This research will test the hypothesis that decision support in critical care is enhanced by incorporating the concept of therapeutic goals into data collection and interpretation strategies. Since a w variety of information is available and a large amount of processing can be done in critical care environments, some strategy is needed to prioritize data collection and reasoning tasks, as equipment limitations may preclude real-time interpretation of all information. """"""""Intelligent monitoring"""""""" is used to describe the task c adapting the data collection and reasoning strategies in response to changes in the observed system. The above hypothesis will be tested by implementing a goal-based intelligent monitoring system in the intensive care unit and evaluating it's performance in clinical use. Several specific aims have been defined to support this objective. First, necessary data will be integrated on a bedside computing platform in the medical intensive care unit in real time. Next, knowledge base requirements for storing a test set of therapeutic goals will be defined, and a software representation of the test goals will be developed to facilitate implementation of a goal-based reasoning module. This module interfaces with an existing bedside intelligent monitoring system and synthesizes data from the monitors, hospital information system, and clinical staff to infer therapeutic goals and support the intelligent monitoring tasks. Evaluation of the system in clinical use is an objective. Finally, maintaining patient safety and confidentiality is an overall specific aim.
These aims allow successful evaluation of the ability of therapeutic goals to facilitate intelligent monitoring in critical care decision support. The insight which results from this testing has the potential to improve the design and development of systems to enhance clinical decision-making, thereby improving quality of care.