Can Induced Anxiety from a Negative Earlier Experience Influence Vascular Surgeons’ Statistical Decision-Making? A Randomized Field Experiment with an Abdominal Aortic Aneurysm Analog

BackgroundIncreasing detection, new screening recommendations, and popular press attention contribute to the rising prevalence of asymptomatic abdominal aortic aneurysms (AAA). Evidence-based guidelines recommend the optimal time to operate is when the aneurysm is 5.5 cm in diameter. Smaller AAAs are periodically monitored with imaging. Recent events and emotional reactions to risk and uncertainty, including anxiety, can cause decision-making to diverge from cognitively based assessments. It is not known whether this applies to vascular surgeons making statistically-optimal, risky decisions. We tested whether an unexpected, recent negative event might influence vascular surgeons’ decisions about a computer-simulation AAA-analog that includes statistical information.Study designWe performed a randomized, computer-based field experiment with evidenced-based statistical information readily available on bursting probabilities. Participants included vascular surgeons with AAA operative experience attending two vascular surgery conferences held in 2005 (n = 81). The intervention was a randomly assigned, anxiety-inducing, bursting balloon versus a nonbursting balloon before a statistical decision-making computer simulation. The main outcomes measure was real-time prospective choice to opt out of expanding AAA simulation. A Cox proportional hazard model was used to assess the likelihood of opting out, while controlling for important covariates.ResultsThe experimental group was more likely to opt out (hazard ratio: 3.32; 95% CI: 1.25 to 8.81), even after controlling for initial anxiety levels, risk attitudes, uncertainty attitudes, use of statistical information, surgical experience, and demographics.ConclusionsExperiencing a negative, potentially anxiety-provoking, preceding event can influence decision-making, even among experienced vascular surgeons who have ready access to statistical risk information.