By Joel Warner
By Michael Roberts
By Alan Prendergast
By Michael Roberts
By Michael Roberts
By Amber Taufen
By Patricia Calhoun
By William Breathes
Denver's 911 dispatch center — which is officially under the Denver Manager of Safety, who supervises the Denver Police Department, Denver Sheriff's Department and fire department — is located in a nondescript brick building next to a towering steel radio antenna near Congress Park. A large room on the second floor is filled with dozens of cubicles roughly organized into different groups representing dispatchers for the fire, police and paramedic units. When a call comes in, it's answered by a Denver 911 call-taker, who determines the nature of the situation and gets the name and confirmed address of the caller. Once this is accomplished, the call is forwarded to a dispatcher for the appropriate agency, who attempts to get specifics on the situation while responders are sent to the scene.
By far the most important measure of true response time is when the clock starts and when it stops. The city's contract with Denver Health stipulates that the clock starts "when the EMS dispatcher receives the call from the call-taker or from the Police or Fire Department" and stops when the ambulance arrives on the scene. James Azuero, who runs the paramedic dispatch at the 911 center, confirms that the clock starts "as soon as the 911 call-taker transfers the call to paramedic dispatch."
Last month, Colwell was asked by 7News reporter Tony Kovaleski when the clock starts and stops on response times. "When I go ask for response times, the response times I'm getting are from the time we get enough information to dispatch the call until the time we arrive," Colwell asserted.
But a series of inter-office memos on "Response Time Compliance" sent by Thomas Cribley, an EMS captain in the paramedic division, to hospital administrators — including Nugent, Colwell and Stephanie Thomas, Denver Health's chief operating officer — and members of the fire department tell a different story. Each memo begins by stating that since December 2004, "there has been little modification in the data analysis" and that the numbers within exclude calls that have "extended response times because of circumstances beyond the agency's control," such as adverse weather conditions or data entry mistakes. But the most notable phrase is the one that defines the "Agency Average" as "calculated from the time a unit is assigned until a unit actually arrives."
This is a key difference from the expectation laid out in the hospital's contract with the city, which states that response times should be clocked from when the call is received — not when an ambulance is finally sent.
One memo, dated May 8, 2007, details how often ambulances are reaching patients in under 8 minutes and 59 seconds during the four quarters of 2006. First quarter 87.79 percent; second quarter 87.45 percent; third quarter 87.48 percent and fourth quarter 86.6 percent, which would put the yearly average at roughly 87 percent, close to what Denver Health reported to the city — without revealing that its clock was starting later than required. And an August 2007 memo from Cribley calculates the average paramedic response time for the third quarter as 6 minutes, 10 seconds — the exact number the hospital reported to the city for the year, but without starting the clock at the exact time required.
The switch from starting the clock at "call received" to "unit assigned" is significant, because on days when there aren't enough ambulances on the street, dispatchers can spend several minutes searching for an available emergency vehicle before one is finally assigned to the call. On busy days, paramedics say, the scramble to find an ambulance can last ten or fifteen minutes.
A Denver Health spreadsheet detailing 1,100 ambulance calls from March 12 to May 12 of this year shows that response times are clearly calculated from the "Unit Dispatch Time" to the "At Scene Arrival Time," ignoring the "Call Recd Time" — which, if factored in, extends some response times to as much as 25 minutes. And one call actually took forty minutes.
A former dispatcher for the Denver paramedics remembers another call that came in late last year for a young boy with an arterial bleed from the neck that didn't give him a lot of time. "So I was trying to give instructions to control the bleeding, and obviously it wasn't being managed well by the folks that were there," she says. "There was no one to send. So time's just elapsing, and that red blinking light is in front of me. And I knew there was not only nobody to send, but nobody in the near future." From the time she took the call from 911, it took about fifteen minutes to find a vehicle. "An ambulance actually turns up in front of the house what seemed like 25 minutes later, but their response time they're reporting to the city could be still 8 minutes and 40 seconds because they couldn't find an ambulance to go. So when they finally scrape one up, that's the start time," she explains. "Which in my mind is criminally deceptive."
Although Cleveland has twice as many ALS ambulances handling the same number of calls as Denver, its reported response-time average in 2007 was 8 minutes, 3 seconds — and it reported meeting the 8:59-or-less standard 78 percent of the time. But according to Cleveland EMS commissioner Edward Eckart, his city not only starts the response-time clock when the call is received, but doesn't stop the clock until paramedics actually reach the patient rather than simply arrive in the ambulance. "I think that's the expectation that the citizens have," Eckart says. "When I pick up the phone and dial 911, I don't really care about all that stuff that's happening behind the scene. I just know I need help, and I need it right now. I think from a quality-assurance, performance-improvement perspective, you really have to look at all of those behind-the-scenes intervals."