Biomedical Engineering Reference
In-Depth Information
syndrome groups that have been adopted for the
particular system. There are a number of ways of
parsing this free text data into the syndrome groups
[16,17]. The parsed data then may be available to
the hospital and/or public health staff long before
they might be available with manual coding or
diagnosis code based systems.
3.5 Pre-Diagnostic Clinical Data
The pre-diagnostic clinical data category is defined
as data that provide an indication of an illness
without a confirmed diagnosis based upon the
results of laboratory tests. This category includes
the experience of trained health care practitioners
and can include patient encounter information from
different ends of the health delivery system. Initial
encounters may occur with phone triage services
like a nurse telephone hotline, a Health Mainte-
nance Organization (HMO) appointment service,
or even a poison center. If data on the call
are recorded in a structured database, then the
information can be categorized and included in
a surveillance system for analysis. Alternatively,
data may be created and be available after a phys-
ical encounter with a practitioner. Based on the
encounter, there will be some documentation of
the encounter, likely a claim for services, possibly
requests for laboratory or other clinical tests, and
perhaps prescriptions for medication.
Receiving compensation for services can be a
lengthy process and varies by the disease or proce-
dure performed. Traditionally this process has been
paper-based and not very timely or useful for
rapidly identifying an infectious disease outbreak.
In recent years the compensation process has been
expedited greatly by the increased use of Internet
and electronic submission of claims. Many physi-
cian offices currently submit their claims imme-
diately after a patient is seen. A National Center
for Health Statistics [18] study ascertained that
approximately 1.9% of individuals seek emergency
room care during the initial phase of their disease
while 64% will seek help from their family prac-
tice physician. Capturing this information greatly
increases the sample size and makes it easier
to identify outbreaks earlier while they are in
their evolution and case numbers are small. Like
diagnosis-coded inpatient visits, this class of health
indicator data represents a valuable source of infor-
mation for surveillance purposes when the data
are captured electronically close to the time of
encounter.
Many HMOs provide a nurse hot-line or reser-
vation triage service manned by experienced
3.3 Other Data Streams
In addition to the coding of diagnoses at admission
or discharge and the emergency department chief
complaint information, other sources of health indi-
cator data are available for inclusion in syndromic
surveillance systems. Three categories of health
indicator data are considered here: confirmatory
diagnostic data sources, medical encounter data
of various types that are pre-diagnostic, and non-
traditional indicators.
3.4 Confirmatory Clinical Data
Confirmatory data have been considered the gold
standard for public health practice and is the foun-
dation upon which traditional public health surveil-
lance is built. Biological specimens are collected
from patients showing signs of illness and analyzed
to identify the organism, or the known laboratory
characteristics of the organism, causing the illness.
Most public health organizations cannot afford to
launch a costly investigation or large prophylaxis
effort without specific knowledge of the illness and
the extent to which it has affected the population.
As a result of these constraints, a confirmation
of the presence of an infectious organism from a
laboratory has been needed, and in the absence of
a recognized widespread health emergency, such
confirmation remains the most prudent approach
to routine disease surveillance. The method used
to capture laboratory data is dependent upon
the laboratory location and informatics capability.
Hospital-based systems should be accessible within
the hospital enterprise. Acquiring data from major
laboratory chains requires their willingness to
provide the data as well as compliance with the
Health Insurance Portability and Accountability
Act of 1996 (HIPAA) and security requirements.
