data is because it has not made sufficient effort to have better data.
Updated cause: Many incident reports of serious fires are not
updated after fire investigators make a final determination as to
cause. This muddies national statistics on the causes of fires with
deaths, serious injuries, and large property loss. Initially, many
such fires are recorded as “under investigation” and then left that
way because the incident report is not updated when the investigation is complete. The result is that the cause of a large fraction
of serious incidents are left as “unknown” in the National Fire
Incident Reporting System (NFIRS). This problem is easy to solve:
Make it department policy to update your incident reports once a
final cause determination is made.
Officer review: Incident reports should be reviewed for accuracy
and completeness before being finalized in your data system. The
supervisors of the officers preparing the reports are good candidates
to do the reviews. They are likely to know about the incidents or
at least the plausibility of the data on cause, loss, equipment used,
alarm systems, response time, etc. They also can check whether
missing data are truly “unknown” or just an omission.
Computer logic review: Before or after supervisory review of a
report, computer software should check its logic and completeness.
My favorite example (not to pick on a great state) was fire incident
data once submitted to NFIRS by California, which had numerous
fires in single-family dwellings reported with height more than 200
feet. Some had mistaken height above sea level as the desired data
element vs. height above ground (number of stories). A house on a
cliff was reported as having a height of 200 feet. This type of error
should be easy to catch by computers. So are other “impossible”
combinations, like an arrival time before the dispatch time, or loss
of $100 million in a single-family dwelling fire. The solution is to
add more logic checks to the incident reporting software.
MORE DETAILED DATA
There needs to be a balance between the incident details desired
by analysts and engineers and the practicality of a firefighter filling
out a long form after attending a fire in the middle of the night.
Most incident reports on computers today are simply copies of
printed forms with a fixed set of data boxes to fill in. However, it is
possible to change the data details requested based on the previous
For example, if a smoke alarm was present and working, did it
give first notice of the fire? If the equipment involved in ignition
of a fire was a fireplace, was the fireplace screen left open or were
accelerants used in the fireplace? Those details can determine what,
if any, public education or equipment safety features are needed.
You cannot afford the space on a printed form for all such follow-up questions, but you can on a computerized form.
ESTIMATING SAVINGS OF LIVES AND PROPERTY
It is often said that the fire service saves “countless” lives and prop-
erty value, but they need not remain countless. There are credible
ways to estimate the property value and lives saved from fires and
other incidents. Such estimates take little of the incident com-
Questions that need
to be answered
for improved community
risk reduction, better
budgeting, and a safer
BY PHILIP SCHAENMAN
There are credible ways to estimate the
property value and lives saved from fires
and other incidents. (Photo by Nyttend.)