Automated Syndromic Classifi cation of Chief Complaint Records

yndromic surveillance, a medical surveillance approach that bins data into broadly defi ned syndrome groups, has drawn increasing interest in recent years for the early detection of disease outbreaks for both public health and bioterrorism defense. Emergency department chief complaint records are an attractive data source for syndromic surveillance owing to their timeliness and ready availability in electronic form. As part of the ESSENCE prototype biosurveillance system, APL has developed an automated process for classifying chief complaint data into syndrome groups, thus increasing the feasibility of chief complaint–based surveillance by avoiding laborious manual classifi cation. Preliminary studies indicate positive results in classifying daily chief complaint data feeds from multiple hospital sources.