Community Diabetes RegistryKey Points

1
Community Diabetes RegistryKey Points

• Wells Shoemaker MD
• PI, Community Chronic Care Network, Santa Cruz
  County, CA

2
Collaboration

• Very rare in USA to have plausible chance for
  public/private sector collaboration for entire
  population. Usually piecemeal.
• HIP Council, Regional Diabetes Collaborative set
  the stage
• IOM invitation (Jan, 2004) validated the merits
  of our early steps
• AHRQ looking for expandable models, invested 1.3
  Million taxpayer funds

3
Who Needs It?

• We have approximately 18,000 people with diabetes
  in our County.
• Many will die early from complications we
  theoretically can delay or prevent
• It costs 120 million per year here, mostly for
  late care. That number will go up.
• Many more people are going to develop diabetes in
  the next 5-10 years, and were not ready to care
  for them.

4
Dont we already have good doctors? Sure.

• Were good locally, really good, but mostly with
  the patients we touch. We may be missing a third
  to a half of our opportunities.
• To turn this epidemic around, we need to get
  people connected with efficient systems
  delivering relatively simple care
• We want to bring them all into the light

5
Chronic Care Model

• We have all seen the diagram.
• The CCCN registry is mostly a provider tool gt
  the right hand arrow.
• We expect the registry to stimulate and support
  activities in the patient activation area, as
  well as in the larger community circles on top.

6
(No Transcript)
7
Chronic Care QI4 Basic Steps

• Collect clinical data for patients with defined
  need, i.e. diabetes
• Deliver that data to the point of carefast,
  accurate, easy, reliable
• Give performance feedback to providers
• Recognize and reward excellence AND improvement
• Share and spread best practices locally

8
TerminologyRegistry is tricky

• Register for the draft, register aliens, register
  HIV patientsall have unpleasant Big Brother
  sensitivities in 2004.
• Were using the term technically, not
  politically. A registry is an accurate list of
  patients with a given condition, with associated
  clinical information which supports better
  decisions for those people.

9
Are all registries the same? No.

• Variable in horsepowersimple list of names
  list up to sophisticated hand-holding, paper
  lists up to elite web tools intertwined with
  practice management
• Some entrepreneurial, some public domain
  variable flexibility and cost
• Variable responsiveness to end-users
• Variable ability to generate reports for small
  and large population samples

10
4 Levels of Registry Use

• Point of Care liveMaximum utility
• DashboardCollate and display core information
  before the patient arrives
• Reports at the practice siteoverdue and high
  risk alertsallows staff to react
• Reports of large populationsrecognize providers
  and patients with needs, measure effectiveness of
  our efforts

11
Three Registries in SC County

• PMGhomegrown, web-based POC tool in use since
  11/03
• 4 data inputs Claims, lab, pharma, and POC
• Road tested in multiple disparate practices
  much feedback from widespread demos
• Built-in real-time report-generating capability
• Penetrated to approx half of PMG adult PCPs
• Already see performance stratification between
  user and non-user PCPs
• Feedback financial incentive for PCP use
• Only HMO patients included

12
Registries in SC County 2

• SCMFDashboard tool for decision support
• Efficient data gopherclaims and labbut
  limited POC input and interactivity
• Paper tool with daily updatewill later blend to
  Epic EMR
• Internal data system generated, not web
• Well penetrated with providers, consistent format
• Avail for all SCMF patientsnot just HMO

13
Registries in SC County 3

• AllianceDashboard tool for Medi-Cal
• Includes claims, lab, pharmaall 3 from
  reliable, local data store
• Hard copy deliverydated, not web-based
• Requires significant end-user process
• Probably most important opportunity for
  population care improvement

14
Why Did We Choose 1

• Web accessibility with advanced POC features
• Report generating capability built in
• Established functionRD, beta testing done
• Local control of architecture Flexibility
• Ability to incorporate new data streams
• HIPAA confidence
• Portability for patients moving within county
• Offered as community service gesture

15
Registry vs. EMR

• A registry is a proven low cost, high utility
  device with low barriers to adoption
• EMR or EHR is high cost, high utility with high
  barriers and history of false starts
• EMR can do more comprehensive work, but .
• Limited deployment and portability now
• Registry ready for immediate service

16
Isnt Registry a Dead End? No!

• We fully embrace the evolution of EMRs locally,
  and well build the registry so that it will
  automatically and painlessly cross-populate
  information.
• Its probably going to be 5 years before EMRs are
  serving gt50 of local residents.
• The registry can achieve significant reductions
  in morbidity for diabetes in 2 years.

17
Synergies

• SCMF and PMG both in pioneer CA
  CollaborativeBreakthroughs in Chronic Care
  Programmodeled after ICSI, using IHI type rapid
  cycle improvement
• Allianceleadership commitment for enrollees for
  preventive chronic care
• RDC thrust for clinical care, education, policy
• Harmonize efforts to confront health disparities
  by elevating performance for ALL

18
Adoption BarriersPMG Early

• Serves HMO-onlyless than 50 of patients?poor
  incorporation into office workflow
• Not portable for patients moving to PPOs
• Redundant work for few practices with full EMRs
• Vulnerability to financial and contractual
  vicissitudes

19
Adoption Barriers Expected 2005

• Change of any kind creates resistance
• Variable computer skills and openness to
  incorporation of electronic tools into work flow
  and record keeping at different sites
• Trust and Disclosure
• Security questions
• Staffing

20
Technical Barriers

• Interfaces customized to payor and administrative
  data inputs
• Multiple lab interfaces with variable cooperation
• Pharma data from multiple sources
• Unique patient identifier challenges
• End-user education and support

21
Outside Buy-In ChallengesNot Yet Solved

• Medicareboth regional and federal sign-off for
  claims information. No central pharma
  recordswould need local pharmacy cooperation.
  Lab multiple local.
• PPOs6 or more separate negotiations

22
What Order?

• Alliance first updata already mature,
  disadvantaged population. Inflow to Safety Net
  Clinics, SCMF independent PCPs likely before
  6/2005
• Medicare (estimate 35-40 of diabetes)
• SCMFcurrently well served, less need
• PPOsprobably less needy population, but boost
  for penetration into PCP practice

23
Whats Next?

• Add other related illnesses, i.e. vascular
  disease
• Consumer access to same database
• Sustainability in SC County 2007 will require
  payor contributions to protect savings derived
  from reduction in costly complications
• Conceptually appealing spread to Alliance
  population in safety net clinics in Monterey
  County. Will need funding to support expansion

24
Whos Watching?

• Big Brother. AHRQ has close oversight and
  expectations for publishable results.
• Commercial insurorsif local collaborative can
  prove effective, disease management funds can be
  directed to local group instead of remote vendors
• Our community
• Our families affected by this illness

25
Why, Again?

• We are competing with an expanded sense of the
  possible, not with each other.
• We can. We Must.
• People are counting on us.