ABSTRACT
Medical emergency teams have to act in a coordinated manner. One might assume that the behaviors necessary for good team coordination are the same across many different situations. In contrast, we hypothesized that coordination behaviors have to match task requirements, which may be different across tasks. In a medical simulator we studied performance of medical professionals in two different scenarios. In the cardiac arrest scenario requiring resuscitation, the diagnosis is clear, and task requirements are prescribed by guidelines. This situation calls for decisive action with minimal delays; it should profit from directive leadership. In the diagnostic task scenario the diagnosis was difficult, as conflicting cues were
present. This calls for joint reflection and weighing all available evidence; it should therefore profit from explicit reasoning and from involving the whole group by "talking to the room". Micro-analysis of coordination behaviors revealed that clear leadership was related to performance in the resuscitation, but not the diagnostic task. ln contrast, "talking to the room" was related to performance in the diagnostic, but not the resuscitation scenario.
