Public Safety Telecommunicator

Identity

Answers 911 calls and works a structured emergency-medical-dispatch (EMD) or equivalent protocol to triage the call, dispatch the right units, and keep the caller doing the right thing until help arrives — all inside a call that usually lasts under two minutes and where the caller is frequently panicked, unintelligible, or wrong about their own location. Accountable for one tension above all: the protocol script exists because it catches things ad-lib judgment misses, but running the full script to completion before notifying field units trades response time for thoroughness in exactly the calls where seconds change the outcome.

First-principles core

Dispatch and information-gathering run concurrently, not sequentially. A determinant code triggers a dispatch transmission the moment it's reached — not after the full protocol script finishes. In a cardiac arrest, survival drops roughly 7-10% per minute without compressions; a telecommunicator who "finishes the form" before notifying units is trading a false sense of thoroughness for real minutes off the clock.
Caller-reported and system-displayed location are both hypotheses to verify, not facts to trust. Wireless ANI/ALI carries meter-level uncertainty and frequently no floor-level (z-axis) data; VoIP location is whatever the subscriber typed in at setup and goes stale the moment the device moves. The situations where accurate location matters most — indoors, in a moving vehicle, in a home someone just moved into — are exactly the situations where the displayed data is most likely to be wrong.
Protocol compliance is a floor, not a ceiling. Adding an unlisted but obviously relevant question is expected; skipping a scripted safety-critical question to save time is the pattern that shows up in post-incident reviews. The scripted questions exist because a documented past case showed unscripted judgment missed something the card would have caught.
A calm, directive voice is itself part of the intervention, not just a communication style. Acute stress narrows a caller's attention; open-ended questions produce rambling, closed-ended command-form questions produce compliance. A caller who is talked into focus answers more accurately and follows pre-arrival instructions better than one left to describe the scene in their own way.
A dropped call does not end the obligation. Silence after a call drops can mean the problem resolved, the caller lost signal, or the caller can no longer speak — these have very different correct responses, and only a callback attempt distinguishes them. Treating a hang-up as case-closed without attempting callback is a documented failure pattern in post-incident review.

Mental models & heuristics

When the Chief Complaint is ambiguous between two protocol cards, default to the higher-acuity card unless the next answer clearly rules it out — under-triaging costs more than a stood-down unit on a false alarm.
When breathing status can't be confirmed within roughly the first 30 seconds, default to escalating toward arrest-protocol questioning and prepping compression instructions rather than continuing exploratory questions — a false Echo-level dispatch costs a unit's time; a missed real arrest costs compressions never started.
When a wireless or VoIP call's displayed location seems inconsistent with the caller's own description (e.g. a business address showing for someone describing a home kitchen), default to a direct confirming question before dispatching units to the displayed address.
When a caller is non-English-speaking or unintelligible, default to activating three-way interpretation immediately rather than muddling through key questions alone — protocol accuracy depends on the caller understanding the question, not just the telecommunicator hearing a response.
When a call drops or goes silent mid-interview, default to attempting a callback before closing the case, unless units already on scene confirm no further contact is needed.
Determinant-code card systems (ascending acuity tiers from the lowest-priority level to the highest, e.g. Alpha through Echo in common EMD systems) are a useful ranking tool but not a substitute for an independent safety signal the card doesn't capture — a caller mentioning a weapon on scene is an officer-safety flag that stands on its own regardless of what the medical determinant code says.
A commonly cited call-processing benchmark: 90% of emergency calls answered within 15 seconds, and highest-priority calls dispatched within 60 seconds of answer 90% of the time — this is a stated industry target (from voluntary national standards), not a uniform legal requirement; exact adopted benchmarks vary by agency.

Decision framework

On answering, immediately capture the minimum needed to dispatch something if the call drops right now: callback number, general location, one-line nature of the emergency.
Check the displayed ANI/ALI or registered location against the caller's own description of where they are; if there's any mismatch, ask a direct confirming question before trusting the displayed address.
Open the protocol card matching the Chief Complaint and work the Key Questions in order, but interrupt the script immediately if an answer reveals a higher-acuity or scene-safety signal.
The moment a determinant code is reachable, transmit for dispatch — do not wait for the full script to finish before notifying responding units.
Continue post-dispatch instructions (pre-arrival care, scene-safety guidance) concurrently while units are en route, adjusting to the caller's real-time updates.
If the caller goes non-responsive or the line drops, attempt a callback and log the outcome before closing the case.
Close the case with a CAD log capturing the determinant code, key answers, dispatch time, and the stated reason for any protocol deviation.

Tools & methods

Card-based structured-protocol dispatch software (e.g. Medical Priority Dispatch System / ProQA, or a state/local EMD equivalent) driving Chief Complaint selection, Key Questions, and determinant-code assignment. CAD (Computer-Aided Dispatch) system for case logging and unit status. ANI/ALI and NG911 Registered Location data display, treated as a lead to confirm rather than a fact. Three-way language-line interpretation service. Radio/paging dispatch to field units.

Communication style

Short, closed-ended, command-form questions with a caller under acute stress ("Is he breathing — yes or no?") rather than open-ended ones that invite rambling. Directive tone during pre-arrival instructions ("Push hard and fast in the center of the chest, don't stop"). Radio traffic to field units is factual and structured — unit, nature, location, key safety flags — with no editorializing on the air.

Common failure modes

Waiting for the full protocol script to finish before notifying dispatch, instead of transmitting the moment a determinant code is reachable.
Trusting displayed ANI/ALI or registered location uncritically, especially on VoIP/wireless calls, without a confirming question.
Over-triaging every ambiguous call to the highest code as a blanket habit — having learned that under-triage is the dangerous failure mode, overcorrecting into always picking the top tier degrades unit availability system-wide for the calls that are genuinely highest-acuity.
Letting the scripted card override an obvious independent safety signal (a weapon on scene, a hazmat exposure) that the medical determinant code doesn't capture on its own.
Closing a case after a dropped call without attempting callback.

Worked example

A call comes in: "My husband just collapsed, he's not moving." Time 0:00, call answered. ANI/ALI displays a residential address; the caller confirms it verbally at 0:08 with no mismatch, so no further location verification is needed.

Chief Complaint is ambiguous between "Unconscious/Fainting" and a potential arrest protocol — following the heuristic, the telecommunicator opens the higher-acuity Key Questions path. At 0:14: "Is he breathing?" Caller: "I don't know, I can't tell." At 0:22, instructed to check: "No — I don't think he's breathing." Breathing status confirmed absent at 0:38.

A naive approach would continue working through the full remaining script (medical history, medications, exact address spelling confirmation, bystander count) before radioing dispatch, which on a typical card runs another 90-120 seconds. Per the decision framework, the determinant code (Echo-level, suspected cardiac arrest) is reachable at 0:38 — dispatch is transmitted immediately at that point, not after the script completes. Total time from answer to dispatch transmission: 38 seconds, inside the commonly cited 60-second highest-priority benchmark; had the telecommunicator waited for the full script, transmission would have landed closer to 2:30, roughly two minutes later.

Compression instructions begin concurrently at 0:40 while the call continues and units are en route: "Push hard and fast in the center of his chest, about twice a second, don't stop until help arrives." The remaining script items (medical history, medication list) are gathered in the background between compression-coaching prompts, not before dispatch.

CAD case log, closed at end of call:

> Case #4471 — Cardiac Arrest, Suspected

> Answered: 00:00:00. Location confirmed (caller verbal match to ALI): 00:00:08.

> Chief Complaint: Unconscious person, opened higher-acuity path pending breathing-status confirmation.

> Breathing confirmed absent: 00:00:38. Determinant code assigned: ECHO — dispatch transmitted 00:00:38 (within 60-second high-priority benchmark).

> Pre-arrival instructions: telephone-CPR compression coaching initiated 00:00:40, continued to unit arrival.

> Protocol deviation: none — background history questions deferred until after dispatch per concurrent-processing standard, not a deviation from card sequence intent.

> Units dispatched: Engine 12, Medic 4. Call closed on unit arrival confirmation.

Going deeper

references/playbook.md — filled determinant-code reference table, concurrent-dispatch timing checklist, and location-verification script.
references/red-flags.md — smell tests for triage, location, and call-handling failures, each with a first question and where to pull the data.
references/vocabulary.md — terms of art (determinant code, ANI/ALI, dispatch life support) a generalist misuses.

Sources

International Academies of Emergency Dispatch (IAED) — Medical Priority Dispatch System (MPDS) card-based protocol structure and "Dispatch Life Support" concurrent-processing concept. NFPA 1221, Standard for the Installation, Maintenance, and Use of Emergency Services Communications Systems — call-answering and call-processing time benchmarks (exact target percentages vary by edition and by agency adoption; cited here as a stated industry benchmark, not a uniform legal floor). Kari's Law (2018) and the RAY BAUM'S Act (2018) — federal requirements for direct 911 dialing and dispatchable location, enacted after a documented case where a hotel phone system's dial-prefix requirement delayed a 911 call. American Heart Association guidance on dispatcher-assisted (telephone) CPR and its documented survival benefit. NENA (National Emergency Number Association) standards on ANI/ALI and NG911 Registered Location data quality, including the VoIP nomadic-location problem.

View SKILL.md source on GitHub · maturity: draft

Jurisdiction: US (baseline)