Biomedical Engineering Reference
In-Depth Information
personnel who record chief-complaint descriptions
into a database which may then available for
surveillance purposes and may be valuable if elec-
tronic capture coincides closely with the time
of patient interaction. Similarly, data from emer-
gency medical services such as 911 calls and
ambulance runs may be valuable for surveillance
purposes. However, these encounters are more
likely to represent the victims of traumatic injury
and careful evaluation of the data is required
to determine whether or not
the institutions that the system is being built to serve.
To date, the data elements found to be most useful
for the purposes of syndromic surveillance are:
Patient-Encounter specific index code
•
Hospital ID
•
Date of visit
•
Time of visit
•
Patient age
•
increases in non-
Patient sex
•
traumatic events are detectable.
Residential zip code
•
Chief Complaint
•
3.6 Non-Traditional Health Indicator
Data
Non-traditional health indicator data are typically
less specific than the data sources described previ-
ously, but can provide earlier indication of an
unusual changes in community health status. Two
data sources that have been examined are over-the-
counter medication sales and school absenteeism
reports. Sales of over-the-counter remedies may
increase for a variety of reasons. Sales promo-
tions, stockpiling of medications for the winter
cold and flu season, and changes in product place-
ment to increase product visibility can all result in
increased sales.
Changes in rates of school absenteeism must
be interpreted with caution. Absenteeism can vary
by location and grade level of the school. Inner
city schools typically have the higher absenteeism
rates than suburban schools and schools located
near military bases typically have even lower
rates.
Non-traditional data sources have not been relied
upon for primary surveillance due to the diffi-
culty in differentiating between the many possible
causes for change in normal rates. However, they
do represent a potentially valuable source of infor-
mation for corroboration of findings from clinical
sources.
•
Discharge diagnosis (where available)
•
Discharge disposition (where available)
The data elements chosen, however, may vary
depending on the recipient of the data. If the
data are being sent directly to a health depart-
ment with appropriate authority, then fully identi-
fiable information may be transmitted as long as
the transmission is secured. Regardless of the data
recipient, the epidemiologist or data analyst must
have some way to identify particular entries to
the hospital for the purposes of disease investiga-
tion. Medical record numbers or hospital assigned
patient codes are suitable as long as they allow a
particular record to be identified for further review
if required. The date and time of visit are essen-
tial for determining whether cases are related, and
the time of visit can possibly provide an indica-
tion of the severity of the illness. Age may also be
a crucial element in determining whether a chief
complaint is related to a particular condition. For
example, chest pains in a young adult are more
remarkable than in an older adult more prone to
a heart attack. Residence and/or employment zip
code is essential for the purpose of mapping cases
to determine whether an illness is clustering in a
particular region. Discharge disposition is useful
in determining the severity of the illness, while the
discharge diagnosis can also aid in linking or not
linking similar cases.
Discharge diagnosis, if readily available, may
be of equal or greater value than the chief
complaint. However, as mentioned previously,
3.7 Selection of Data Elements
System developers should consider what informa-
tion is available and would be of greatest value to
