Real-Time Analytics for Legacy Data Streams in Health: Monitoring Health Data Quality

Healthcare organizations are increasingly using information technology to ensure patient safety, increase effectiveness and improve efficiency of healthcare delivery. While the use of health information technology (HIT) has realized many improvements, it has also introduced new failure modes arising from data quality and IT system usability issues. This paper presents an approach towards addressing these failure modes by applying real-time analytics to existing streams of clinical messages exchanged by HIT systems. We use complex event processing provided by the Event Swarm software framework to monitor data quality in such systems through intercepting messages and applying rules reflecting the syndromic surveillance model proposed in [4]. We believe this is the first work reporting on the real-time application of syndromic surveillance rules to legacy clinical data streams. Our design and implementation demonstrates the feasibility of this approach and highlights benefits obtained through improved operational quality of HIT systems, notably better patient safety, reduced risks in healthcare delivery and potentially reduced costs.