EVALUATION OF PHYSICIANS' COGNITIVE STYLES

Monday, October 21, 2013
Key Ballroom Foyer (Hilton Baltimore)
Poster Board # P2-24
Decision Psychology and Shared Decision Making (DEC)

Benjamin Djulbegovic, MD, PhD1, Jason W. Beckstead, PhD2, Shira Elqayam, PhD3, Tea Reljic, BS4, Ambuj Kumar, MD, MPH5, Janis A. Cannon-Bowers, PhD4, Stephanie Taylor, MD4, Athanasios Tsalatsanis, PhD6, Brandon M. Turner, PhD7 and Charles N. Paidas, MD, MBA4, (1)USF, Tampa, FL, (2)University of South Florida College of Nursing, Tampa, FL, (3)De Montfort University, Leicester, United Kingdom, (4)University of South Florida, Tampa, FL, (5)Center for Evidence Based Medicine, University of South Florids, Tampa, FL, (6)USF Health, Tampa, FL, (7)Stanford University, Stanford, CA
Purpose: Patients’ outcomes critically depend on the accuracy of physicians’ judgments. Yet, little is known about individual differences in physicians’ decision-making styles. The purpose of this work is to assess physicians’ individual differences in cognitive styles related to medical decision-making. We also tested the relation with age, experience and degree and type of training.

Method: We collected data on a) demographics, b) responses to 6 well-validated scales [maximizing (as alternative search and decision difficulty) and satisficing, analytical vs. intuitive reasoning, need for cognition, intolerance toward ambiguity, tendency to seek evidence to inform decision-making, and disposition to suppress intuitive responding (CRT)]; c) responses to four types of conditional inference (MP- modus ponens; MT-modus tollens; DA-denial of the antecedent; AC-affirmation of consequence).

Result: 221 physicians consisting of trainees [61% residents, 14% fellows] and 25% faculty (attendings) completed the full survey. 54% were female; mean age: 33.73 [range: 25 to 69]. All 6 constructs show acceptable psychometric properties. There was a negative correlation between the number of incorrect AC inferences and satisficing (r=-0.25;p=0.02) and CRT (r=-0.3;p=0.0003). Satisficing had a positive correlation with analytical thinking (r=0.23;p=0.03), which correlated negatively with the number of incorrect DA inferences (r=-0.232;p=0.04). Surprisingly, we found a negative correlation between age and satisficing (r=-0.240;p=0.013). As expected, one component of maximizing (alternative search) decreased with age (r=-0.220;p=0.037); however, scores on decision difficulty were not related to age.

Conclusion: We report the first study of cognitive styles in physicians. Those scoring higher on rational thinking made fewer inferential mistakes. We found a negative correlation between age and maximizing: as they advanced in their career, physicians are less willing to spend time and effort in exhaustive search for solutions. The finding has immediate implications for educational prescriptions (such as developing evidence-based resources that are immediately accessible).  Contrary to the expectations we found that satisficing decreases with age. This raises the question for re-interpretation of satisficing concept as traditionally understood.