MEDICAL SPECIALTY DIFFERENCES IN TREATMENT TIMING DECISIONS FOR EXPANDING ABDOMINAL AORTIC ANEURYSM REPAIR: A RANDOMIZED, COMPUTER SIMULATION STUDY

Purpose:  Even when well-established, evidence-based guidelines exist, physicians often fail to follow them.  Caring for patients with expanding asymptomatic abdominal aortic aneurysms (AAA) requires a series of choices between continuing watchful waiting (WW) versus doing surgical repair.  Published guidelines support delaying repair until AAA rupture risk exceeds the 5% surgical mortality risk when a AAA reaches 5.5cm in diameter. We have previously shown that vascular surgeons' timing decisions are strongly influenced by a preceding (bad) outcome.  We experimentally tested the influence of preceding events on physician decisions in three specialties: surgeons, geriatricians, and anesthesiologists.    

Methods:  A field experiment was conducted at professional society meetings for surgeons (n=63), anesthesiologists (n=92), and geriatricians (n=67). Participants completed a questionnaire and an incentive-compatible, computer-based simulation of management of a patient with a AAA.  The simulation presented serial CT-scan images of an expanding AAA, including its size and rupture likelihood.  At each update, participants chose between continuing WW (and accepting the rupture risk), or doing surgery (and accepting the 5% operative risk).  After completing a conditioning round randomly presenting either a AAA rupture, a surgical fatality, or a successful surgery, the participants completed the simulation.  The dependent variable was the number of times participants continued WW.  Time-to-event analyses, for the event of choosing surgery, were conducted.

Results:  All groups chose repair significantly earlier than the statistically-recommended 10 visits (surgeons - 7.0 visits; anesthesiologists - 7.3 visits; geriatricians - 7.7 visits).  Surgeons in the WW group waited only 5.5 ± 0.6 visits before operating, corresponding to an AAA size of 4.7cm and a rupture risk of 1.2%. Geriatricians in the successful surgery group waited longest, 8.0 ± 0.7 visits. After controlling for demographics, risk attitudes and baseline anxiety, both surgeons (OR = 1.99) and geriatricians (OR = 1.59) exposed to the WW condition operated significantly sooner. Anesthesiologists were not influenced by prior experience.

Conclusions: We found specialty-dependent and specialty-independent effects on physician choices.  All groups operated earlier than guidelines suggested. Prior experience most impacted surgeons and least impacted anesthesiologists.  These results support the idea that non-statistical factors including specialty choice affect how physicians use available statistical information.  Possible explanations include specialty-specific effects of anxiety, regret, mindset, and responsibility. Further work is required to understand the mechanisms underlying these effects.