psychological foundation of clinical decision-support
TRANSCRIPT
4/29/14
1
Alan H. Morris, M.D.
Psychological Foundation of Clinical Decision-Support
Utah Health Services Research Conference
U of Utah, 30 April2014
Clinical Uncertainty- Complexity: >236 Variable Categories 1. Perception / Opinion
2. What does not work well 3. Nudging towards goals 4. Feasible solutions
Survivor stories drive screening toward more Over diagnosis#
More Intensive#Screening#
More "Survivor"#Stories#
More#Overdiagnosis#
More Useful#Screening#
Appears To Be#
H. Gilbert Welch MD, MPH#
Raffle AE, Gray JAM. Screening : evidence and practice. Oxford ; New York: Oxford University Press; 2007. xiv, 317 p
Peter or#Andrew?# ECMO, Prayer#
More #ECMO, Prayer#
More Useful#ECMO,Prayer#Appear To Be#
1. Perception / Opinion 2. What does not work well 3. Nudging towards goals 4. Feasible solutions
Central line associated blood stream infection: Hand hygiene
at insertion: 62% (6-71% comply: infection control)
2014;42(2):94-9
1. Perception / Opinion 2. What does not work well 3. Nudging towards goals 4. Feasible solutions
Many make bad decisions we
would not have made if .. full
attention.. had information and
cognitive abilities.. self-
control. p. 5. Yale University Press. Kindle Ed. 2008-04-08.
Libertarian Paternalism Libertarian: people be free to
opt out (liberty-preserving) Paternalistic: choice architects
should influence behavior to make people healthier.
Sunstein, Cass R.; Richard H. Thaler (2008-04-08). Nudge: Improving Decisions About
Health, Wealth, and Happiness (p. 5). Yale University Press. Kindle Edition.
4/29/14
2
Arranging cafeteria • able to ↑ or ↓ consumption of many foods up to 25 %
• Influenced by small changes in context
• Power: focus user attention
Sunstein, Cass R.; Richard H. Thaler (2008-04-08). Nudge: Improving Decisions About
Health, Wealth, and Happiness (p. 1-2). Yale University Press. Kindle Edition.
Doctor is a choice architect • must describe alternative treatments to patients
• organizing decision context • there is no neutral (uninfluenced) design.
Sunstein, Cass R.; Richard H. Thaler (2008-04-08). Nudge: Improving Decisions About Health, Wealth, and Happiness (p. 1-2). Yale University Press. Kindle Edition.
“Opt in” – Germany, 12%
“Opt out” – Austria, 99%
Drivers License Organ Donation
http://www.nytimes.com/2009/09/27/business/economy/27view.html?_r=0&adxnnl=1&pagewanted=print&adxnnlx=1396455073-WPMVY2uodw+NxB1OxMBNrw
Dual Process Theory (cognition)
Slow Fast Conscious
~10-40 bits/s Burden: conserve
X
Unconscious ~11x106 bits/s
No Burden X
Aha! 2
1
System 2 Thinking
(conscious)
Dual Process Theory
Beginning student
SLOW- little data 10-60 bits/sec
15
Skilled performer
Eyes closed No music “Autopilot” System 1
Thinking (unconscious)
Dual Process Theory
System 2 Thinking
conscious
System 1 Thinking
unconscious
Dual Process Theory - unburden conscious
Traditional: Teach more
Only small improvement
(6-71% comply: infec control)
1. Perception / Opinion 2. What does not work well 3. Nudging towards goals 4. Feasible solutions
Simple Guideline
Bedside Paper Protocol
Bedside Computer Protocol (eProtocol)
4/29/14
3
% M
easu
rem
ents
Blood Glucose (mg/dl)
8
6
4
2
0 0 40 80 120 160 200 240 280 320
Simple Guideline
→ Target Range
% G
luco
se M
easu
rem
ents
Bedside Paper Protocol Bedside Computer Protocol
Expect 6-9 additional ICU Deaths/100 patients
% G
luco
se M
easu
rem
ents
Blood Glucose (mg/dl)
8
6
4
2
0 0 40 80 120 160 200 240 280 320
Western USA Southeast USA Northeast USA Singapore
→ Target Range
Blood Glucose eProtocol-insulin
% M
easu
rem
ents
Blood Glucose (mg/dl)
8
6
4
2
0 0 40 80 120 160 200 240 280 320
ICU Type Pa-tients
Measure-ments
Research 493 21,321 Clinical Care 2,296 109,458
→ Target Range
% G
luco
se M
easu
rem
ents
eProtocol-insulin: Research vs. Practice
1. Perception / Opinion 2. What does not work well 3. Nudging towards goals 4. Feasible solutions