Biology Reference
In-Depth Information
such as international transmission or accidental release of agents. We explore
the rationale, methodology, and user interface within the BioCaster system.
15.2 Existing Global Health Information
Systems: An overview
Our early discussions with domain experts in the public health commu-
nity revealed that timely information is especially important on the fol-
lowing: (1) outbreak of newly emerging diseases such as novel pandemic
influenza, (2) the moment of transition from animal-to-human and sus-
tained human-to-human transmission, (3) the importation of exotic dis-
eases across international borders, and (4) accidental or deliberate release
of biological, chemical, radiological, or nuclear agents. Such are a precise
reflection of the concerns addressed in the WHO's revised International
Health Regulations (IHR) (Gostin 2004). Given the practical barriers to
compliance that many countries face due to competing economic pressures,
GHISs using open media sources have a special role to play in helping
make health threats more transparent to the world public health commu-
nity, thereby contributing to the ongoing capacity building required to
implement the revised IHR.
The challenges in developing automated GHISs able to understand
events in news reports should not be underestimated. Traditional keyword
searches and page rankings are now a generic technology; however, these
may not be precise enough to locate the customized and detailed informa-
tion required for disease surveillance. Despite the advances made in track-
ing user health queries about influenza, Internet penetration is still lacking
in many regions of the world that are most at-risk for emerging and reemerg-
ing diseases (Ginsberg et al. 2008). Having the timeliest information clearly
requires human experts to have access to rare localized reports in the early
epidemic stages; often, these are only available in local languages. Finding
such reports requires sophisticated automatic natural language processing
(NLP) to detect the weak initial signal of an epidemic's “storm front” and
help control efforts during a pandemic surge. It is also necessary to consider
the need for interdisciplinary collaboration—effective GHISs require a com-
bination of expertise at the technological and domain levels that takes time
and resources to develop.
In this section, we provide the results of a brief survey of several rapidly
developing GHISs. The findings here are a result of an early 2009 inves-
tigation conducted for the Japanese Ministry of Health in collaboration
with Global Health Security Action Group (GHSAG) member systems
(Collier et al. 2009). We briefly examine the similarities and differences
of news sources, domain knowledge, processing, and the dissemination
