Recorded presentation

Carla V. Rodriguez, MPH¹, Howard Greller, MD², Rick Heffernan, MPH¹, Jane A Greenko, EMT-P, MPH¹, Debjani Das, MPH¹, Don Weiss, MD, MPH¹, and Robert S Hoffman, MD². (1) Bureau of Communicable Disease, New York City Department of Health and Mental Hygiene, 125 Worth St., Box 22a, New York, NY 10013, 212-788-4341, crodrigu@health.nyc.gov, (2) NYC Poison Control Center, 455 First Avenue, Room 123, New York, NY 10016

Background: Carbon monoxide (CO) is a leading cause of unintentional poisoning deaths. On 12/10/03, over 40 workers were evacuated from a factory due to elevated CO levels. Nineteen patients were transported to three hospitals. Only one hospital reported the event to the regional Poison Control Center (PCC).

The New York City (NYC) Department of Health and Mental Hygiene (DOHMH) conducts syndromic surveillance for disease outbreaks by monitoring patient chief complaint data from 44 emergency departments (EDs) daily. Because we did not routinely monitor CO poisoning syndromes, the system did not detect this cluster. To test the system’s usefulness for detecting CO events, we retrospectively analyzed ED data for evidence of the incident.

Methods: ED visits were categorized into “CO syndrome” if chief complaint included the words “smoke”, “gas”, “fumes”, “C.O”, or indicated headache, but not respiratory, fever, flu-like illness, diarrhea, or vomiting. Using a modified spatial scan statistic, we simulated a daily, prospective analysis for CO clusters using ED data from 01/05/03 – 1/13/04.

Results: Examination of age, gender, and admission time for 156 citywide CO visits on 12/10 identified 14 patients clustered at two hospitals associated with the event. Prospective analysis detected 41 clusters (p<.05) during the one-year period, including the 12/10 event (p=0.015, RR=2.13), which ranked 15th in signal strength based on p-value.

Conclusion: Syndromic surveillance for CO poisoning illustrates a new application for such systems. Based on these results, we will pilot the use of daily, prospective spatial analysis for early detection of CO poisoning clusters.

Learning Objectives: This session will provide participants with the ability to

• Describe spatial scan analysis and syndromic surveillance activities as they are applied in one jurisdiction. Discuss issues regarding their application to environmentally related health outcomes.
• Discuss and assess the feasibility and applicability of these methods to health outcomes of concern in their own jurisdictions.
• Discuss, debate and identify candidate syndromes and data gathering methods for syndromic surveillance. Develop a plan for conducting syndromic surveillance activities.

Keywords: Outbreaks, Surveillance

Related Web page: www.satscan.org; www.syndromic.org

Presenting author's disclosure statement:
Organization/institution whose products or services will be discussed: 1) Kulldorff M. Prospective time periodic geographical disease surveillance using a scan statistic. J R Statist Soc A. 2001;164(1):61-72. 2) Mostashari F, Kulldorff M, Hartman JJ, Miller JR, Kulasekera V. Dead bird clustering:a potential early warning s
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.

Recorded presentation