A ride arm snaps on opening night, passengers fly 50 feet through the air, and the first radio report names the wrong ride. How does your comm center build a picture from calls that don't match?
The 2017 Ohio State Fair opened on July 26 with 71 amusement rides cleared for operation after state inspection. The Fire Ball — a KMG-manufactured pendulum ride that swings passengers 40 feet in the air while spinning them at 13 revolutions per minute — had passed its inspection that same day. By 7:20 that evening, it had launched eight people into the air.
A metal arm connecting a four-person passenger car to the ride's main spoke fractured mid-cycle. The car swung free and struck a base support beam before separating, ejecting riders at speed. Tyler Jarrell, 18, was thrown approximately 50 feet and struck the ground near a Gordon Food Services truck in the fairground. He died on impact. Seven others — ranging in age from 13 to 42 — were scattered across the immediate area, three of them critically injured.
What the comm center saw, and when. Color coding indicates the operational dimension.
The first transmission is the entire dispatch picture at the moment of first report: a trooper, breathless, naming the wrong ride, reporting one victim, giving only a compass direction on a massive fairground. No mechanism of injury. No patient count. No specific location. Dispatchers at the Columbus 911 Emergency Communications Center were working with a single data point that was factually incorrect about what had just happened.
What followed was a multi-agency mass casualty response at a crowded mass gathering, simultaneously a crime scene investigation, a crowd control operation, and an MCI. Within three minutes, Columbus Fire Engine 18 and an EMS unit were through Gate 7. Within an hour, 25 witness statements had been collected, a police helicopter was overhead, and mounted officers were managing a crowd that had turned hostile outside the ride operator's trailer. The ride operator initially fled the scene. All 71 other fair rides were shut down by gubernatorial order before midnight.
The fatal victim was found near a food truck, 50 feet from the ride. The OSHP incident report reflects that the initial report described an unconscious male with no pulse near the GFS truck — not at the ride. These look like two separate calls for two separate incidents until someone makes the connection. There was also a fourth layer: patients inside the ride structure who could not self-extricate, separate from the ejection victims on the ground. The mechanism of injury is a 40-foot pendulum ride that is now also a potential crime scene.
Wrong ride, right fairground. The first trooper called it as the Scrambler. The Fire Ball is not the Scrambler. Dispatchers receiving that information have no reason to doubt it — the trooper was there. But any caller descriptions about the ride will not match. When do you override a first responder's on-scene identification based on caller reports? Civilian callers are often accurate about what they're describing but imprecise about where they are; first responders are precise about location but may have incomplete information about mechanism.
Victim separated from incident scene. The fatal victim was 50 feet away near a food truck. These look like two separate calls for two separate incidents until someone makes the connection. CAD tools that flag geographically proximate incidents — and call-taker protocols for rapid linkage during mass casualty events — are the difference between two parallel response tracks and one coherent operation.
Unknown patient count from a crowd scene. Witnesses saw "people flying." Patient count and condition are unknowable until units arrive. Most MCI protocols are tiered by confirmed patient count — but when you can't confirm, what do you key off? Some centers go off mechanism alone (a ride failure in a crowd of 50,000 gets a full MCI response regardless of reported numbers). Others scale based on confirmed patients. Decide which model your center uses before the next event.
Concurrent demands have no clean priority order. A critical patient dying is not more important to manage than a crowd of hundreds turning hostile near emergency personnel. In practice, these get worked in parallel, and comm center capacity becomes the limiting factor. Spinning up additional staffing when the MCI is declared — not after the crowd turns — is the supervisor function.
Sustained scenes outlast burst protocols. The governor shut down rides. The fair stayed open. 50,000 people remained on the grounds. The active investigation, the crowd control operation, and normal fair-related calls were all running simultaneously for hours. Sustained high-tempo events need shift-overlap planning, documentation discipline, and dispatcher fatigue tracking — not just burst-response staffing.
No right answers. Tap a question to expand the analysis. Use one or all — whatever fits your time.
The Scrambler vs. the Fire Ball — these are different rides in different locations. As 911 calls start coming in from bystanders describing a pendulum ride, they won't match the Scrambler. Does your dispatcher correct the location in CAD based on caller reports, or defer to the first on-scene report?
What's your center's policy on overriding a first responder's location with civilian caller information? This is a real tension: civilian callers are often accurate about what they're describing but imprecise about where they are. First responders are precise about location but may have incomplete information about mechanism. Who wins?
The picture-building question matters more than the override question. If five callers are describing a pendulum ride and the trooper said Scrambler, the dispatcher's job isn't to pick a winner — it's to surface the conflict to the responding units immediately. "On-scene reports Scrambler; multiple callers describing pendulum ride. Confirm ride name on arrival."
Unconscious male near a food truck. Ride accident on the east side of the fairgrounds. In a high-call-volume environment, these can enter the queue separately, be assigned to different call takers, and begin separate response tracks. The connection depends on someone noticing the geographic proximity, or a unit on scene radioing that a patient was found at a distance.
Does your CAD have tools that flag geographically proximate incidents as potentially related? Do your call takers have a protocol for rapid linkage during mass casualty events? How many minutes do you think it would take your center to recognize these are the same incident?
The link is operationally significant. If they're treated as separate incidents, you may dispatch a second medic to the food truck while the first is at the ride — running parallel responses to the same mechanism. Linking them lets the IC build one casualty picture instead of two.
Most MCI protocols are tiered by confirmed or estimated patient count. When you can't confirm patient count — because the mechanism is unusual, the scene is chaotic, and your witnesses are panicked — what do you key off?
Some centers go off mechanism alone: a ride failure in a crowd of 50,000 gets a full MCI response regardless of reported numbers, because the potential is too high.
Others scale based on confirmed patients. What does your center do? And who makes that call — the dispatcher, a supervisor, or the first unit on scene?
The over-respond/under-respond asymmetry. If you over-respond to a 2-patient event, you have units returning to service in 30 minutes. If you under-respond to a 12-patient event, patients deteriorate while you scale up. The cost of over-response is recoverable; the cost of under-response often isn't.
There is no clean priority order here — a critical patient dying is not more important to manage than a crowd of hundreds turning hostile near emergency personnel. In practice, these get worked in parallel, and comm center capacity becomes the limiting factor.
How many dispatchers are working this incident in your center? Is there a supervisor on the radio? Do you have a protocol for spinning up additional staffing when an MCI is declared?
The ride operator locate — which is both a witness issue and a potential criminal investigation — can it run concurrently, or does it pull resources from the medical response?
The capacity question is operational, not theoretical. Two dispatchers can manage two parallel tracks. Five demands four. The supervisor's job in those first ten minutes is to call the additional staffing in, not to wait until the existing staffing is visibly underwater.
The governor ordered rides shut down, but the fair itself continued. 50,000 people were still on the grounds. The active investigation, the crowd control operation, and normal fair-related calls were all running simultaneously. For a comm center handling a fairground detail, this is an extended event — not a burst and recovery, but hours of sustained elevated activity.
Does your center have a protocol for managing sustained high-tempo events? How do you handle documentation, shift overlap, and dispatcher fatigue when an incident runs five-plus hours?
When does this become a shift briefing issue for the incoming crew? The handoff from the responding shift to the next shift carries the institutional memory of the call — what was done, what's pending, who knows what. That handoff in a sustained MCI is a more structured conversation than a normal shift change.
Five questions. Answer, then submit for inline feedback. Progress saves locally.