The 130th Annual Meeting of APHA

Abstract #46515

Man-huei Chang¹, M. Kathleen Glynn², Lori C. Hutwagner³, Carol Peterson⁴, Donald E. Ward⁵, Jordon Lewis⁶, Diana Jordan⁷, Steven Wiersma⁵, Sherrie Bruce⁸, Samuel L. Groseclose¹, and Tracee Treadwell⁸. (1) EPO/DPHSI/SSB, Centers for Disease Control and Prevention, 4770 Buford Highway NE, MS K-74, Atlanta, GA 30341, 770-488-8412, mchang@cdc.gov, (2) NCHSTP/DH, Centers for Disease Contorl and Prevention, 1600 Clifton Road, MS-E47, Atlanta, GA 30333, (3) DHS/ESB, Agency for Toxic Substance and Disease Registry, 1600 Clifton Road NE, MS- E31, Atlanta, GA 30333, (4) Los Angeles County HD/Acute Communicable Disease Unit, California Department of Health, 313 N. Figueroa Street, Los Angles, CA 90012, (5) Bureau of Epidemiology, Florida Department of Health, 4052 Bald Cypress Way, Bin A-12, Tallahassess, FL 32399, (6) Hillsborough County Health Department, Florida Department of Health, 1105 E. Kennedy Boulevard, P.O. Box 5135, Tampa, FL 33675, (7) Pinellas County Health Department, Florida Department of Health, 500 7th Avenue South, St. Petersburg, FL 33701, (8) NCID/BPRP, Centers for Disease Control and Prevention, 1600 Clifton Road, NE, MS C-18, Atlanta, GA 30333

Background: Short-term, alternative data sources with limited trend information may be necessary to supplement traditional surveillance in detecting potential bioterrorism events. Statistical quality control methods may be used to identify unusual patterns in short-term surveillance data.

Methods: Syndrome-based surveillance was conducted for 3 weeks in selected emergency departments (EDs) during two high-profile events: the 2000 Los Angeles Democratic National Convention (DNC) and the 2001 Tampa Super Bowl and Gasparilla Festival activities (SBGF). Syndrome data were analyzed once daily during DNC and twice daily during SBGF using four aberration detection methods (Shewhart P and C, moving average, and cumulative sum control chart) and the chi-square test. We calculated the aberration detection lag time (interval between ED visit and application of aberration detection) for both events, and compared aberration detection by method and syndrome.

Results: A median of 57 visits/ED/day (range 2-237) were reported during DNC and 118 (56-242) were reported during SBGF. Overall, the median aberration detection lag time was less than 24 hours and was similar between the two events (DNC 23.8 hours; SBGF 20.3 hours). P and C chart methods performed similarly and detected the most aberrations among all methods. Aberrations were more likely to occur at low number of ED visits for critical rare syndromes, with the inverse for common syndromes.

Conclusion: Quality control methods might be useful for the timely identification of aberrant patterns in surveillance data that could be used to direct public health interventions. Further evaluation of the application of these aberration detection methods will benefit public health surveillance allowing early detection of potential bioterrorism.

Learning Objectives:

• Articulate the benefit of appling quality control methods on hospital emergency department syndrome-based data for early detection of potential bioterrorism events.
• Articulate the usefulness of aberration detection methods on data with no historical baseline
• Assess the needs for implementation of aberration detection methods in a short-term non-traditional surveillance system.

Keywords: Bioterrorism,

Presenting author's disclosure statement:
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.