Results from the National Capitol Region Emergency
Department (ED) Syndromic Surveillance System
David Blythe1, Leslie
Edwards1, Dipti Shah1, Julie Casani1,
Karen Mattews2, Eileen Steinberger3, Mark
Wegner1, John Davies-Cole4, LaVerne Jones4,
Elizabeth Souza4, Jennifer Capparella4, Leslie
Branch5, Denise Sockwell5, Lori Hutwagner6.
1DHMH, 2Johns
Hopkins Bloomberg School of Public Health, Baltimore, Md, 3University
of Maryland School of Medicine, Baltimore, Md, 4District
of Columbia Department of Health, Washington, D.C., 5Virginia
Department of Health, Richmond, Virginia, 6CDC, Atlanta,
Ga.
Introduction: ED Syndromic surveillance systems
have been proposed as one method for quickly detecting unannounced
biological attacks. We report results from a regional system that
has been operating in the National Capitol Region since September
11, 2001.
Methods: ED logs from selected Maryland, District
of Columbia, and Virginia hospitals are collected daily. Each ED
visit is assigned to one of eight syndrome categories (death, sepsis,
rash illness, respiratory illness, gastrointestinal illness, unspecified
infection-like illness, neurological illness, and all other visits).
Using techniques modified from the cumulative summation (CUSUM)
aberrancy detection method, the daily proportion of each syndrome
category and the daily census are evaluated to determine whether
an expected threshold has been exceeded.
Results: Overall, thresholds were exceeded for
76 syndrome categories on 57 (42%) of 135 days under surveillance:
5 of the 12 days for which thresholds could be calculated in September;
13 days in October; 14 in November; 16 in December; and 9 in January.
Of the 76 total category thresholds exceeded, 15 (20%) were for
the unspecified infection-like illness category, 11 (14%) rash illnesses,
9 (12%) neurological illnesses, 6 (8%) each for the death and sepsis
categories, and 5 (7%) for gastrointestinal illnesses. Eight (11%)
were for the “other” category; 10 (13%) were for changes
in ED census. The number of thresholds exceeded per day varied from
none to three.
Conclusions: Regional ED syndromic surveillance
systems can provide consistent daily information about ED visits.
Depending upon the aberrancy detection methods employed, thresholds
may be exceeded frequently.
Submitted to American Public Health Association
(APHA) Annual Meeting 2002; accepted for oral presentation
Maryland Department of
Health & Mental Hygiene — Epidemiology & Disease Control
Program |
are PDF. Download
Adobe Acrobat Reader for viewing .pdf files
