Biology Reference
In-Depth Information
professionals formed a Society for International Disease Surveillance (Mandl
et al. 2004) dedicated to improving the science basis of early event detection.
CDC began to support extramural, peer-reviewed research to improve early
event detection while creating a framework for evaluating the capacity of sur-
veillance systems to rapidly detect outbreaks (CDC Working Group 2004).
The development of the BioSense program illustrates the importance of
using the objectives of a surveillance system to drive its design. The initial
objective of the federally funded BioSense program was to create a national
system for early detection of disease outbreaks (Buehler et al. 2009; Eban
2007). CDC initially used DOD and VA data, and hired contractors to build a
vast data mining system where detection algorithms would identify potential
disease spikes. CDC would oversee the analysis of data and send out alerts
to state health departments. A series of false alarms resulted in states being
skeptical and prompted concerns that the system being designed and imple-
mented would not and could not meet its overarching object of providing an
early warning system (Eban 2009). Critics pointed out that syndromic surveil-
lance systems such as BioSense could only be expected to detect outbreaks
of a certain size; the anthrax outbreak of 2001, for example, was simply too
small and geographically dispersed to be picked up by a centralized data
mining operation at a distant federal facility (Reingold 2003). And, if an out-
break had to have a certain critical numerical mass, could the data mining
operation be expected to be more timely than other, more localized efforts led
by astute clinicians and local public health officials? Finally, the opportunity
costs associated with responding to these externally generated false signals—
because the natural consequence of increasing the sensitivity of the system
is to decrease specificity and diminish predictive value positive-became too
much for resource-strapped state and local health departments to bear.
As a result, the objective of the BioSense system changed from that of
early detection to one of providing situation awareness (Eban 2007; Buehler
et al. 2009). Situation awareness is the current catchword for public health
officials' ability to quickly scan the health service delivery environment to
gauge how widespread—or not—a particular health threat might be. For
example, reports of influenza-like-illness that typically do not have labora-
tory confirmation may give a reasonably accurate picture of how widespread
influenza activity is within a given locale, particularly when combined with
data, including laboratory results, from other sources.
With the infusion of federal preparedness funds, state and local jurisdic-
tions have developed a variety of syndromic surveillance systems, and,
in many locales, officials have used these systems to successfully supple-
ment information from other sources. Typically, these local systems have
been found to be useful in characterizing situations in which the health
threats are widespread, such as those associated with influenza or certain
environmental disasters, such as wildfires with smoke exposure (Buehler
et al. 2009). Despite a significant investment in the BioSense program, esti-
mated at $184 million by 2008 (U.S. GAO 2008), many state and local health
