A tool for training in decision making for emergency general surgery--explicit training is possible through facilitation.

Background: Emergency general surgery is increasingly recognized as an area for improvement within the National Health Service in the United Kingdom. Our aim was to test the feasibility of a tool we developed in assessing and facilitating surgical decision making by trainees in the emergency setting.

Methods: A decision-making tool based on common decisions in surgery was designed and incorporated prospectively into patient clerking. The 10 decisions available to surgeons were divided into 3 major headings: discharge decisions, observation decisions, and operative decisions. "Ideal decisions" were derived from the final diagnosis on discharge by 2 independent clinicians and compared with initial decisions made by surgeons. We analyzed the data with a view to validate the tool using 2 methods. Firstly, we looked at the percentage of correct decisions made by different grades of surgeon, and secondly, we analyzed the trends in deviations from ideal decisions.

Results: The decision-making tool was completed for 136 emergent surgical admissions over 2 months. Interns made less "ideal" decisions compared with residents (45% vs 56%, respectively; p = 0.10, Fisher exact test) and attending surgeons (45% vs 70%, respectively; p = 0.0001). Interns made more "admit and observe or investigate" decisions compared with residents (63% vs 55%, respectively; p = 0.27) and attending surgeons (63% vs 51%, respectively; p = 0.01), who were more likely to decide to operate on or discharge patients. A repeat study over the subsequent 2-month period has shown comparable findings in proportion of ideal decisions for each grade of surgeon.

Conclusions: There are significant differences in the proportion of "ideal decisions" and types of decisions between interns and attending surgeons suggesting a learning curve and potential for using this tool in decision-making training. The tool inherently facilitates discussion over decisions made by making these explicit. Repeating the study revealed an internal consistency.