Purpose: We wanted to evaluate the feasibility of conducting syndromic surveillance in a primary care office using billing data.
Methods: A 1-year study was conducted in a primary care practice; comparison data were obtained from emergency department records of visits by county residents. Within the practice, a computer program converted billing data into de-identified daily summaries of International Classification of Diseases, Ninth Revision (ICD-9) codes by sex and age-group; and a staff member generated daily summaries and e-mailed them to the analysis team. For both the practice and the emergency departments, infection-related syndromes and practice-specific thresholds were calculated using the category 1 syndrome codes and an analytical method based upon the Early Aberration Reporting System of the Centers for Disease Control and Prevention.
Results: A mean of 253 ICD-9 codes per day was reported. The most frequently recorded syndromes were respiratory illness, gastrointestinal illness, and fever. Syndromes most commonly exceeding the threshold of 2 standard deviations for the practice were lymphadenitis, rash, and fever. Generating a daily summary took 1 to 2 minutes; the program was written by the software vendor for a fee of dollar 1,500. During the 2003-2004 influenza season, trend line patterns of the emergency department visits reflected a pattern consistent with that of the state, whereas the trend line in primary case practice cases was less consistent, reflecting the variation expected in data from a single clinic. Still, spikes of activity that occurred in the practice before the emergency department suggest the practice may have seen patients with influenza earlier.
Conclusions: This preliminary study showed the feasibility of implementing syndromic surveillance in an office setting at a low cost and with minimal staff effort. Although many implementation issues remain, further development of syndromic surveillance systems should include primary care offices.