Data collection is a sensitive issue when putting the fire service in front of politicians and the public. There is a lot of data out there, and while some of it is complete and useful, much of it leaves bits and pieces out – and as a result, the reporting component of data management becomes inaccurate or incomplete – or both.
Let’s look at two simple examples: NFIRS and On-Time Arrivals (OTA) for first due companies at emergencies.
Many fire agencies use NFIRS to report on their emergency incidents. Accessing the NFIRS website and adding data can be time consuming. As a result, some company commanders will enter only the basic set of data to complete an entry. NFIRS recognizes that some entries are limited, but there are other issues as well.
NFIRS wants to ensure that all entries made are as simple and fixed as possible. As a result, there are no variations on what the data may be. That’s fine, until you come across a situation where the data must be false.
Example: Your Engine gets a run for a structure fire. You and your crew are suited up and out of the station in 70 seconds. You’re en route for three minutes. On arrival, you discover a “fire out” with food on the stove. After filling out your manual report (soon, you can use Command Journal!), you return to your station. After grabbing a cup of java, you sit down to enter the incident into NFIRS.
After the basic data is entered, you get to the timing section. Whoops! NFIRS does not permit you to arrive on scene with a fire out. You must take a minimum of one minute to control the fire and one minute to extinguish it. That’s not a big push in terms of time, but how many “fire out” incidents do you respond to? Over time, with hundreds of departments fudging the data to meet NFIRS requirements, how does that affect overall suppression time? Or any other factor? And that is only one simple example.
Another example is On Time Arrival at incidents. The NFPA policy (NFPA 1710) on en route and arrival time states that a target of 90% on time is essential. The objective is fine and nobody would argue that aspect of the policy. But, is it truly reasonable?
The LAFD has gone thru some public pain regarding on time data and accuracy of data in that regard. Analysis by media has demonstrated that LAFD dispatchers (firefighters, not civilians) do not get dispatch data to first due fire companies within one minute as often as the “national standard” says they should (within one minute). Getting companies out of quarters within one minute is a reasonable standard, incidents requiring full PPEs notwithstanding. But the fact that LAFD (and other agencies) take some time to determine the quality, severity, and location of the incident must be taken into account as well.
The fact of the matter is, there may not be data to support accurate on-time arrivals in LA, and there may not be such statistics or core data in most cities. As a baseline, are all fire districts the same size in terms of geo area served? Are the resources in each district the same, and are they always available (obviously not). Add to that the issue of how is “en route” time calculated? By GIS/GPS device? By pressing an en-route button on the computer in the rig? By radio?
Arrival times are important to fire companies. FDNY has several boastful fire companies that proclaim “Never Late!” which is positive reinforcement for the excellent service they provide in their districts. But, does FDNY publish their response times? What fire agency in the United States publishes anything about their response times and compliance with NFPA 1710? Not many.
And NFPA 1710 is not the last word on the topic, either. Response times are broken up into multiple policies. NFPA 1221 includes policy related to communications and dispatch – noting that dispatchers must process incoming calls and get them to first-due fire companies within one minute 95% of the time. If someone from the public or political sector begins to look at response times, which document will they rely on? What level of familiarity relative to the NFPA will anyone interested in response times need to have?
The root of the overall issue is data. And knowing what data to collect, harvest, and manage will become central to accurate reporting in the future. We’ll comment on these issues more as time goes by.