The Public Health Agency of Canada is currently utilizing a syndromic surveillance prototype called the Canadian Early Warning System (CEWS). This system monitors several live data feeds, including emergency room chief complaint records from all seven local hospitals, Telehealth (24/7 nurse hotline) calls, and over-the-counter drug sales from a number of the large chain drug stores. Data trends are analysed for aberrations as early indicators of outbreak events. Collaborators on this Winnipeg, Manitoba-based pilot include the Winnipeg Regional Health Authority and IBM Business Solutions. Algorithms currently in CEWS include the 3, 5 and 7-day moving averages, CUSUM and the CDC’s EARS. We seek to investigate the performance of these algorithms in view of the fact that their detection ability may be dependent upon data source and/or the type of outbreak event.
Objective
To determine the sensitivity, specificity and days to detection of several commonly used algorithms in syndromic surveillance systems.