Analytic Evaluation of a Standardization Effort for the DISTRIBUTE Emergency Department Surveillance Project

• Activity Number: 336
• Type: Topic Contributed
• Date/Time: Tuesday, August 2, 2011 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistics in Defense and National Security
• Author(s): Howard Burkom*+ and Donald Olson and Marc Paladini and Atar Baer and Debra Revere
• Companies: The Johns Hopkins University and International Society of Disease Surveillance and New York City Department of Health and Mental Hygiene and Public Health-Seattle & King County and University of Washington
• Keywords: disease surveillance ; signal-to-noise ; influenza ; CUSUM

Abstract

The Distribute project supports collection of aggregate emergency department (ED) data to analyze care-seeking behavior for influenza-like illness (ILI). Forty-two public health sites provide data from jurisdictions representing ~60% of the US population, and capturing ~50% of all ED visits nationwide. Each jurisdiction uses a local preferred syndrome definition (?Preferred? ILI). Definitions vary widely, making analytical comparison across sites difficult. A pilot project was undertaken to evaluate use of a standardized definition of ILI across sites (?ILI-S?). Six sites provided 4 years of data using a syndrome definition comprising 3 components-fever/cough, fever/sore throat, and flu-defined by a formal code set. Evaluation measures included a signal-to-noise calculation based on epidemic intervals reported by the participating sites and a CUSUM control chart for evaluation of detection timeliness. Results of comparative visual and statistical analysis indicated a) clearer inter-regional comparisons for ILI-S than for Preferred ILI; b) similar differences among age groups across definitions; and c) no significant loss in signal-to-noise ratio for the ILI-S groupings.