ANCHORING BIAS IN GYNECOLOGIC ONCOLOGY
Candidate for the Lee B. Lusted Student Prize Competition
Method: Gynecologic oncologists practicing in New England were surveyed electronically by completion of a brief online questionnaire that included three clinical scenarios and one non-clinical scenario. Respondents were asked to estimate life expectancy for a patient with platinum-resistant epithelial ovarian cancer, the probability of pulmonary thromboembolism in a postoperative patient, and the height of the world’s tallest building; they were also asked to evaluate the performance of a hypothetical medical student. Respondents were randomized to one of two survey versions; each version included either a “high” or “low” arbitrary anchor in the question without changing the relevant content of the scenario. Mean estimates were calculated and significant difference was evaluated by the Student’s T test.
Result: 55 responses were obtained from 94 gynecologic oncologists (59% response rate). 85% of responses were from staff gynecologic oncologists, 15% from fellows in training. Respondents provided with a high anchor were significantly likely to give a higher estimate of life expectancy (15.5 months vs. 10.8 months p=0.005), probability of pulmonary embolism (38% vs. 20% p=0.02) and the height of a building (3243 feet vs. 1527 feet, p<0.0001). There was no difference in medical student evaluation regardless of anchor, nor in likelihood of ordering diagnostic imaging or therapy for presumed pulmonary embolism.
Conclusion: Gynecologic oncologists may be significantly affected by anchoring bias in counseling and treating patients. More effort must be directed towards awareness of this bias and investigation of possible remedies to ensure that all patients receive the same quality and content of care.
See more of: The 36th Annual Meeting of the Society for Medical Decision Making