Oregon Health Leadership Council: High Value Patient Centered Care Model The Clinical Performance Improvement Network

1
Oregon Health Leadership CouncilHigh Value
Patient Centered Care ModelThe Clinical
Performance Improvement Network
September 19, 2012

• Denise L. Honzel
• Executive Director

2
Oregon Health Leadership Council--History

• Commissioned by the business community in the
  summer of 2008
• PurposeDevelop solutions and actions to keep
  health care costs and premium increases closer to
  the CPI
• Incorporated as a 501(c) 6, statewide membership
  organization in 2010
• 14 member Board of Directors
• 29 Council members 7 major medical groups 8
  hospitals/health systems 12 local/national
  health plans Oregon Assoc. of Hospitals and
  Health System and Oregon Medical Assoc.
• Director of Oregon Health Authority (ex-officio)
  
• Quarterly reporting back to the business
  community

3
Areas of focus for the Council

• Initially, four committees to oversee work
• Value Based Benefits
• Evidence Based Best Practicehigh cost imaging
  low back pain elective deliveries before 39
  weeks
• Administration Simplificationsingle sign on
  standardized EDI for claims, eligibility best
  practicesplan websites, pre-authorization,
  credentialing
• Reimbursement and Payment Reformmedical home
• In 2011 began work on a major effort focused on
  re-designing care delivery and financing for the
  Medicaid and uninsured populations--now moving
  forward as Community Care Organizations (CCOs)

4
High Value Patient Centered Care
Demonstration--Timeline

• In early 2009, the Payment Reform group
  recommends a multi-payer medical home
  initiative focused on high risk adult population
• Reviewed the successful model of Boeing that
  delivered lower costs (up to 20), higher quality
  and patient satisfaction, improved employee
  productivity with results beginning within 12
  months
• In fall of 2009, hired Dr. Pranav Kothari,
  Renaissance Health with a goal to launch the
  initiative within a year, learn along the way
• Convened medical group and health plan advisory
  committees to develop the care and financing
  models

5
Hypothesis Care Model and Outcomes

• Focus on the to 10 highest risk patients
• Intensive care management with RN Care Manager in
  the medical office, following a defined care
  model would produce the greatest impact on
  care/costs
• At the end of the 2 year demonstration, results
  would show
• Lower overall costs compared to a matched control
  group with reduced ED visits, reduced inpatient
  admits/days and initial increase in PCP visits
• High quality health outcomes, high levels or
  member and provider experience

6
Hypothesis Process Needs

• Multiple payers to get the size
• Common model of care
• Common payment model and standard contract for
  all plans to use
• Committed medical groups
• Scientific approach to the evaluation
• Collective funding to support infrastructure
• Common set of health plans policies and practices
  to implement

7
Common Care Model Components

• Access
• Dedicated care manager for each patient
• 24/7 access for urgent care
• Access to care team via email/phone
• Facilitate access for non-MD services, integrated
  w/PCP
• Care Delivery Elements
• Rules-based, care planning and management
• Intentional, integrated care coordination
• Recognition of social and behavioral health needs
• Information
• EHR, registries, quarterly feedback for the team
• Movement towards broader data transparency

8
Common Care Model Components

• Coordination
• Care transition management (post ED, hospital)
• Medical neighborhood of specialists, service
  agreements
• Care giver and social support systems
• Intensive Care Management
• Motivational interviewing, readiness assessment
• Team based pre-visit planning, systematic Rx
  review
• Team huddles, group visits
• Advanced directives, end-of-life care programs
• StaffingRN as a team lead with support from the
  care team one RN for 200 patients enrolled

9
Consistent Payment Approach

• All payers use the same methodology same
  contract
• Upfront PMPM for participating patients to pay
  for the RN care coordinator
• Standard fee-for-service for medical care
  provided
• Shared savings between the medical groups and
  payers for achieving savings and quality metrics,
  paid at the end of the demonstration
• Quality measuresmaintain or improve and achieve
  Quality Compass at 50th percentile for diabetes,
  hypertension, cholesterol management measures
  patient satisfaction administer pre, post
• Cost savings50 share up to specified limit
  based on difference over difference of
  intervention group vs. tightly matched control
  group

10
Rolling Out the Demonstration

• Early 2010, 8 health plans and state expressed
  interest
• Plans identified 25 medical groups statewide that
  they felt could deliver on the model
• OHLC issued an RFP
• Staff interviewed each group
• Fourteen medical groups selected a 15th group
  met qualifications except not the volume needed
• Due to volume or challenges with data, 5 health
  plans and state agree to move forward

11
Medical Groups Selected
A









Adventist (Portland) High Lakes (Bend)
Legacy (NW Port./Tualatin Medford Medical (Medford)
NW Primary Care (Portland) No. Bend Med. Ctr. (Coos Bay)
OHSU (Portland) Oregon Med. Group (Eugene)
PeaceHealth (Eugene) Portland Clinic (Portland)
Portland Family Practice (Portland) Providence (Portland/Newberg)
Tuality (Hillsboro) Westside Internal Medicine (Beaverton)
12
Health Plans and State Groups
Health Net ODS
PacificSource Providence
Regence PEBB (State employees)
OEBB (School employees) OMIP (High risk pool)
DMAP (Medicaid FFS)
13
Implementation

• July 2010, each health plan began negotiating
  with each medical group30 day target, finished
  within 45 days
• During that time, plans began identifying
  patients to be invited into the demonstration,
  using plan predictive risk models
• Top 10-15 highest risk adult patients
• All complex and chronic included
• Asked plans to exclude patients with ESRD,
  transplants, OB, Trauma with no other co-mordity
• In September 2010, medical groups hired 23 RNs
• In October 2010, 23 RNs participate in a four
  day training

14
Implementation

• Beginning October 1, 2010, using the lists
  provided by the plan, nurses start to invite
  patients to participate-- original goal to have
  all enrolled by end of December, extended until
  early March 2011
• Intake visits and care plans completed early in
  2011
• Baseline quality measures for each patient
• LDLs, A1c and blood pressure reported
• Patient satisfaction surveys (modified national
  survey) completed
• Early 2011, began quarterly meetings on the
  medical group and health plan project managers
  and care managers

15
Where we are now

• 3,600 patients voluntarily enrolled
• In July 2011 began, reporting quarterly
  utilization reports to medical groups, from
  claims data
• ED of visits, avoidable rates, diagnosis, by
  day of week, by member
• Inpatient of admissions, by DRG,
  discharge/admission day of week, by member, LOS
• Rx Drugs filled, Imaging, Lab
• Working with Oregon HealthCare Quality Corp and
  Milliman to refine the medical group reporting
• Producing summary utilization reports for OHLC
  and demonstration participants

16
Where we are now

• Additional 1.5 day training for the care managers
  in Sept. 2011
• Agreed to extend the program through Feb. 2013 to
  allow for input from physicians on patient
  selection minimize churning
• Second enrollment period occurred Jan.-March 2012
  to address attrition
• Plans provided new lists nurses more selective
  in choosing those to invite
• Contracted with Dr. John McConnell, OHSU for the
  shared cost savings evaluation Dr. Meredith
  Rosenthal, Harvard University, Dr. Kothari,
  Renaissance Health will be advisors.
• In February 2012, weekly Office Hours with care
  managers
• Working on a sustainable model, post
  demonstration

17
Results to date

• While promising, its too early to make
  conclusions not comparing to a control group,
  some enrollment data need reconciliation
• Looking at utilization data through March 2012,
  compared to the pre-intervention time period, we
  are seeing
• Downward trends in emergency room visits,
  inpatient admits and inpatient days/1000
• Office visits for primary and specialty care
  initially stayed flat or increased slightly then
  decreased (phone/email not included
• Reduced use of RX for chronic conditions
• Slight downward trends in imaging
• Final cost savings wont be available until the
  end of the demonstration and analysisfall 2013

18
Learnings Medical Group-Plan Partnership

• Medical groups and Plans able to move beyond
  contracts and work creatively togetherfor data,
  care delivery
• Medical groups and Plans appreciate personal
  interactions, in person meetings
• Medical groups appreciate requests for input
• Communication beyond direct participants involved
  has been uneven for all parties sometimes leading
  to confusion, complications and eroding
  partnerships
• Medical groups and Plans challenged to manage new
  billing processes
• Study design (control group, enrollment, etc)
  limits real-time adjustments to care model and
  patient selection

19
Learnings Data Transparency/Exchange

• Exchanging accurate, timely data between plans
  and medical groups can impact care decisionsneed
  timely ED and hospital admit/discharge data
• Challenge in the data exchange accuracy, timing,
  capabilities to manage/use data
• Risk models are a good starting point to
  identifying risk need mechanisms to include
  clinical intelligence not all the right
  patients
• Attribution capabilities uneven
• Need for more real-time assessment of
  progressclaims lag not suited for dynamic
  interpretation

20
Learnings Medical Group Culture/Readiness

• Some medical groups fully engaged others less so
  difference in readiness not small vs. large
  group, but cultural
• Front-line MD compensations still almost entirely
  productivity based these efforts largely
  perceived as noise by physicians
• Physicians appreciate care manager support,
  especially after seeing patient improvement
• Support for care managers varies, depending on
  clinic buy in

21
Learnings Care Model

• Patients appreciate care management from their MD
  team very high engagement when advocated by PCP,
  well-accepted by clinic
• Non-visit based care (especially email) critical
  to serving this population needs and for care
  manager panel management
• Early, regular communication with patients
  surface unique needs
• Initial intake assessments yielded deep patient
  discussions, revelations around barriers
  although difficult to schedule
• Workforce needs skill building to engage patients
• Health Plan disease management can be helpful
  partner with medical group care management
• Gap around education resources for patients,
  behavioral health, pharmacy, community resources

22
Next Steps

• As demonstration continues, further work on the
  processes to support the model of care
• Continue collaboratives to share best practices
  and learning's
• Complete work on a sustainable model for the long
  term - care model, roles/responsibilities,
  reimbursement models
• Measure results

23
For further information

• Denise Honzel, Oregon Health Leadership Council
• Denise_at_orhealthleadershipcouncil.org
• Dr. Pranav Kothari, Renaissance Health
• pk_at_renhealth.net