DemandDriven Research: Working through DeliveryBased Networks

1
Demand-Driven Research Working through
Delivery-Based Networks

• Irene Fraser, PhD, Director, CDOM
• AcademyHealth Annual Research Meeting
• June 8, 2004

2
RAND Study Quality of Health Care Often Not
Optimal

• Doctors provide appropriate health care only
  about half the time

Percentage of time
E. McGlynn, S. Asch, J. Adams, et al., The
Quality of Health Care Delivered to Adults in the
United States, N Engl J Med, 2003
3
To Improve Health Care, Decision-Makers Need and
Want Evidence

• Clinical decisions - EB Medicine
• What to include in drug formularies
• How to define medical necessity
• Management decisions -- EB Management
• How to pay hospitals, physicians, nurses
• How many/what kind of staff to recruit
• With whom to merge, affiliate, contract
• How to organize processes of care
• What kinds of IT systems to purchase
• Policy Decisions EB Policy

4
Original research
18
variable
Negative results
Dickersin, 1987
Submission
46
0.5 year
Kumar, 1992
Koren, 1989
Acceptance
Negative results
0.6 year
Kumar, 1992
Publication
1714
Expert opinion
35
0.3 year
Poyer, 1982
Balas, 1995
Lack of numbers
Bibliographic databases
50
6. 0 - 13.0 years
Antman, 1992
Poynard, 1985
Reviews, guidelines, textbook
9.3 years
Inconsistent indexing
Implementation
17 yrs to turn 14 of original research to the
benefit of patient care
5
A Particular Problem for Evidence-Based
Management

• EB Medicine
• More targeted goal (improve practice)
• Researchers and users are both scientists
• Read common journals
• Patients the unit of analysis

• EB Management
• Multiple goals (quality, efficiency, profit)
• Managers not scientists
• Managers and scientists read different things
• Organizations the unit of analysis (N problem)

6
Lessons from Marketing

• First rule of marketing
• Build product to fit the need
• Second rule of marketing
• To have the right product, may need to change the
  way you make it

7
Identifying Need What Do Decision-Makers Want?

• Six user-centered meetings
• Small (20-30) groups
• Combination of industry leaders, purchasers,
  policymakers, researchers
• Discussion at several NAC meetings
• Focus on both content and dissemination/use
• Five lessons for HOW we do our work

8
1 Design Studies that Answer User Questions

• Move from description to prediction and
  explanation
• Focus on independent variables that are
  modifiable
• Provide details on HOW to implement

9
2 Present Findings in Decision-Makers Time and
Space

• Define evidence as they do
• Perhaps redefine when something is ready for
  prime time
• Limit caveats
• Be willing to generalize to broader universe
• Add tools to our list of products

10
3 Change Incentive Systemsfor Researchers

• Currently, researchers
• Get tenure by publishing in journals managers
  dont read
• Get credit for teaching other researchers, not
  managers
• Get grants for academic research, not applied,
  translational research
• Find it easier to publish quantitative, not
  qualitative, research

11
4 Build User-Researcher Collaborations and
Dialogue

• Provider-based networks
• Challenge grants
• Organizational researchers with management
  experience
• Managers with research training
• Practitioners in classrooms

12
5 Change Dissemination

• Let early adopters, not researchers, educate next
  wave of users
• Give presentations at trade association meetings,
  not just research meetings
• Collaborate with, mimic, or work through
  consultants
• Find or create equivalent of Harvard Business
  Review for health care
• Expand funding for implementation research

13
Bonus Tip Dont Over-Complicate Findings

• Winners of Washington Post Contest on Ways to
  Make Life Harder
• Supermarkets arrange goods alphabetically
  lettuce, lamb and lysol are now contiguous
• Thursdays become drive-on-the-left-side-of-the-ro
  ad day
• Subdivide time zones into 60 parts, observing
  local times to the minute
• Alphabetize phone books by first names

14
SUPPLY-DRIVEN MODEL
DECISION-MAKERS
RESEARCHERS

• Questions
• Hypotheses
• Grant Applications
• Study
• Writing
• Re-writing

PUBLICATIONS
Leadership
Politics
Evidence
Culture
15
SUPPLY-DRIVEN MODEL WITH MARKETING
Knowledge transfer
DECISION-MAKERS
RESEARCHERS

• Questions
• Hypotheses
• Grant Applications
• Study
• Writing
• Re-writing

PUBLICATIONS
Leadership
Politics
Evidence
Culture
16
DEMAND-INFORMED MODEL
User Needs Assessment
Knowledge transfer
DECISION-MAKERS
RESEARCHERS

• Questions
• Hypotheses
• Grant Applications
• Study
• Writing
• Re-writing

PUBLICATIONS
Leadership
Politics
Evidence
Culture
17
DEMAND-DRIVEN NETWORK MODEL
DECISION-MAKERS
RESEARCHERS
Info Tools
Info Tools
PUBLICATIONS
18
3 Examples of Research Networks

• Integrated Delivery System Research Network
• Center for Health Management Research
• HIV Research Network

19
What is the IDSRN?

• 9 practice-based research partners
• Abt Associates Inc.
• Center for Health Care Policy and Evaluation
• Denver Health
• Emory Center on Health Outcomes and Quality
• HMO Research Network
• Marshfield Clinic IDS Research Consortium
• Research Triangle Institute - UNC Network
• University of Minnesota Consortium
• Weill Medical College of Cornell University
• 40 collaborators
• Work through task orders

20
Strategic Advantage Size and Breadth

• IDSRN includes
• Over 50 million patients
• Majority of US physicians
• Majority of acute inpatient facilities
• 2,250 outpatient clinics
• 450 long term care facilities
• 56 rehabilitation facilities
• 30 home health agencies
• 63 dental facilities

21
Size and Breadth Persons Served Located in All
States
22
Strategic Advantage Population Diversity

• Payer mix privately insured, Medicare (3
  million), Medicaid (2 million), uninsured (0.5
  million)
• Geographic mix urban, inner-city, suburban and
  rural (6.3 million) residents
• Demographic mix ethnic and racial minorities
  (gt5.8 million), children and adolescents (gt14
  million), persons aged 65 (7 million)

23
Strategic Advantage Data, Research,
Implementation Capacity

• Large, robust databases (e.g., administrative,
  clinical, registries)
• Clinical, demographic, geographic data diversity
• Nationally-recognized academic and field-based
  researchers
• Expertise in data manipulation, methods, emerging
  IDS policy/management issues
• Operational leadership involved in setting agenda
  and using findings

24
Strategic Advantage Speed

• From request for proposals to award 9 weeks
• Average project completed in 16 months
• Excellent mechanism for rapid-cycle work
• data availability
• expertise in data manipulation, methods
• established partner collaborations
• want solutions for own delivery system issues

25
Examples of Products in 2003

• Presentations to IDS Operational Leadership (12)
• Presentations at live/web-assisted conferences
  (50)
• Scalable, scenario-appropriate models (for local
  to international adaptation) (15)
• Training session and workshop tools (47)
• How to guides, workbooks (many!)
• Publications in peer-reviewed journals (12)
• Press releases, briefings and briefs (10)
• National conference (1)

26
Impact 830 Downloads of Weills BTP Model from
AHA Web Site AHA TA Provided
27
Impact IDSRN Results Influence Users Practice
or Policy Decisions

• Providence Health System hired transition
  pharmacist to reduce errors using an electronic
  medication list
• CMS/OMH using CLAS guides as training tools in
  MCO workshops nationwide
• Findings influenced Aetnas decision to begin
  collecting race/ethnicity data
• Evercare CEO making NP staffing decisions for
  nursing homes based on pneumonia care findings

28
IDSRN Three Types of Impact
29
Home Pagehttp//www.AHRQ.gov

• Center for Delivery, Org. Markets
• http//www.AHRQ.gov/about/cods
• Irene Fraser ifraser_at_ahrq.gov