Abstract
We describe our preliminary experience in using a computer rule-based system to assist real-time public health surveillance. The rules operated on HL7 messages transmitted by hospital information systems and included laboratory test ordering, results, radiology reports, emergency room & outpatient clinic visits and hospital admissions. Data elements were mapped to standard vocabularies and then run through the inference engine, in a data-driven process. The aim of the system was to detect events of possible public health significance, as well as reportable infectious diseases. The program has been actively used by the University-based infection control team, beginning with the 2002 Olympic Winter Games. It represents a novel surveillance system that uses objective, quantifiable events to support monitoring of community populations.