The promise offered by fully computer-based patient records for improving the quality of patient care and advancing medical knowledge through research is enormous. Therefore, concerted efforts should be made to overcome legal and technological barriers that stand in the way of full development of computer-based records and record systems.
In the future, with increasing use and development of artificial intelligence systems, computer-based patient records may be expected to become interactive, providing diagnostic assistance and even treatment recommendations. An interactive patient record promises improved quality of care, but the interaction of such ''smart" systems with computer-based patient records will also raise a host of legal and policy issues that are beyond the scope of this paper. Among them will be allocation of responsibility (and liability) for errors in the artificial intelligence system, whether caused by faulty hardware, faulty software, or error in the system's medical rules. The more advanced such systems become, the more questions they will generate about the practice of medicine and whether nonphysicians can use these systems to diagnose and treat patients without physician involvement. In addition, systems that can diagnose or treat patients without intervening professional involvement may be classified and regulated as medical devices under food and drug laws. Finally, these "smart" systems can be expected to lead to a redefinition of the physician's role, as they begin to perform functions that formerly only a physician could perform.