
6. Measure Progress Continuously monitor progress andimprove with experience.
Identify key process and outcome measures to monitorpractice change implementation.
Monitor agreed upon COT patient care data, providingand discussing data with clinic staff and medicalproviders at monthly meetings.
COT, chronic opioid therapy.
48 JABFM January–February 2017 Vol. 30 No. 1 http://www.jabfm.org
tient agreements (contracts). Fifteen of the twentyLEAP sites had clinic policies and/or patient agree-ments in place. In some clinics, the discussionsabout what to include in these 2 documents pro-vided a scaffold for on-going hallway conversationsamong providers and staff that resulted in a sharedunderstanding about their approach to COT man-agement. Some clinics discovered a lack of align-ment between language in the patient agreementand their clinic policies and worked to align the 2documents.
“I explain to them the policies, that I might call themfor a pill count or a urine drug screen just so they knowwho I am. But for the most part, I have pretty goodrapport, I think, with them. Like, ‘I am not here toattack you; we’re here to help you. We want to makesure you are taking them safely.’ And it seems to goreally well.” –RN Pain Registry Manager, LEAP SiteCommon elements of clinic policies are found inTable 3. Clinic policies addressed common situa-tions such as new patients who present with anopioid refill request, activities to monitor the safetyof established patients, and policies around stan-dards of care for COT refills. Clinic staff used the
policies to discuss workflow redesign efforts such aschecking the state Prescription Drug MonitoringProgram’s (PDMP) database so that policies wereapplied to daily patient care.
Patient agreements were sometimes used astools to guide discussions with patients about therisks associated with COT, alternative treatments,and signs of aberrant behaviors such as repeatedrequests for early refills that would raise concerns.They were often renewed annually and some pro-viders used them to have discussions about taperingCOT with patients for whom the benefit of con-tinued use of COT was questionable.
Building Block 3: Implementing a Registry forPopulation Management“My registry… is kind of helpful because it has every-thing. It says what the patient has and if they are out ofcompliance, like they need a new pain agreement thatexpired 2 months ago.” –Pain Registry Manager,LEAP Site A registry or some form of trackingsystem was commonly used to monitor COT pa-tients between visits and manage refills requests. Afew had a designated registry manager, usually amedical assistant or nurse, with protected timeeach week or month to update the registry andidentify patients with care gaps. In several clinicsreports from the registry were reviewed duringprevisit planning such as a morning huddle toprepare for a visit. Some sites routed COT refillrequests through the registry manager to identifycare gaps and address them in between visits. Inaddition, data from the registry was commonlyused to track agreed-on measures for quality im-provement in the area of COT management.Some of the items included in a typical registryare found in Table 4.
Table 4. Common Elements Seen in registries
Date of renewal of patient agreement (signed by patient)Current morphine medication equivalent dose of opioid
medicationsDate of most recent PDMP checkDate and result most recent urine drug screenPEG scores (trended at regular intervals)Opioid risk tool scoreMedication list reviewed for concurrent use of sedativesPHQ screen for depression
PDMP, prescription drug monitoring program; PEG, pain, en-joyment, general activity; PHQ, patient health questionnaire.
Table 3. Examples of Common Clinic Policies toSupport Management of Chronic Opioid Therapy
New patients currently on COTAll new patients require a urine drug test and copies of
prior medical records prior to an opioid prescriptionStandard elements of a pain assessment on all new patients
Established Patients COT ManagementNo refills on Monday and FridaysNo early refills for lost or stolen prescriptions or or a
police report for such a refillFace-to-face visit intervals required for a refill based on
level of risk28-day supply only (to avoid running out on weekends)Advanced notification period (e.g. 4 business days) for a
refill request to be processedRandom urine drug screening frequencyFrequency of required PDMP check and who is responsibleFrequency and documentation of screening for depression
and post traumatic stress disorderMonitoring for co-prescribing of sedatives
Others:No initiation of opioids to treat headaches, fibromyalgia or
chronic low back painStandards for when a referral is required to a pain specialist
or mental/behavioral health specialist (e.g., aberrantbehaviors, high dose such as "100 morphine medicationequivalent)
COT, chronic opioid therapy; PDMP, prescription drug mon-itoring program.
doi: 10.3122/jabfm.2017.01.160183 Clinic Re-Design for Opioid Management 49
Building Block 4: Conducting PlannedPatient-Centered VisitsClinics often used huddles or some type of chartreview the day before to prepare for visits withchronic pain patients. They identified care gapssuch as an out-of-date patient agreement or anoverdue check of the state PDMP and used the visitas an opportunity to close those gaps. Staff andclinicians would occasionally rehearse how to havedifficult conversations with patients about aberrantrefill behaviors or abnormal urine drug tests.
“And it was also the process of learning how to say noin a kind way, because it is very difficult when somebodywho is misusing—it is much easier when somebody whois not misusing, but when somebody is misusing and theyknow they are misusing, they get very confrontational.The staff over the past 7 years has learned how to notinflame that.” –Nurse Practitioner, LEAP Site
They often shared “scripts” with each otherabout language they used in talking with patientswho “expect” their prescription pain medicationsand are reluctant to hear about alternative treat-ments.
One important patient-centered componentthat LEAP sites often mentioned was moving awayfrom the traditional visual-analog pain scale to as-sess the effectiveness of current management strat-egies to a more robust patient-centered assessmentof pain, function, and enjoyment of life. The Pain,Enjoyment, General activity (PEG) scale was be-coming more widely used across sites.11–12 Severalproviders mentioned that tracking the PEG scale atevery visit helped them avoid inappropriate COTdose escalation and sometimes led to conversationsabout tapering the dose.
Building Block 5: Identifying Resources for ComplexPatients“Both patients and providers were very concerned thatthey did not have access to traditional resources such asPhysical Therapy to support patients with chronic pain sowe created group clinics for chronic pain staff by abehavioral health consultant, a physical therapy assis-tant… and a health coach from our clinic.” –BehavioralHealth Social Worker, LEAP Site
Clinics recognized that some patients requiredmore support and treatment for addiction, opioiduse disorder, or mental/behavioral health issuesthan was currently available within their clinic sys-tem. Some identified existing community resourcesand built stronger linkages with them. Others de-
veloped resources within their setting to deliverthese services to the patient directly. A few clinicsused shared appointment (group) visits with a stan-dard curriculum to improve patient self efficacyaround managing their chronic pain. In 1 LEAPsite patients on high-dose opioids were required toattend these sessions to refill their opioid medica-tion. Seven of the clinics had a buprenorphine pre-scriber who provided medication-assisted therapy.
Building Block 6: Measuring progress“My [work with] chronic pain [patients] is my favoritepart of my job because I love to feel like I can help them.”–Medical Assistant Registry Manager, LEAP SiteClinics often had improvement teams with a spe-cific focus on COT use. They selected processmeasures such as proportion of patients with anup-to-date agreement in the chart, proportion ofpatients with a PDMP check twice in the past year.Important outcome measures that clinics foundvaluable to track included: average PEG scalescores11–12 (see Building Block 4), the proportionof patients on high-dose COT (eg, daily morphine-equivalent dose !90), and the monthly number ofpatients who transitioned from opioids for acutepain to those on COT. Clinics reported these mea-sures and discussed them during monthly staffmeetings and leadership team meetings.
DiscussionThrough data gathered from site visits to primarycare clinics that exemplified team-based models ofcare, we identified 6 common change strategies or‘building blocks’ used by innovative primary careclinics to improve COT management. The themesthat emerged consolidate and systemize best prac-tice approaches to addressing the complex chal-lenge of managing COT patients in primary caresettings. These findings are not meant to be com-prehensive or exclusive of other approaches orstrategies that might be equally effective or neces-sary and clearly must be adapted to local contextand resources.
Ongoing challenges faced by smaller primarycare clinic settings included a lack of validated clin-ical performance measures for improving care pro-vided to COT patients, difficulty identifying ade-quate community resources such as access tomental/behavioral health and addiction services,and concerns about a paucity of evidence-based
50 JABFM January–February 2017 Vol. 30 No. 1 http://www.jabfm.org
alternatives to COT for chronic pain. At least 1LEAP site made the decision to no longer prescribeopioids for managing chronic pain because of theseand other concerns. Providers in many of theLEAP sites that made systematic changes to sup-port safer opioid prescribing reported that tacklingthese challenges restored a sense of joy and fulfill-ment in their daily work.
The recent release of new COT-prescribingguidelines combined with growing evidence of pa-tient harm with use of COT for chronic noncancerpain will require changes in workflow and clinic-wide systems to adopt and implement the newstandards.6 Changing systems of care is differentfrom changing provider prescribing habits, but ex-perience suggests that system change in primary caresettings is a critical component to sustained change inprovider behaviors.13 The practical steps and strate-gies represented in the 6 Building Blocks were usedby innovative clinics to address the use of COT intheir patient population and should be considered indesigning improvement initiatives in other primarycare settings. It is important to note, however, thatthese new guidelines and the associated workflowredesigns to implement them cause burdens of theirown. Unless they can be demonstrated to significantlyimprove patient outcomes, while also decreasing pro-vider and staff burnout, there may be resistance toimplementation. In addition, primary care clinicsalone cannot stem the tide of opioid overuse withinlocal communities; it will require community-wideinitiatives that include all prescribers.
The content is solely the responsibility of the authors and doesnot necessarily represent the official views of the Robert WoodJohnson Foundation or the Agency for Health care Researchand Quality. The authors thank the members of the RWJFLEAP National Advisory Committee, the site visit teams, andthe leadership and staff of the 30 exemplar practices.
The Group Health Research Institute’s Institutional ReviewBoard reviewed the LEAP project and classified it as exempt.
To see this article online, please go to: http://jabfm.org/content/30/1/44.full.
References1. Von Korff MR, Franklin G. Responding to Amer-
ica’s iatrogenic epidemic of prescription opioid ad-diction and overdose. Med Care 2016;54:426–9.
2. Rudd RA, Aleshire N, Zibbell JE, Gladden RM.Increases in drug and opioid overdose deaths—United States, 2000–2014. MMWR Morb MortalWkly Rep 2016;64:1378–82.
3. Report of the International Narcotics Control Boardfor 2015. United Nations Publication Sales No.E.16.XI.1. New York: United Nations. March 2,2016.
4. Chen JH, Humphreys K, Shah NH, Lembke A.Distribution of opioids by different types of Medi-care prescribers. JAMA Intern Med. 2016;176:259 – 61.
5. Matthias MS, Parpart AL, Nyland KA, et al. Thepatient-provider relationship in chronic pain care:providers’ perspectives. Pain Med 2010;11:1688–97.
6. Dowell D, Haegerich TM, Chou R. CDC guidelinefor prescribing opioids for chronic pain—UnitedStates, 2016. Recommendations and Reports 2016;65:1–49.
7. Sullivan M, Hudson T, Bradley CM, et al. Na-tional analysis of opioid use among veterans. Pro-ceedings of the 30th Annual Meeting of the Amer-ican Academic of Pain Medicine; 2014 March 6 –9,Phoenix AZ.
8. Von Korff M, Dublin S, Walker RL, et al. Theimpact of opioid risk reduction initiatives on high-dose opioid prescribing for patients on chronic opi-oid therapy. J Pain 2016;17:101–10.
9. Trescott CE, Beck RM, Seelig MD, Von Korff MR.Group Health’s initiative to avert opioid misuse andoverdose among patients with chronic noncancerpain. Health Affairs 2011;30:1420–4.
10. Ladden MD, Bodenheimer T, Fishman NW, FlinterM, Hsu C, Parchman M, Wagner EH. The emerg-ing primary care workforce: preliminary observa-tions from the primary care team: Learning fromeffective ambulatory practices project. Acad Med2013;88:1830–4.
11. Krebs EE, Lorenz KA, Bair MJ, et al. Developmentand initial validation of the PEG, a three-item scaleassessing pain intensity and interference. J Gen In-tern Med 2009;24:733–8.
12. Krebs EE, Bair MJ, Damush TM, Tu W, Wu J,Kroenke K. Comparative responsiveness of painoutcome measures among primary care patientswith musculoskeletal pain. Med Care 2010;48:1007–14.
13. Solberg LI. Improving medical practice: A concep-tual framework. Ann Fam Med 2007;5:251–6.
doi: 10.3122/jabfm.2017.01.160183 Clinic Re-Design for Opioid Management 51
1
An Introduction to Population Health and Use of Registries to
Improve Outcomes for your PatientsDec 5, 2018
2
▪ Click the arrow to expand or collapse this pane
▪ Click the hand to “raise your hand”
▪ Click Questions to expand or collapse the Questions box
▪ Messages and links from staff will appear here
▪ To share a comment or question with staff, type here and click Send. (Please keep in mind: at our discretion staff may address your comment/question publicly during the webinar, privately during the webinar, or privately at a later time.)
Webinar Logistics: Asking Questions
3
▪ If you plan to speak, join the webinar via computer or the
GoToWebinar smart phone app.
▪ If joining audio by dialing the number (not the IOS or Android app), you
will need to enter your “Audio PIN” to speak. Otherwise, you will be in
listen-only mode.
▪ “Raise your hand” in the GoToWebinar control panel when you want to
talk; the meeting organizer can unmute your line.
▪ Check that your computer or phone is not muted, and talk.
Webinar Logistics: SpeakingIf you would like to ask a question directly please ‘raise your hand’ and we will unmute you to ask your
question
4
An Introduction to Population Health and Use of Registries to
Improve Outcomes for your Patients
Jeff Hummel, MD, MPH
Michelle Glatt, MPH, PA-C
Dec 5, 2018
5
Goals for this session
By the end of this session, hopefully you will understand
• the essential role registries play in population management
• the range of tools available for creating registries
• the content of registries, who uses them, and how they are used
6
Introduction: Why are we talking about Population Health?
2008 - Institute for
Healthcare
Improvement
Triple/Quadruple
Aim
Population Health
Patient Experience
of Care
Care Team Satisfaction
Per Capita
Cost
High Value Care
7
Step #1: Defining the Population
Who are the patients/clients?
Which communities do they come from?
What barriers to health might they have?
What health conditions might they need treatment for?
What services might they need to maintain optimal health?
Patient Registry
8
The one way to define a population that ISN’T really practicing population health:
Defining the population as (only) the patients/clients that walk through the
doors of your organization that day
9
Registry Basics
10
Registries
What is a registry? ▪ A list of all people in a specific population
▪ Key information on the care of each person
What is a registry used for?▪ Identify action the care team need to take
Who uses a registry?▪ Care team member/Care or Case Manager
When is a registry used?▪ At an interval determined by the population
11
The Importance of Registries
• Ability to see key health parameters of an entire population in a single view
• The value of care is determined by outcomes and cost at a populations level
• If we’re going to manage care we need to see the care gaps must be closed to
▪ Improve clinical outcomes
▪ Reduce avoidable costs due to ED use and hospitalization
12
13
Data Definitions Used to Define Target Population
• Demographic data: Age, Gender
• Health condition: Problem List entry
• High Risk Medications: opioids, warfarin
• Care transitions: ADT data from Hosp/ED
• Pay attention to criteria for entering and leaving the population
▪ Some turn over slowly, some rapidly
▪ Criteria different for each target population
14
Entering and Leaving a PopulationLength of time In
the PopulationEntering
Definition
Leaving Definition
Diabetes Dx Code Long Term Lv Practice
Depression PHQ-9 Episode of Care Pt Choice
Anti coagulation Rx/Lab Episode of Care D/Ced Rx
Children 0 - 5 Birth In age range 5th Birth Day
Pregnant Women 1st PNV Episode of Care PP Visit
D/C Planning Admission In-Pt D/C
High Need/High Cost > 5 ED/Yr Indeterminant CBO Cert
Cognitive Impairment MMS < 25 Long term Lv Practice
15
There are Many Ways to Make a Registry
• Spread Sheet▪ Enter into Excel by hand
▪ Patient List functionality in EHR export to Excel
• Canned reports from EHR
• Queries run from EHR data base
• Third Party Registry Vendors
• The right way for the situation depends on▪ Available technology
▪ Available skills
▪ Cost
16
16
• Who has F11 on problem list?• Presence in registry• Who has received > 2 opioid refills
within past 90 days?
Documentation at last visit of• Morphine Dose Equivalent• Pain Enjoyment General Activity
(PEG) Score• Prescription Drug Monitoring
Program
Whose MDE is > 50/90?Who is also on sedative/hypnotic?Who has had a fall or near fall in the last month?
• Follow up MDE in 1 month?• Follow up MDE improving?• If not improving next step?
Who have we assessed?
What did we find?
Who is in the target population?
What did we do?
Care Gap 1
Care Gap 2
Using a Registry to Manage a Population
17
Reporting from a Registry to Track Improvement
• The denominator is everyone in the population
• Care gap 1 is difference between people who have been assessed (numerator) everyone in the population (denominator)
• Care gap 2 is the difference between people meeting the clinical goal and the total number assessed
• A registry gives us a way to track care gap closure over time
18
Who is in the Target Population?
0
50
100
150
200
250
300
350
400
450
500
550
Jan-04
Feb-04
Mar-04
Apr-04
May-04
Jun-04
Jul-04
Aug-04
Sep-04
Oct-04
Nov-04
Dec-04
Jan-05
Feb-05
Mar-05
Apr-05
May-05
Jun-05
Jul-05
Aug-05
Nu
mb
er
Number of Anticoagulation Patients
19
Who Have We Assessed?
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
Jan-04
Feb-04
Mar-04
Apr-04
May-04
Jun-04
Jul-04
Aug-04
Sep-04
Oct-04
Nov-04
Dec-04
Jan-05
Feb-05
Mar-05
Apr-05
May-05
Jun-05
Jul-05
Aug-05
Perc
en
tPercent with INR Monitored within 35 days
20
What did we find?
40%
42%
44%
46%
48%
50%
52%
54%
Jan-04
Feb-04
Mar-04
Apr-04
May-04
Jun-04
Jul-04
Aug-04
Sep-04
Oct-04
Nov-04
Dec-04
Jan-05
Feb-05
Mar-05
Apr-05
May-05
Jun-05
Jul-05
Aug-05
Perc
en
tPercent with INR at Target Range
21
Use Case 1: Managing a Chronic Opioid
Therapy Population in Primary Care
22
Setting
• Primary Care Clinic part of a 5-clinic system managing patients on Chronic Opioid Therapy using 6 building blocks approach
• The clinic does not present itself as an opioid treatment facility
• At least 1 provider in each clinic is trained on buprenorphine induction and management
• They do not encourage referrals from outside their delivery system
• Once induction is complete, patients are managed by their own PCP
• Purpose of Registry: Management of population to prevent overdose
23
Defining the Population
• Clinical Concept: People on chronic opioid therapy
• Data definition:▪ Add to registry if
–F11 code on Problem List and > 2 Opioid refills in PDMP in quantity sufficient for 30 days in past 3 months
–Positive answer to opioid screening question on SBIRT
▪ Remove from registry if Chronic Opioid Therapy managed elsewhere–F11 Code not on Problem List, or
–Patient is has received no refills for opioid in past 90 days
24
Information to include in the Registry
• Demographic information
• Data to identify care gaps
▪ Date controlled substance contract signed
▪ Date last office visit/contact of any type
▪ Morphine Dose Equivalent MDE - date & value
▪ Date PDMP last checked
▪ Concurrent sedative hypnotics
▪ Medication Assisted Therapy status
▪ Date of last urine drug test - date & value
25
How the Registry is Used
• Chronic Disease Population Manager uses the registry on a weekly basis to find care gaps▪ Follow a protocol in closing care gaps
▪ Confer with care team
▪ Outreach to patient
▪ Scheduling follow up
▪ Referral to community-based resources
• Medical Assistants use registry on daily basis in prep for huddle patients on schedule
26
Use Case 2: Managing People with Opioid Use
Disorder Receiving Care at Behavioral Health Clinic
27
Setting
• Behavioral Heath Agency without co-located primary care services
• Formal referral agreement with FQHC that has capacity for Medication Assisted Therapy for Opioid Use Disorder
• Community-based resources for treatment of other Substance Use Disorders
28
Defining the Population
• Clinical Concept:
• Client receiving care at Behavioral Health Agency who has an Opioid Use Disorder
• Data Definition:
▪ Add to registry if:
–Opioid use disorder documented in chart (Dx code)
–Positive answer to opioid screening question on SBIRT
▪ Remove from registry if:
– Successfully treated and off opioids (removed Dx code)
29
Information to include in the Opioid Use Registry
• Demographic information
• Data for decision-making• Currently using – within past 7 days: Yes/No
• If yes, has Naloxone
• If yes, willing to participate in Substance Use Disorder treatment including Medication Assisted Therapy
• If Yes, Change Readiness Assessment Using Prochaska model
• If in Action Stage then referral for Opioid Use Disorder
• Follow up to determine if referral completed
30
How the Registry is Used
• Care Manager reviews registry clinicians on weekly basis regular basis to identify
• People for whom substance use disorder assessment to be done at next visit
• People needing Naloxone
• People due for change readiness assessment
• People in contemplative stage: date follow-up evaluation scheduled
• People in action stage: date & status of referral for Substance Use Disorder treatment
31
Other Behavioral Health Populations to Consider using a Registry
• Population defined by• Suicidal Ideation
• Eating Disorder
• Self Harming Behavior
• Registry use• Who is in the target population
• Who has been assessed for severity using a validated tool
• Tracking improvement using validated tools
• Identifying people lost to follow up
32
Use Case 3: Managing Depression in Adults with Diabetes in a Primary Care
Clinic with the Collaborative Care Model
33
Setting
• Primary Care Clinic part of a 5-clinic system
• Each clinic has:
• A centralized diabetes management program
• An onsite social worker who is a member of a collaborative care BH team with other team members at local BHA
34
Defining the Population
• Clinical Concept: Adults with diabetes
• Data definition:
• Diagnostic code for diabetes on problem list
• Process for identifying diabetes patients without problem list entry
35
Information to include in the Registry
• Demographic information
• Evidence-based data for decision-making
• Date & value most recent depression screen with PHQ-2 and if positive PHQ-9
• If PHQ-9 > 10 date of follow up visit
• Date & value most recent HbA1c
• Date & value most recent BP
• Dates and values for other evidence-based monitoring standards for diabetes
36
How to Use the Registry
• Care Manager reviews registry on weekly basis to identify people due for assessment:
• Depression screening: everyone is screened at least once a year
• Everyone with PHQ – 9 > 10 gets an evidence-based intervention using shared decision-making
• PHQ-9 follow up – everyone with PHQ – 9 > 10 is scheduled for 1 month follow up to see if the intervention worked
• Reminders to Primary Care teams about patients on schedule for the day and gaps to close during the visit
37
Improved Management of Depression
0
10
20
30
40
50
60
70
80
90
16-Apr 11-May 15-Jun 27-Jul 12-Oct
Denominator
PHQ screen
PHQ9 score >10
Scheduled Followup
Referral success
Tota
l nu
mb
er
of
Pati
ents
wit
h D
iab
etes
38
Now that we’ve seen some use cases lets go back to basics:
Who are the patients/clients?
Which communities do they come from?
What barriers to health might they have?
What health conditions might they need treatment for?
What services might they need to maintain optimal health?
Patient Registry
39
Registries are the foundational step for both population management and quality improvement activities
Collect information
Organize patient by
patient information
(a registry!)
Plan for Updates
Patient outreach and
care coordination
Driving Performance
for better health
outcomes
Improving care by deciding as an organization what good care looks
like and transparently driving toward that goal over time via systems
improvements
Improving care by focusing on a set of services, tests, or goals that an
individual patient might need/want
Basic Population Management
Basic Quality Improvement
40
Population Management & Quality Improvement
Patient outreach and
care coordination
Driving Performance
for better health
outcomes
Basic Population Management
Basic Quality Improvement
Define an ideal set of goals for members of the populationSelf-Management Goals Set?
Community and Social Support Linkages Made?Medical Referrals Complete?
HbA1c tests, Blood pressure, INR complete/controlled?Follow-ups completed as recommended?
The percent of patients in this population that have met the goalsShare transparently by care team, location, organization, county, state
Use to measure systems changes made using some kind of process improvement or quality improvement methodology
WhaWhaW
Organized Evidence
Based Care
41
Multi-Condition
• Patients with Diabetes, Depression and hypertension
Unengaged
• Patients without colon, breast or cervical cancer screening
• Patients assigned by the MCO without establishing care
Social determinants of Health
• Patients with housing insecurity, food insecurity, and/or transportation issues
Risk
• Using a risk calculation to determine which of your patients might need the most support
Multi
• Patients with housing insecurity with diabetes, depression, and an elevated risk score
Moving beyond traditional registries -Identifying Complex Populations
© Qualis Health, 2016
It is often the case that we are not collecting the
information that we actually think is
important in a systematic way.
42
Summary• Registries help care managers see a target population and
care gaps for the population. • They require data definitions
• Membership in the population• Looking for gaps in an evidence-based standard of care• Follow up to see whether gaps were closed
• Registries require a population health infrastructure• The technology can be simple or complex• It requires a team approach to care with someone dedicated
to the job of • maintaining the accuracy of information in the registry• using the registry to guide evidence-based actions to close care gaps