Syndromic surveillance: Wave of the future for bioterrorism preparedness
Old walls between hospitals, public health are coming down
The threat of bioterrorism and the continuing emergence of new infectious agents have spurred the development of syndromic surveillance systems, which may detect clusters of cases much earlier than traditional methods. What does that buy you? Time.
"Syndromic surveillance is trying to pick up a day, two days, even hours in some instances, of a larger pattern of indicators that suggest the need for a public health investigation," says Daniel Sosin, MD, MPH, director of the division of public health surveillance and informatics at the Centers for Disease Control and Prevention (CDC).
Syndromic surveillance systems typically use health-related data that precede diagnosis but may signal an outbreak that warrants public health response. Some surveillance systems use the International Classification of Diseases (ICD) diagnostic codes with a set of syndromes caused by the major bioterrorism agents.
Other systems abstract data from emergency department logs, 911 calls, or nurse call lines. Syndromic surveillance systems often draw on data sources that already exist, but have not been designed specifically for public health surveillance purposes. "Clinical providers always have and will continue to have a major role in disease reporting and public health surveillance," Sosin says. "Probably, still the most important surveillance system we have is the telephone. The health care provider says, I’m seeing something unusual; are you seeing something in the community?’ That direct connection between a clinician [and the public health department] will continue to be important."
The CDC began trialing such programs in 1999, setting up enhanced syndromic surveillance activities for high-profile community events such as national political conventions. As part of those efforts, the CDC has identified syndrome categories indicative of the clinical presentations of several critical bioterrorism-associated conditions.1 (See table of syndromes.) The Department of Defense’s ESSENCE program (Electronic Surveillance System for the Early Notification of Community-based Epidemics) also uses broad syndrome groups using ICD codes that approximate natural infectious disease outbreaks or bioterrorism.2 These syndrome groups currently are under routine surveillance at military medical treatment facilities. Other public health agencies also have developed syndrome-based definitions and code groupings according to their own data sources and surveillance goals.
The idea is not to create a national system, but guide local health departments in creation of syndromic surveillance programs that are appropriate for their resources and community needs, Sosin says. "Public health truly is local. The identification of public health threats day in and day out has to happen at the local health department. Different local health departments have different risk scenarios; the threat is different in different jurisdictions; and they have different resources to respond. So the thresholds will and should vary. We have a lot to learn yet. It isn’t appropriate that we go out with a single model and tell everybody to do it," he explains.
To assist in the continued development of these systems, the CDC has posted a draft framework document that can be used for evaluating or creating syndromic surveillance programs.3 The model and several other CDC documents on syndromic surveillance have been posted on a web site for comment and review. (See editor’s note at end of article.) Eventually, such systems in one form or another are expected to be adopted throughout the nation, further bringing down the traditional walls between infection control and public health.
It will be important that the surveillance systems are designed to capture the best data from both a clinical and public health viewpoint. Therefore, infection control professionals (ICPs) should consider participating if syndromic surveillance systems are being developed by their local public health departments.
"An important thing about these guidelines and this framework is that if they are employed, then we are not going to send hospital epidemiologists on wild-goose chases," says J. Marc Overhage, MD, PhD, an advisor to CDC on the issue and professor of medicine at Indiana University School of Medicine in Indianapolis. "Ensuring the quality of the process up front is going to be very critical for ICPs so that they are not chasing shadows. In some of the bioterrorism surveillance systems, there is a little bit of that going on today. These guidelines bring some rigor and some thoughtfulness to these systems in advance so ICPs are not subjected to bad systems that add unnecessary work."
The CDC concedes that the utility of the new systems for early detection and response to outbreaks has not been well established. Significant costs may be incurred in developing and managing these surveillance systems and investigating false alarms, the CDC framework draft states. That makes it even more important that there is two-way communication between public health agencies and the clinical community.
Ideally, the clinicians and laboratories report cases and clusters of cases of unusual diseases, while health departments provide consultation on case diagnosis and management, alerts, surveillance summaries, and clinical and public health recommendations and policies.
"[Syndromic surveillance systems] should become valuable tools for infection control practitioners if they meet the standards that are set out [in the framework document]," Overhage points out. "Part of the goal of the framework is to develop systems that are robust enough that the conclusions can be trusted, and the ICP can say, Here’s something I should at least look at, if not worry about.’ They don’t have to do all the legwork to find that thing."
While the systems can detect the signs of a bioterrorism attack, such marriages between public health and clinical care also will be helpful as the traditional boundaries within the health system continue to blur and disappear.
"Lengths of stay are becoming dramatically shorter," Overhage says. "Our average length of stay at the hospital is 2.3 days. That patient goes home, but they still have an infection, are still being treated with antibiotics. They just became a public health problem instead of a hospital problem. There really is that interface; that handoff happens all the time now. It will increasingly happen as we do not even admit patients to the hospital, but manage them [from the onset] as an outpatient."
But syndromic surveillance systems still are in a phase of relative infancy, and must await the development of more sophisticated models. For example, syndromic surveillance systems for bioterrorism are limited in their ability for early detection of single cases or small outbreaks. In such cases, early clinical manifestations of diseases that may be due to terrorism are common and nonspecific. Individual case detection and follow-up investigation of all people with non-specific syndromes that could be due to one of the terrorism agents would put unreasonable demands on public health staff, the CDC framework draft emphasizes.
"At this stage, syndromic surveillance is most useful for large outbreaks, the catastrophic kinds of events," Sosin explains. "It is very important to us to detect [those] at an early stage, so it doesn’t diminish the potential role of syndromic surveillance. But for detecting 11 cases of inhalational anthrax, it is not very practical; because at the early stages, it presents like an upper-respiratory infection that is fairly common. We don’t want to be triggering public health investigations based on very small numbers and individual cases."
That situation might improve over time, as researchers identify health indicators that are more specific to particular bioterrorism agents. "It will take some time to decipher, to learn what indicators are more sensitive, more specific, and more closely aligned to the kinds of events we are trying to detect," he says. "Improvements in the data sources and improvements in our detection and analytical methods could help us do a better job of weeding out the noise from the signals we want to respond to."
(Editor’s note: The CDC encourages review and comment on its syndromic surveillance framework document at www.cdc.gov/epo/dphsi/syndromic.htm. If you would like to share comments, suggestions, and relevant examples that could be used to strengthen the CDC framework document, send them to Daniel M. Sosin, MD, MPH, Director, Division of Public Health Surveillance and Informatics Epidemiology Program Office, Centers for Disease Control and Prevention, Mailstop K-77, 4771 Buford Highway N.E., Atlanta, GA 30341-3717.)
- Centers for Disease Control and Prevention. Syndrome Definitions for Diseases Associated with Critical Bioterrorism-Associated Agents. Atlanta; 2003. Web: www.bt.cdc.gov/ surveillance/syndromedef/index.asp.
- Centers for Disease Control and Prevention. Draft Framework for Evaluating Syndromic Surveillance Systems for Bioterrorism Preparedness. Atlanta; 2003. Web: www.cdc.gov/ epo/dphsi/syndromic/framework.htm.
- Department of Defense, Global Emerging Infections System. ESSENCE: Electronic Surveillance System for the Early Notification of Community-based Epidemics. Washington, DC. Web: www.geis.ha.osd.mil/geis/surveillanceactivities/essence/essence.asp.