1248. Near Real-Time Norovirus Disease Surveillance Using BioSense Emergency Department Chief Complaint-Based Syndromic Data
Session: Oral Abstract Session: Hot Topics in Public Health
Saturday, October 5, 2013: 11:30 AM
Room: The Moscone Center: 220-226
Background: Noroviruses are the leading cause of sporadic and epidemic gastroenteritis in the United States (US).  Timely monitoring of norovirus activity could rapidly identify season onset and elevated levels of activity, thereby focusing prevention and control efforts.  We assessed whether BioSense, a national-level, near real-time, electronic surveillance system that maps emergency department (ED) visit data (ICD-9-CM codes and chief complaint text) to 78 sub-syndromes, could monitor norovirus activity.

Methods: BioSense ED visits mapped by chief complaint text to “Diarrhea” or “Nausea and Vomiting” (N/V) sub-syndromes as a monthly proportion of all ED visits were compared with monthly reported norovirus outbreaks (January 2007–April 2010). Linear regression models were fitted separately for five age groups (0–4, 5–17, 18–64, ≥65 years, and all ages). For the 0–4 year group, laboratory-reported test data for rotavirus, a major cause of gastroenteritis in this age group, were also included in the model. We restricted analysis to six states with uninterrupted 1) BioSense, 2) norovirus outbreak, and 3) rotavirus antigen test data (≥120 tests/year). In addition, weekly national-level BioSense and outbreak data (2009–2011) were used to estimate norovirus season onset, peak, and end.

Results: During the 40-month period, the six states reported 277,433 Diarrhea and 1,165,414 N/V visits, 1048 norovirus outbreaks, and 32,455 rotavirus antigen tests (13% positive). The proportion of visits mapped to Diarrhea or N/V both had strong linear relationships with norovirus outbreaks in each age group (P <0.001), with stronger correlations for Diarrhea (R = 0.828–0.926) than N/V (R = 0.729–0.866) in each age group. Including rotavirus activity significantly improved the 0–4 year age group models (P <0.01). Diarrhea ED visit proportions exhibited winter seasonality attributable to norovirus, with rotavirus also contributing in those aged <5 years, and were used to estimate onset, peak, and end of season within four weeks of observed dates.

Conclusion: BioSense ED visit data mapped by chief complaint to Diarrhea correlated strongly with reported norovirus outbreaks, and could provide a near real-time indicator of norovirus disease activity in the US.

Brian Rha, MD, MSPH1, Sherry Burrer, DVM, MPH-VPH2, Soyoun Park, PhD, MS1, Tarak Trivedi, BS1, Umesh D. Parashar, MBBS, MPH1 and Benjamin A. Lopman, PhD, MSc1, (1)Centers for Disease Control and Prevention, Atlanta, GA, (2)Centers for Disease Control and Prevention, Chamblee, GA


B. Rha, None

S. Burrer, None

S. Park, None

T. Trivedi, None

U. D. Parashar, None

B. A. Lopman, None

Findings in the abstracts are embargoed until 12:01 a.m. PST, Oct. 2nd with the exception of research findings presented at the IDWeek press conferences.