Presentation Abstract

Session: 631-645-Trauma VI
Date/Time: Friday, Mar 22, 2013, 12:18 PM -12:24 PM
Location McCormick Place, Room N427
Presentation Number: Paper 645
Title: Improving Decision-Making in Fracture Care: Cognitive Bias and Rational Choice
Classification: +Other (Trauma)
Keywords: Tibial Plateau; Epidemiology/Procedural/Outcomes
Author(s): Joseph Bernstein, MD, Haverford, Pennsylvania
Jaimo Ahn, MD, PhD, Philadelphia, Pennsylvania
Abstract: INTRODUCTION: Cognitive biases, as first described by Kahneman and Tversky, can result in irrational decision making. This work applied to behavioral economics won the Nobel Prize in 2002. The purpose of this study is to determine if these biases are found when the treatment options for fracture related care are considered.
METHODS: A survey was presented to 131 subjects. Five clinical scenarios describing a tibial plateau fracture were provided, and respondents were asked to select among treatments offered. Four weeks later, the scenarios were presented again, with slight modifications: (1) To test for an anchoring bias, the respondents were asked to state the maximal rate of surgical complications they would accept; but in the second iteration, an artificially high value was suggested to them. (2) To test for a peer pressure bias, the respondents were first asked to choose between arthroscopic and open treatment; and when it was seen after the first iteration that arthroscopy was favored, in the second iteration, they were told that most people would choose open surgery. (3) To test for a framing bias, respondents were asked twice to choose between surgery and therapy to correct arthrofibrosis: in the first instance, surgery was presented as a means of gaining motion relative to the post-fracture state, whereas in the second instance, surgery was presented as a means of preventing loss of motion relative to the pre-fracture state. (4) To test for an emotional valance bias, respondents were asked if they would prefer a prophylactic fasciotomy to prevent a compartment syndrome, but in the second instance only were they shown photographs of necrotic muscle and an amputation. (5) And last, to test for a distraction bias, respondents were first asked to select between strong or weak anticoagulants, with offsetting benefits; but in the second iteration, they were presented with the same two choices plus a third and obviously inferior choice-a strong anticoagulant with high risk of complication, which, though not appealing itself, makes the original strong anti-coagulant appear more appealing in comparison.
RESULTS: In all cases, cognitive biases lead to response inconsistencies, as shown in the table.
CONCLUSIONS: The selection among treatment options in fracture management should be able to utilize an expected-value calculation. Namely, for each option, patients should consider the possible outcomes, the likelihood of attaining these outcomes and the values they place on them; and rational choice dictates that the option with the greatest expected value should be selected. Yet our data indicate the potential for irrational choices when patients contemplate fracture management options. If true patient-centered care is to be offered, surgeons need to be aware of the cognitive biases which lead to irrational decision making.
Scenario First iteration Second iteration
Case #1. Maximal acceptable complication rateBaseline:
Mean response = 19%
When told that 78% was a reasonable rate:
Mean response = 32%
Case #2. Open vs minimally invasive surgeryBaseline:
Open surgery = 4%
When told that open surgery was chosen by 85% of the others:
Open surgery = 18%
Case #3. Surgery vs therapy for arthrofibrosisWhen surgery is framed as a mean to gain motion:
Surgery = 40%
When surgery is framed as a means to avoid loss of motion:
Surgery = 60%
Case #4. Prevention of compartment syndromeInformation only:
Fasciotomy = 42%
Information plus pictures of complications as well:
Fasciotomy = 72%
Case #5. DVT preventionStrong anticoagulant = 25%
Weak anticoagulant = 75%
Obviously inferior strong anticoagulant (as distraction) = 0%
Original Strong anticoagulant = 43%
Original Weak anticoagulant = 57%

Disclosures: Presenting Author
7 - Clinical Orthopaedics and Related Research
Author 1
2 - Synthes; 3B - Merck Synthes

Please remember to clear your previously entered search criteria when doing a new search.

6300 North River Road
Rosemont, Illinois
Phone 847.823.7186
Fax 847.823.8125

Need Technical Assistance?
OASIS Helpdesk
Hours: 8:00 AM - 5:00 PM CT

Get the AAOS Annual Meeting Proceedings in eBook format