Response interval histograms for EMS calls
The big advantage of the histogram is that it allows you to visualize the distribution of all of the response interval data
In part one, we described several common measures for examining how quickly an EMS system is responding to calls — target response intervals (TRI), averages, 90th fractiles, and the percentage of compliance in meeting a target response interval.
In this column, we show you how to look at response interval information in a specialized bar chart format called a histogram.
A response interval histogram shows how many calls fall into adjacent but non-overlapping time frames. The number of calls that fall into each time frame is shown by the height of a bar. The more calls that fall into a time frame, the taller the bar will be.
Figure 1 shows all 44,229 lights and sirens responses (without cancellations or downgrades) for an EMS system for an entire year in a response interval histogram format. Each bar in Figure 1 represents a consecutive 60 second (1 minute) time frame.
The height of the first bar shows how many responses occurred within a 0 to 60 second time frame (the actual number is shown in small text right above each bar — 279 responses arrived in 60 seconds or less).
The next bar is for a 61 to 120 second time frame, as so on — all the way out to 3600 seconds (60 minutes).
Figure 1 – Response interval histogram in 60 second time frame increments.
Figure 2 and 3 show the same data, but use 240 second (4 minute) and 480 second (8 minute) time frames, respectively. You can make the time frames as narrow or wide as desired — it just depends on how much detail you would like to have in the graph.
Figure 2 – Response interval histogram in 240 second (4 minute) time frame increments
Figure 3 – Response interval histogram in 480 second (8 minute) time frame increments
The big advantage of the histogram is that it allows you to visualize the distribution of all of the response interval data.
Notice in Figure 2 that the height of the bars tapers off as you get past the peak in the 240 to 480 second (4 – 8 min.) time frame. However, there are still some lights and sirens calls that are taking a very long time to arrive. Some calls are still showing up after the 1200 second (20 min.) time frame.
This system had a target response interval (TRI) of 12 minutes. The data shows that 86 percent of their calls complied with that TRI. The 90th fractile for this data falls between the 13 and 14 minute bars.
There were eight calls that took between 3120 and 3600 seconds (52 to 60 minutes) to arrive! You do not want to risk missing that kind of important information by just looking at the percent of compliance to the TRI or the 90th fractile value.
The response interval histogram is a powerful data visualization tool that let's you see what ALL of your response interval data looks like so you don't overlook important details. It should be included in all of your standard response interval performance reports.