Intermed

1
Intermed
2
Function type

• Distributed research center

3
Distributed research center
Harvard BW Decision Systems
McGill
Harvard Mass General
Stanford Medical Informatics
Columbia U
Columbia Presbyterian
University of Utah
4
Resource diagram
5
InterMed history

• NIH experiment in distant collaboration
• Prior work by all PIs
• Initial funding 1994-1997
• Outside evaluator brought in 1995
• Follow-ups funded by the Army AHRQ and NLM

6
Alternative history

• Confusion
• Frustration
• Focusing
• Productivity
• (Question for discussants Was the confusion/
  frustration part a necessary phase?)

7
Scientific Products

• Initial goals were broad
• Make progress in the field of medical decision
  support
• Improve standardization between regions
• Needed credibility of multiple sites

8
Disparate goals (incentives)

• Clinicians wanted to efficiently develop
  something that would be useful in the field
• Researchers each wanted to push the envelope in
  their area of interest. Would this project focus
  on
• CPSC research?
• AI research?
• Medical research?
• Differences in goals between locations

9
Could these challenges negotiated?

• Collaboration readiness?
• Technology readiness?

10
Collaboration readiness?

• Barriers of distance, discipline, institution
• All PIs had a family connection to Octo
  Barnetts lab

11
Technology readiness?
12
Technology used

• Email, email lists
• Teleconferencing
• Timbuktu application sharing (for
  tele-presentation)
• Ontolingua controlled vocabulary system

13
Strategic use of technology

• Conference calls
• FTF meetings

14
Conference calls

• PI-only calls were negotiation methods of last
  resort
• Efficiency of conference calls improved
  significantly over time
• Email used to prepare for conference calls
• Email used to prepare for and offload content
  from conference calls
• Role specialization was a key to increased
  productivity later in the collaboratory

15
FTF meetings

• Noted increased productivity of conference calls
  after FTF meetings

16
GLIF

• Post-1996 Coordinated effort to develop a
  machine-readable system for describing clinical
  decisions
• Settled on GLIF Guideline Interchange Format

17
Example GLIF for cough treatment
REPLACE WITH CHRONIC COUGH
18
GLIF
19
GLIF
20
Challenges still remained

• Communication gaps
• Example of unstated background knowledge by
  clinicians