Introduction: Syndromic surveillance monitors trends in nonspecific health indicator data to detect disease outbreaks in a timely manner; however, only a limited percentage of persons with mild illness might exhibit behaviors that could be detected by syndromic surveillance.
Objectives: The objectives of this study were to 1) examine the demographic characteristics of New Yorkers with recent flu-like or diarrheal illness, 2) describe behaviors associated with having flu-like illness, and 3) estimate the citywide burden for selected illnesses by calculating the syndromic multiplier (i.e., the number of citywide illnesses represented by each visit to an emergency department [ED]).
Methods: A cross-sectional telephone survey of 2,433 adult residents of New York City (NYC) was conducted during March 19-March 31, 2003, and October 27-November 23, 2003. Respondents were asked about flu-like illness, behaviors related to flu-like illness, and diarrheal illness during the 30 days before the interview. Estimated numbers of citywide illnesses were compared with ED visits for flu-like and diarrheal illnesses that were recorded by the NYC syndromic surveillance system for the same periods.
Results: Every ED visit for flu-like illness represented approximately 60 illnesses among city residents; every visit for diarrheal illness represented approximately 251 illnesses. Among adults who reported a recent flu-like illness, 53.2% purchased over-the-counter (OTC) medications; 32.6% reported missing school or work; 29.1% visited a physician; 21.4% called a physician for advice; 8.8% visited an ED; and 3.8% called a nurse or health hotline for advice. Of those who reported multiple behaviors, respondents most commonly reported purchasing OTC medications as their first response to a flu-like illness.
Conclusions: Population-based survey data can be used in conjunction with syndromic surveillance data to better understand the relation between nonspecific health indicators and the burden of certain illnesses in the community, and to assess the representativeness of different syndromic data sources.