Our long-term objective is the development of expert systems using fuzzy logic for sophisticated, rapid analysis of physiologic data in intensive care units. Our immediate objectives are to tune and extend our Phase I pulmonary artery catheter data system to alarm nursing staff on invalid data, and to conduct clinical trials to test whether the system will improve patient care. The existing single-bed system will be adapted to multi-bed and Hewlett-Packard, Marquette and Space Labs physiolgoic monitors. Eighty patients at ICUs in two hospitals will participate in final system development, and an additional 200 patients will participate in a clinical trial of system effectiveness in reducing the number and duration of pulmonary artery catheter abnormalities. (Our tests indicate that the incidence of catheter malfunction may be badly underestimated.) We will also study the expansion of the expert system to include additional information including SvO2 and cardiac output, and to deliver more informative messages to nurses and physicians. Finally, the system will be made suitable for commercial use.