Presentation on SHARP projects: Medication reconciliation, tracking medical lab tests, systematic yet flexible systems analysis, and preventing wrong patient errors. Houston, TX April 4, 2012

1. Project 4: Cognitive Information Design
& Visualization
Project Leaders: Project Co-Is: Consultants: Postdocs, GRAs,
Programmers:
Todd R. Johnson Jorge Herskovic John Flach
Ben Shneiderman Elmer Bernstram Eliz Markowitz
Catherine Plaisant Surreya Tarkan
Tiffany Chao
Project manager:
Chitra Shriram
w w w. s h a r p c . o r
g

2. Top 3 Accomplishments in
Years 1-2
 Medication Reconciliation:
Spatial Layout with Animated Transitions
 Test Result Management:
Table Design & Retrospective Analysis
 Systematic Yet Flexible Systems Analysis:
Framework for analyzing HIT interface
design

3. Medication Reconciliation:
Spatial Layout with Animated Transitions
 What’s unique?
 What’s identical?
 What’s equivalent?

4. Medication Reconciliation:
Spatial Layout with Animated Transitions
LEVELS OF EQUIVALENCE
Equivalenc Criteria Example
e
Form Identical except for Advil = Ibuprofen
brand vs. generic Senormin = Atenolol
Functional Same therapeutic intent Atenolol & Propanolol
both betablockers
Partial Form or functional equivalence, Advil 100 mg
but Acetaminophen 200mg
differ in dosage, frequency,
route
None Unique in form & function
 “Automated medication reconciliation and complexity of care transitions”
Bozzo Silva, Bernstam, Markowitz, Johnson, Zhang and Herskovic, AMIA 2011

5. Medication Reconciliation:
Spatial Layout with Animated Transitions
DEMO
Prototype by Tiffany Chao
VIDEO AVAILABLE AT
www.cs.umd.edu/hcil/sharp

6. Medication Reconciliation:
Spatial Layout with Animated Transitions
Twinlist running on Microsoft Amalga Platform Twinlist adapted for problem list reconciliation in
Contact: Hank Rappaport cancer risk assessment software (Hughes
riskApps™)
“Best medication reconciliation interface I Contact: Kevin S. Hughes, Massachusetts General Hospital
have seen” Shawn Murphy, MD, PhD, Harvard
University
“A sigh of relief when I saw Twinlist”
Melinda Jenkins, PhD, FNP

7. Test Result Management:
Table Design & Retrospective Analysis
GOALS:
Help clinicians see what needs attention
 Rich tabular displays
Facilitate taking action
 Clarify responsibility
 Embed operations to save time
Allow retrospective analysis

8. Test Result Management:
Table Design & Retrospective Analysis
Sample
Current
Design
VA
View
Alerts

9. Test Result Management:
Table Design & Retrospective Analysis
Sample
Current
Design
VA
View
Alerts

10. Test Result Management:
Table Design & Retrospective Analysis
Sample
Current
Design
VA
View
Alerts

11. Test Result Management:
Table Design & Retrospective Analysis

12. Test Result Management:
Table Design & Retrospective Analysis
 Color-Coding, Icons, and Ranking

13. Test Result Management:
Table Design & Retrospective Analysis
 Compact Layouts

14. Test Result Management:
Table Design & Retrospective Analysis
 Improved Headers

15. Test Result Management:
Table Design & Retrospective Analysis
 Integrated Actions

16. Test Result Management:
Table Design & Retrospective Analysis
 DESIGN GUIDELINES: sample of 28 in total
 Rank the table according to one or more column attribute(s), arranged
vertically down (Few, 2004)
 Use color coding of columns to show the ranking, by default the most
severe value must appear at the top of the table while ensuring that the
most important values are still visible
 Round data displayed in table cells entries where it is not misleading to
do so (MSCUI, 2008) but show more precision if space permits
 Avoid a heading that is significantly wider than the data it is indicating
(MSCUI, 2008) by splitting such headers into two or more lines
 If a certain column always has the same value, it could be removed to
save space
 Keep table structure consistent from table to table, for example, distinct
columns in tables should be placed at the rightmost end to allow for
alignment of the same columns on the left side

17. Test Result Management:
Table Design & Retrospective Analysis

18. Test Result Management:
Table Design & Retrospective Analysis

19. Systematic Yet Flexible Systems Analysis:
Framework for analyzing HIT interface
20 design
 Systematic, consistent approaches
 Can improve
 Efficiency, Safety, Effectiveness
 Examples
 Standard operating procedures, Clinical guidelines
 Decision support, Hard stops in EHRs
 But flexibility is needed to accommodate
variation

20. SYFSA in a Nutshell
21
Idealized:
Logical
constraints on
the work,
independent
of any system System:
A redesigned
system that
matches logical
constraints as
closely as
Natural: How possible
the current
system
supports and
constrains the
work

21. Idealized Space for Medication Ordering
 Procedural Flexibility: 3 bits
 Functional Flexibility: Approximately 14.58
 Estimate from RxNorm: ~ 24,000 prescribable
drugs, including those not in RxNorm
Refills Quantity Form Frequency
Confirm Dose Name
Quantity Refills Frequency Form

22. Open Vista Natural Space

23. OpenVista: Natural Space
OpenVista –
The Natural Space
Too much procedural
flexibility: 9.5 bits

24. Natural Space of “Droogle” By Peter V. Killoran M.D
an e-Prescription demonstration project inspired by Google
DEMO
 Droogle
 Procedural
Flexibility close to
ideal: 1 bit
Dose Form
Ingredient Quantity Frequency Duration Refills Confirm
Form Dose

25. RxTerm Medication Entry Demo App
from The National Library of Medicine, NIH
From: http://rxterms.nlm.nih.gov:8080/
Keystrokes: war

26. RxTerm Medication Entry Demo App
from The National Library of Medicine, NIH
Keystrokes: war TAB

27. RxTerm Medication Entry Demo App
from The National Library of Medicine, NIH
Keystrokes: war TAB 5

28. RxTerm Medication Entry Demo App
from The National Library of Medicine, NIH
Keystrokes: war TAB 5 TAB
182 different drugs require only 4 keystrokes: letter TAB number TAB
Functional flexibility matches ideal

29. RxTerm Medication Entry Demo App
from The National Library of Medicine, NIH
Procedural flexibility: 5.1bits
Maximum interface efficiency: ~ 68% (4 keystrokes)
Minimum interface efficiency: ~ 5% (50 character free text entry)

30. Reducing Wrong Patient Errors: BONU
Animated Transitions & Photos S

31. Reducing Wrong Patient Errors:
Animated Transitions & Photos

32. Reducing Wrong Patient Errors:
Animated Transitions & Photos

33. Reducing Wrong Patient Errors:
Animated Transitions & Photos

34. Reducing Wrong Patient Errors:
Animated Transitions & Photos
Error Recognition Rate for each Group
0.7
63%
0.6
0.5
43%
36% 0.4
0.3
0.2
7% 0.1
0
Control
Control Animation
Animation Photo
Photo Combined
Combined
The combination of animation & photo resulted in a significant increase
in error recognition rate relative to the control & animation groups 
Dramatic implications for commercial systems
(Taieb-Maimon, Plaisant & Shneiderman, 2012)

35. Year 3 & 4 Planned
Deliverables
Medication Reconciliation
Lab Tracking
Systematic Yet Flexible Systems
Analysis
Visualization Guidelines

36. Anticipated Challenges in Years 3-
4
Clinical User Testing
Industry Collaboration in Technology
Transfer
Resources for Widening Impact

37. Join us for HCIL symposium
May 22-23, 2012
Includes Medical Informatics Workshop on Day 2
www.cs.umd.edu/hcil/soh