John W. Robinson, M.D., Ph.D., LLC
Statistics ▪ Analytics ▪ Data Science
Real-Time Clinical Events Monitoring with Health Care Claims
Healthcare organizations often need to monitor a population for the occurrence of specific clinical events, such as the assignment of particular diagnoses or the provision of particular services or procedures. Health care claims offer a readily available source for identifying such events; however, information contained in claims is often not available for case management and quality improvement purposes until after claims have been processed for payment and selected data elements have been extracted and loaded into a data warehouse. By that time, the clinical information may no longer be current or may have been altered during the claims adjudication process.
When claims are received via Electronic Data Interchange (EDI) transactions, automated "real-time" mining of claims for important clinical events is an expeditious alternative to waiting for completion of claims processing and data warehousing procedures. EDI claims, which have an inherently hierarchical structure, can be readily converted to Extensible Markup Language (XML) format, which can in turn be directly queried for diagnosis codes and procedure codes that represent the specific clinical events targeted for monitoring. When a targeted code is identified in a claim, patient and provider identifiers and other selected information from that claim can be loaded into a clinical events database for aggregation, analysis, and report generation.