Clinical decision support systems and medical information systems are both of great interest in medical informatics. Ideally these two types of systems should be closely linked to provide maximum educational and clinical utility. Usually medical informatics development emphasizes one type of system to the exclusion of the other. The primary objective of this proposal is to develop a generic, linked system in which medical knowledge represented in the decision support module is applied in the clinical setting in the context of an existing information system. The major advantages are practical: the data required by the decision support system is already available; no special data collection or entry is needed. The advice given by the decision support system is presented in clinical """"""""real-time"""""""" to optimize its effect. The second objective is to maximally exploit non-artificial intelligence in medicine (AIM) systems for representing medical knowledge and decision making processes when appropriate. This is cost effective and, if sufficiently robust, may be adequate for many types of medical decision problems. In this proposal we plan to develop general methods and apply them to a single clinical setting: hypertension care. The specific work will test the hypothesis that an automated therapeutic management review system may be created, using a non-AIM rule-based structure coupled with a system designed to frequently gather pertinent clinical information, to provide sophisticated analysis of current medical management for providers. The system will survey current clinical data and, following a stepped care approach to the treatment of hypertension, advise providers as to when they should increase therapy, modify therapy, or reduce therapy in appropriate circumstances. The prototype system will be studied in a general medicine practice. Testing will involve the examination of its performance in terms of reliability and validity, its psychological impact on providers and their perception of benefit and satisfaction, and its cost and impact on stepped care therapy. If this development and feasibility project is successful, we will consider submitting a proposal to subject the developed system to more definitive study.
