Integration of knowledge-based system and database for identification of disturbances in fluid and electrolyte balance.

We describe a knowledge-based system which automatically identifies fluid and electrolyte disorders in intensive care patients. The knowledge-based system was built and interfaced to an existing patient data management system (PDMS) in Kuopio University Central Hospital to evaluate the potential of knowledge-based techniques in information management and decision support in the high dependency environment. Because of the integration, the system does not require any manual data input, and it provides a natural extension and increased performance to a current patient data management system used in clinical practise. The paper discusses design considerations and gives the system description. The evaluation of the experimental system in clinical use showed that it performed almost as well as junior clinicians of the intensive care unit.