intelligent decision aids (IDAs) must be able to promote both high-level decision making under uncertainty and the ability to develop strategies for planning for and preventing stressful events (Kontogiannis and Kossiavelou, 1999). The most successful IDAs for team training purposes are those that mimic usual event escalation processes and contributors, including imagined action consequences, anticipation of rare events, and prioritization of tasks when time is limited. IDAs also can be used to provide information about an event or situation, to present multiple perspectives about potential contributors and possible outcomes, and to monitor task performance. In addition, they have potential relevance for facilitating contingency planning through the use of information displays concerning difficulties encountered in the past, critical errors associated with similar actions, and resources needed to activate the plan. Because the use of IDAs for assistance with decision making in highly stressful conditions is new, experiments and field evaluations of their effectiveness must be an integral part of their use (Kontogiannis and Kossiavelou, 1999).
Reports on methods for monitoring team processes are few, with most evaluations of team performance focusing primarily on clinical outcomes rather than error or error avoidance. Although favorable outcomes are commonly interpreted as an indication of the absence of error, this assumption needs to be documented more clearly. Moreover, because the development and maintenance of effective teams are essential to safe care delivery processes and ideal outcomes, efforts need to be made to monitor and describe those collaborative groups and work teams that consistently produce safe care. Identifying teams and organizations as benchmarks for outcomes is insufficient; understanding and mimicking their processes also is required.
Strategies for evaluating team performance range from day-to-day quality assessment processes to formal investigations of team impact. Inherent in all discussions of the impact of interdisciplinary teams on patient safety and other care delivery outcomes, however, is the need for continuous assessment of team performance and impact. This continuous process is highlighted in a model of collaboration described by Sorine and colleagues (1996), who identify five essential components of the collaboration cycle, each requiring close monitoring of process and outcome. In Sorine et al.’s model, performance guidelines drive compliance agreements, which in turn influence preparedness training and implementation procedures. Once the procedures have been implemented, verification and improvement efforts are undertaken to ensure the quality and consistency of behaviors. These