On Saturday, Dec. 20, 2008, Continental Flight 1404 prepared for takeoff at Denver International Airport (DIA). At the time, Denver was experiencing normal winter weather, which was cold and breezy but not actively snowing, although there was snow on the ground. At the airport, 110 people along with five flight crew members boarded Flight 1404 bound for Houston out of the Mile High City. The aircraft pushed away from the gate, the passengers settled themselves, and the plane taxied to Runway 34R. At approximately 18:18, the Boeing 737-500 was 2,600 feet into its takeoff roll and going approximately 150 mph when it suddenly veered left off the runway. The airplane went across the snow-covered field, crossed a taxiway and crashed into an elevated road that sheared off the landing gear and briefly bounced the aircraft into the air. The plane landed hard, causing a crack in the rear fuselage, and rode on its belly and engines. It scraped along the field, narrowly missing DIA_s Aircraft Rescue and Firefighting Station 4 (ARFF 4). As the plane came to a rest, the port engine broke off and a fire engulfed the right side of the aircraft.
Scene Assessment & Incident Management
Covering DIA at the time of the crash were Denver Health Paramedics Jeff Benson, Dave Nieberlein, Brian Schimpf and Joy Stephens. Stephens was responding to a report of dyspnea in A Concourse when the “Red Alert” went out for an aircraft crash; the other three available paramedics responded to their pre-assigned rendezvous points in their Suburbans and were escorted to the crash site. Despite early communications that a smaller aircraft was involved and even one report that no crash had occurred at all, monitoring the Airport Operations and the Federal Aviation Administration (FAA) Tower radio channels revealed that a large aircraft had crashed and was on fire. The responding paramedics recall initial reports stating as many as 350 “souls” were on board.
As they approached the crash site, the EMS personnel saw thick black smoke coming from the burning plane. The aircraft was resting on its belly approximately 200 yards to the north of Fire Station 4. Most of the passengers had extricated themselves from the wreckage and started walking toward the station, some taking pictures and others calling 9-1-1. They were wandering around the perimeter of ARFF 4 as the aircraft firefighting apparatus were deploying foam onto the burning aircraft.„
Benson, Nieberlein and Schimpf arrived on scene at essentially the same time. They quickly assumed key roles for the incident: Benson would be the Triage Officer, Nieberlein would be the Treatment Officer, and Schimpf would serve as the initial EMS Operations and establish EMS communications. Nieberlein made all the decisions as to the order in which patients were transported.„
Because of the cold weather, the team decided to use the nearby fire station for treatment and secondary triage. They organized the station into sections according to their MCI management plan. Firefighters were asked to direct all passengers to the apparatus bay as the Triage Area (TA), which could easily accommodate more than 100 people. (The apparatus bay normally contains three ARFF apparatus.) The living area of the station would be the Casualty Collection Point (CCP) for more seriously injured patients.
After the uninjured passengers were moved to the apparatus bay, three injured patients remained outside the aircraft. A number of other passengers, however, were also outside taking pictures of the crash site on camera phones. The firefighters were asked to start collecting everyone outside and move them into the fire station.„
Benson began triaging passengers in the TA. Nieberlein began assessing the patients at the CCP and identified the highest priority patients, who would be transported first. Schimpf was the liaison with Airport Operations and learned that the system_s current estimate of passengers was 150. Schimpf also confirmed that the Mobile Command Post, disaster trailer, buses and escort vehicles had been requested to respond to the scene.„
The Mobile Command Post is designed to allow different agencies from the airport, the Denver Health paramedics, fire department, law enforcement agencies and American Red Cross to interact easily. The Disaster Trailer is stationed on the airport grounds and holds supplies for mass casualty incidents (MCIs), such as a large number of backboards, cold-weather clothing, extra medical supplies and replacement batteries. Buses are used to keep uninjured passengers warm and transport them back to the terminal. Escort vehicles are necessary to meet incoming transport ambulances at the perimeter staging gate and escort them across the still-operational airport to the scene.„
Schimpf confirmed with the Denver Fire Department that there were no known fatalities. He also confirmed that the fire was almost extinguished and that an interior search of the aircraft was about to be performed. Based on the number of identified parties, Schimpf then contacted Denver Health Paramedic Division (DHPD) Dispatch to ensure they were aware of the scope of the incident, request two ALS ambulances (one emergent and one non-emergent) and also request a member of the DHPD command staff.„
As the patient count continued to steadily grow as passengers filtered into the fire station, the three paramedics met every few minutes to share information. Schimpf continued to update DHPD Dispatch and ordered additional ambulances. He also confirmed that DIA_s Gate 1 was the staging area for all responding ambulances. DIA has multiple perimeter fence gates that allow access to the airfield, and it was important to ensure only one staging area was designated for incoming EMS resources. Airport operations personnel would escort ambulances to the crash site as requested by EMS Operations, and a patient loading area was established at the exit from the CCP in the fire station.
Fire personnel and the paramedics worked in concert to manage the large volume of patients and uninjured passengers. Patients were triaged with primarily complaints of back pain and smoke inhalation. All patients with neck or back complaints were immobilized, and those with altered mentation or dyspnea were placed on oxygen and closely monitored.
Lieutenant Dave Crowl was the first arriving on-duty DHPD command staff member. EMS Command was turned over to Lt. Crowl on his arrival. Schimpf then became the Transport/Passenger Accountability Officer. In this role, he recorded the ambulance agency, number, patient names and dates of birth as each ambulance left.„
Of the 11 hospitals in the Denver metro area that Denver Paramedics transport to, three are adult Level I trauma centers, two are Level II trauma centers and two are pediatric trauma centers. Patient destination was determined by the base physicians at the emergency department (ED) at Denver Health Medical Center based upon patient complaint and type of injuries, as well as available destinations and distribution of previous patients. The Denver Health physicians used the Web-based EMSystem program to communicate with other facilities and gather bed counts to ensure no hospitals were overwhelmed with patients. The physicians were notified of each EMS patient_s condition by the transporting crew upon departure from the airport and decided the most appropriate destination. The ED physicians then notified the receiving hospitals of the incoming patients.„
Captain Stephen Jackamore, whose responsibilities include all EMS services at the airport, had finished his afternoon shift and was driving home when he was notified of the incident. He returned to his division headquarters and responded to DIA in a command vehicle. While en route, both Lt. Crowl and Capt. Jackamore monitored radio traffic at DIA, and both had separately contacted Denver Health Medical Center_s emergency department attending physicians to ensure they had been notified of the crash and potential MCI. After Capt. Jackamore arrived at the airport, he responded to the command post and formally assumed the role of DIA EMS Command.
An MCI Operations channel was preprogrammed into all radios to facilitate communication at MCIs without disrupting other communications. This prearranged channel was activated for all DHPD personnel responding to, or otherwise involved in, the airport incident. The response also required an EMS interoperability (interagency) channel that all agencies in the metropolitan region could access, called “Green 1.” This channel was used for DHPD Dispatch to communicate with mutual-aid ambulances responding to calls in the city and to DIA. In the end, the EMS response utilized three radio channelsƒthe DIA paramedics_ radio channel, the MCI channel and the Green 1 channel.„
Thirty-six patients were initially identified as requiring transport to area hospitals. Receiving hospitals included Denver Health Medical Center (Level I trauma center), Medical Center of Aurora (Level II trauma center), University of Colorado Hospital (Level II trauma center) and Swedish Medical Center (Level I trauma center).„
After the initial triage, all passengers were kept in a single group and their condition re-evaluated every 10 minutes. This allowed for injuries and medical conditions not identified in the initial triage to be recognized and treated. Three more patients were identified during these re-evaluations, and two of them were transported. The third refused transport.„
A total of 17 ambulances, including six from DHPD, were used to transport the injured to the hospitals. The incident also utilized ambulances from Rural/Metro of Colorado, Northglenn Ambulance, Action Care Ambulance and Pridemark Paramedic Services. This was in accordance with the prearranged disaster plan developed by members of the Denver Metropolitan EMS Consortium.„
The four patients initially triaged as being the most serious were treated and transported from the scene in less than 30 minutes of the first paramedic_s arrival. The other 32 patients were all treated and transported from the scene by 20:22. The 79 uninjured passengers were ultimately transported by bus to the Continental President_s Club in Concourse A. They were kept together with paramedics until after 22:00. Two final patients developed complaints during this process and were transported from the Concourse A location at 22:28.
Triage tags were not utilized during this MCI. On arrival at the scene, the medics quickly established that there were no fatalities or critically injured patients. Based on the initial triage process, all patients fell into the walking wounded (green) triage category as defined by the current triage tag system. This included patients who were ultimately admitted to the hospital with such diagnoses as back fractures or respiratory complaints. The most common initial complaints were abrasions/hematomas, head/neck/back/extremity pain, and minor dyspnea and cough due to smoke inhalation.
After triage, prioritization for transport was initiated. Because all passengers were confined to the relatively small area inside the fire station and paramedics were in constant verbal communication with them, this approach was deemed effective for this specific triage situation. After reflection, the responding paramedics still believe that trying to implement the use of triage tags may have only complicated this process. However, from a patient record perspective, the use of triage tags would have been beneficial, even though every patient_s name was recorded as they left through the only available exit.
With the removal of all patients and passengers from the crash site, the paramedics_ focus shifted to accountability. The patient and passenger lists were tallied and compared to the manifest from Continental Airlines. The paramedics_ list had a total of 115 patients and passengers; the Continental flight manifest listed 112 passengers and crew members. As a secondary search of the crash site and surrounding area was made by EMS and DFD to ensure no patients were left behind and that their count was accurate, the lists continued to be compared. After 15 minutes, Continental informed the paramedics that the count of 112 did not include three “lap babies,” who were being held by parents during the flight and not listed on the flight manifest. The paramedics_ figure of 115 people on board was correct, and all passengers were accounted for.„
This incident underscores the importance of disaster plans and preparation, early incident command actions, calmly enacted procedures, having an effective means of communication, and sound destination determinations.
Planning:In this case, the DHPD disaster plan worked well, partly because we didn_t have a large number of critically injured patients. MCIs across the world suggest that large numbers of truly critical patients is uncommon, but the management of this scene could certainly have been affected had we encountered a significant number of critical patients.„
When developing plans for responses to airports, EMS managers should be aware that there_s no requirement for the deployment of EMS resources to the 19 “Category X” airports in the U.S., which are “the nation_s largest and busiest airports as measured by the volume of passenger traffic,” according to the FAA. These Category X airports have requirements for the number and type of firefighting equipment deployed on the airport grounds, but„not EMS. Nor is there direct notification to medical resources of crashes through CrashNet or other automated resources. Be aware of this shortcoming and cooperatively address this issue on an individual airport basis.„
One issue that became apparent during the review of our response to this event was that it wasn_t specifically stated in our plan that ambulances would respond emergently to the scene once a crash had been confirmed but before reliable information from the scene would be able to determine what resources would be needed. The airport MCI policy has been updated to address this issue and now requires emergent response.„
Preparation:In addition to a disaster plan, preparation allowed for this incident to be successfully managed for a number of reasons: Reliable equipment was available, Denver Paramedics participated in several drills in the past few years, and we had solid working relationships through the Denver Metropolitan EMS Consortium.
With respect to training, our paramedics had also participated with DIA, private ambulance providers, area hospitals, fire departments and other agencies for the recent Democratic National Convention that occurred in Denver in August 2008. We had two staging and escort drills, a hostage drill, a tornado drill, a WMD drill and a full-scale plane crash drill in 2007, which included the area hospitals involved in this incident. These drills significantly helped prepare responders to know what to do, where to go and what was expected of them at the airport incident.
Early command:The way responders handle the first 10 minutes of any incident will determine how smoothly that incident will progress. Some of the most important actions are those performed on initial arrival, allowing other resources and those arriving later to function appropriately.„
By establishing clear patient and passenger areas, clear ingress and egress routes, methodically triaging patients, and communicating clearly, paramedics were able to set up this call for successful management.The National Incident Management System (NIMS) is important for interagency cooperation above the individual provider level. This operation should be constantly practiced. Initial priorities should include:<
- Establishing clear lines of communication and notification;
- Establishing lines of authority and responsibilities;
- Triaging and separating victims by priority;
- Requesting or confirming the response of adequate additional resources;
- Prioritizing treatment; and„
- Constantly re-triaging.„„„
Triage tags have significant value at major incidents but do have limitations. In this case, due to the relatively close proximity of all patients and other factors, the paramedics chose not to use them.„
Calm approach:Responding crews must avoid overreactions. Just because a plane has crashed, we shouldn_t change how we act on scene. As in all potential MCIs, field providers triage people, determine the number of patients, and estimate the need for additional resources.„
Communications:Our communications were more efficient than during other area events due in large part to the prearranged use of the EMS Green 1 channel. However, the differences in radio programming between agencies will prove challenging. For example, the CrashNet System notifies airport operations, the control tower and the fire department of a crash, but it does not specifically notify EMS resources. Although all medical providers had previously determined an effective means of communication, it wasn_t the case with all involved agencies. Encouraging all who might be involved in events like this (including non-medical partners) to determine an operative means of communication in advance could prove beneficial.„
Destinations:In this incident, the scene paramedics assigned patients to ambulances, and as ambulances departed, crews contacted the base physicians for destination determination. This process proved effective because the base station physicians had the most direct access to hospital bed counts, as well as updates in status and capabilities through the EMSystem program and direct communication. Having a centralized determination of destination allows appropriate distribution of patients from a scene without overwhelming any one institution and thereby simply supplanting the disaster from the scene to the hospital.„
Real-time tracking of patient destinations is difficult but important in such situations. Tracking may have been easier if destinations had been determined in conjunction with the base station before the patient left the scene.„System Changes
Upon review of the incident, we implemented several changes to enhance our disaster response.
- A fully staffed field ambulance has been placed on the DIA grounds to provide immediate transport and to supplement the airport paramedics.
- Our automatic response to potential or actual crashes has been improved to create an automatic emergency response of four ambulances, with two air evacuation helicopters placed on standby alert. This upgraded response has been tested in exercises and will continue to be adjusted and evaluated. Also, the notification of paramedic resources through dispatch has been formalized and corrected.
- Patient dispersal practices, including the practice of having solely base physicians determine the destination of all patients, has been adjusted to allow for cooperative determination„
between base physicians and on-scene paramedics.„„
The first paramedics arrived to the scene within minutes of the crash notification, the first ambulance arrived 22 minutes after the first notification, and 38 patients were transported within two hours of the crash. Having an established and practiced disaster plan for this location helped ensure the incident was managed successfully. As with any MCI, it provided lessons that resulted in system changes at DHPD to further improve our approach.JEMS
Christopher Colwell,MD;John Johnston,EMT-P, BA;Brian Schimpf,EMT-P;„Dave Nieberlein,EMT-P;Bob Petre,EMT-P;Scott Bookman,EMT-P, BA;James Robinson,EMT-P;Stephen Jackamore,EMT-P, JD;Josh Gaiter,MD;Aaron Eberhardt,MD;Jeff Benson,EMT-P; and„David Crowl,EMT-P, all contributed to this article on behalf of the Denver Health Paramedic Division. They thank the Denver Post photo staff for their assistance with this article.
Be Ready for MCIs:www.jems.com/medical_force_multiplier