Immediately following September 11, 2001, the District of Columbia Department of Health began a syndromic surveillance program based on emergency room (ER) visits. ER logs are faxed on a daily basis to the health department, where health department staff code them on the basis of chief complaint, recording the number of patients in each of the following syndromic categories: death, sepsis, rash, respiratory complaints, gastrointestinal complaints, unspecified infection, neurological, or other complaints. These data are analyzed daily and when a syndromic category shows an unusually high occurrence, a patient chart review is initiated to determine if the irregularity is a real threat.
A time series analysis of the data from this system has shown that with the application of a variety of detection algorithms, the syndromic surveillance data does well in identifying the onset of the flu season. In addition, simulation studies using the same data have shown that over a range of simulated outbreak types, the univariate and multivariate CUSUM algorithms performed more effectively than other algorithms. The multivariate CUSUM was preferred to the univariate CUSUM for some but not all outbreak types.
Objective
This paper evaluates an ER syndromic surveillance system based on simulation studies and comparisons with other surveillance systems.