The link between an ambulance and an emergency department is the least exercised radio path in most EMS systems, and it is the one that carries a patient’s clinical picture to the people who have to be ready when the doors open. It runs on channels nobody else uses, into a hospital base station that usually has no clear owner on either side of the relationship, and it is often patched by a dispatcher who has four other things going on at the same moment. This piece covers med channels, patching through dispatch, the report format that works, how diversion status is declared and distributed, and what comes apart during a mass casualty incident.

The hospital is not on your radio system

Almost every other radio path a fire or EMS agency uses stays inside the public safety world, where somebody has a maintenance contract, a system administrator, a coverage map and a budget line. The hospital link crosses out of that world into a private organization whose radio equipment was probably bought once, installed by a contractor who is no longer under contract, and has never appeared in a capital plan since. Hospitals do hold FCC licenses for the medical channels they operate, and the base station is legitimate licensed infrastructure, but the people who work next to it are clinicians rather than communications staff, so the institutional knowledge about it tends to live in one nurse or one biomed tech who has been there twenty years.

That produces a specific set of failures that have nothing to do with radio engineering. The antenna and coax get disturbed during a roof replacement and nobody notices because the radio still receives strong local signals. The base station gets moved during an ED renovation into a corner where the speaker faces a wall. The volume knob gets turned down during a busy night and stays down for a week. The unit gets unplugged for an outlet, or its battery backup dies quietly, or it ends up in a nurse station that the new triage layout left unstaffed for most of the shift. None of this shows up until a crew with a critical patient cannot raise anyone.

The other structural problem is that responsibility for the link is split so thin it effectively belongs to nobody. Your communications shop does not own hospital equipment, the hospital’s facilities department does not understand it as clinical equipment, the regional EMS council may set the protocol without owning any hardware, and the state EMS office regulates the report without touching the radio. When I have seen this work, it worked because one named person on each side agreed to answer for it, and because a test call happened on a schedule with the result written down.

HEAR, the UHF MED pairs, and what happened to telemetry

The older layer is VHF. The Hospital Emergency Administrative Radio system, universally called HEAR, centered on 155.340 MHz, with 155.280 and 155.400 used in some regions for related medical coordination. HEAR was designed for exactly what its name says, which is administrative and coordination traffic between hospitals and ambulances, and a good deal of it is still in service in rural areas because a VHF simplex or repeated channel covers ground that nothing else covers at that price. Its weakness is that everyone within propagation range shares it, so a busy afternoon can put three counties of traffic on one channel.

The layer most urban and suburban systems grew up on is the UHF medical set, MED-1 through MED-8, paired with base station transmit in the 463 MHz range and mobile transmit 5 MHz higher in the 468 MHz range, with MED-9 and MED-10 sitting just below at 462.950/467.950 and 462.975/467.975. Narrowbanding added interstitial channels between the original assignments in many regions, and regional naming conventions for those additions are not uniform, so confirm the exact channel names, frequencies, bandwidths and tone or NAC values for your area against the current National Interoperability Field Operations Guide and your state EMS communications plan rather than against a codeplug that has been cloned forward since before 2013.

In the analog era, those same pairs carried biotelemetry. A three-lead rhythm strip was modulated onto the voice path as an audio subcarrier and printed at the hospital, which is why older paramedics talk about sending a strip and then keying up to talk. That method was marginal on a good day, it did not survive narrowbanding gracefully, and it has been almost entirely replaced by 12-lead transmission over cellular data directly into the receiving hospital’s system. The practical consequence for a communications officer is that the radio path now carries only voice, and the clinical data path is a separate cellular dependency with its own coverage holes and its own vendor relationship that nobody in your radio shop is watching.

Many regions have layered a third option on top, putting hospital contact on a trunked talkgroup on the county or statewide P25 system, or onto a push-to-talk-over-cellular group. That is convenient, and it inherits every dependency of the host system, so the question worth asking before you retire the conventional med channels is what the crew does when the trunked system is loaded or the site is down and the patient is still critical.

The patch through dispatch and its failure modes

Where the ambulance cannot reach the hospital directly, the common workaround is a console patch, in which the dispatcher bridges an EMS talkgroup or channel to the med resource so the crew and the ED can hear each other. This works, and it is worth setting up in advance with a documented patch configuration rather than building it from scratch while a crew waits, though it carries costs that dispatchers know well and field crews often do not.

A patch adds audio delay, because the console has to key the second system and wait for it to come up before audio passes, so the first syllable or two of every transmission disappears unless the speaker pauses after pressing the button. It also creates doubling in a way that is invisible to both ends, since the crew and the ED are on separate systems and cannot hear each other’s carrier, which means neither side gets the usual cue that somebody else is talking. If the two sides are on systems with different late-entry or hang-time behavior, transmissions clip in ways that sound like equipment failure but are just timing. The single most common operational problem is the patch that gets built and never torn down, which then ties up a resource and pipes unrelated traffic into the emergency department for the rest of the shift.

The other two options each trade something. Having the dispatcher take the report by radio and relay it to the hospital by telephone puts a non-clinician in the middle of clinical information and introduces transcription error at exactly the point where precision matters, which is why most regions abandoned it. Having the crew call the hospital on a cell phone works well for audio quality and privacy, and it is what a great many crews actually do, at the cost of losing the recording, losing dispatch’s awareness of the destination, and losing any ability to reconstruct the exchange later during a quality review or a complaint investigation. If your system tolerates cell phone reports, the protocol should still require the crew to advise dispatch of destination and patient priority on the air so the record exists somewhere.

The report: what the receiving nurse actually needs

The purpose of a pre-arrival report is to cause the right people and the right room to be ready when you come through the door, and that goal is different from the purpose of a patient care report, which is to document. A report that runs three minutes and narrates the call in chronological order gives the charge nurse no way to act until the end. The disciplined version runs somewhere between thirty and sixty seconds and leads with the information that determines the room assignment and the team.

A workable order starts with who you are and what level of service you carry, moves to the patient’s age and sex, and then states the chief complaint or the mechanism of injury. After that you give the pertinent findings, which should include a current set of vital signs along with mental status, followed by whatever you have already done and how the patient responded to it, your estimated time of arrival, and what you are asking the hospital to have ready. Several formalized frameworks exist and any of them beats free improvisation. MIST and its expanded form ATMIST, built around age, time, mechanism, injuries, signs and treatment, came out of military and trauma practice and works well for trauma. SBAR, which organizes a clinical communication as situation, background, assessment and recommendation, is embedded in hospital handoff training and is familiar to the person on the other end. Pick the one your regional protocol names, teach it, and hold people to it, because the value comes from the receiving end knowing what is coming next.

Put the request early rather than at the end. A charge nurse who hears in the first ten seconds that you need a trauma bay, a cardiac cath activation, a psychiatric hold room, an isolation room, an interpreter, or four units of blood available on arrival can begin moving people before you have finished the rest of the report, which is preparation time the hospital does not get if the same request surfaces after ninety seconds of narrative. When you need physician orders under online medical control, say so at the top of the transmission and state the order you are asking for in specific terms, then confirm it back verbatim after you receive it, because online medical direction is a recorded verbal order and both sides need the record to match.

On privacy, the HIPAA Privacy Rule permits disclosure of protected health information for treatment purposes, which is what a radio report is, and it does not prohibit unencrypted EMS radio reports or require you to encrypt the med channel. It does expect reasonable safeguards, which in practice means giving the clinical picture without broadcasting the patient’s name, address or Social Security number over the air, and saving identifiers for the handoff at the bedside or a telephone line.

Alerts that start a clock

Several categories of patient trigger a prehospital alert that is not a report at all but a trigger for an institutional response with a measured time interval attached to it. A STEMI alert based on a field 12-lead sets the cardiac catheterization laboratory in motion, and the interval being measured runs from first medical contact to balloon inflation, which means the clock is already running while you are on the radio. A stroke alert moves a CT scanner, a neurologist and often a thrombolytic decision, and the single most important piece of information you can transmit is the last known well time, with the name and callback number of the witness who established it, because that time determines eligibility for treatment more than anything else you will say. Trauma team activation follows criteria published by the receiving trauma center under state trauma system rules, and those criteria are specific enough that a crew can and should state which criterion the patient meets rather than offering an opinion.

Sepsis alerts, pediatric alerts, obstetric alerts and burn center notifications exist in many regions with the same structure, and they all share one property worth understanding on the communications side, which is that the alert is an interruption of the hospital’s routine that costs money and staff time whether or not the patient turns out to qualify. Over-activation erodes the response over time while under-activation costs lives, so the criteria are written down for a reason and the radio format should force the crew to state the criterion.

Where field 12-lead ECG transmission is in use, it almost always rides on a cellular data path into the hospital’s system or into a third-party notification platform, and it succeeds or fails independently of the voice radio. Build the protocol so that the voice report says explicitly whether the 12-lead was sent and whether the crew received confirmation, since the common failure is a crew that believes the tracing arrived and a cath lab that never saw it. Where transmission is not available, the interpretation statement and a verbal description of the findings still support activation under most regional protocols, and that is worth confirming with your medical director rather than assuming.

Diversion: what it means and what it does not excuse

Ambulance diversion is a hospital’s request that incoming ambulances go elsewhere, and it comes in varieties that mean different things. Emergency department saturation diversion means the ED has no capacity for undifferentiated patients. A diagnostic or CT diversion narrows that to one capability being out of service, so stroke or trauma patients get routed elsewhere while other patients are fine. Trauma bypass takes the trauma center off line for major trauma while the emergency department continues to operate normally. Critical care or telemetry diversion covers the case where the ED can see patients but has no inpatient bed to admit them into. Internal disaster status is declared when the hospital itself has the problem, whether that is a power failure, a water main break, a fire alarm or a hazardous materials contamination inside its own ED, and that category is the one that reliably means the facility cannot safely receive anybody.

The legal point that gets missed is that diversion status does not suspend the hospital’s obligations under the Emergency Medical Treatment and Labor Act. Once a patient presents to a hospital’s dedicated emergency department, that hospital owes a medical screening examination and stabilizing treatment within its capability, regardless of what its diversion board says. Diversion is a request directed at EMS routing decisions before arrival, and it does not authorize turning an ambulance away at the ramp. Regional protocols reflect this by making diversion advisory rather than binding and by carving out patients in extremis, patients who cannot tolerate the additional transport time, and patients whose condition requires a capability only that facility has.

Whether diversion accomplishes anything is a genuinely contested question in the emergency medicine literature. A Government Accountability Office report on emergency department crowding published in 2003 found that diversion was widespread among American hospitals, with a large share reporting having gone on diversion during the year studied, and the report is worth pulling directly from GAO rather than relying on secondhand summaries. The most useful natural experiment came from Massachusetts, where the Department of Public Health ended ambulance diversion statewide effective January 1, 2009, over substantial objection from hospitals. Published evaluations after that ban generally reported that the predicted collapse in emergency department operations did not occur, and if you are arguing this question inside your own region, go read those evaluations rather than taking my summary for it.

What diversion does reliably do is move a problem from one building to another, and when several hospitals in a region go on diversion at the same time, it moves patients into longer transports and takes ambulances out of service in an area that is already short. That is why most regional systems include a rule that when some threshold of facilities is on diversion at once, diversion is cancelled regionwide and everyone takes their share.

How diversion status reaches the ambulance

Three distribution methods are in common use, and most regions run some combination of them. Many states and metropolitan regions use a web-based bed and status platform, often administered by the state hospital association, a regional healthcare coalition or the state health department, in which a charge nurse sets the facility’s status and the change is visible immediately to dispatch centers and to crews with a tablet. Some regions push status by radio, with the dispatch center broadcasting changes on the EMS channel and repeating current status at set intervals. Others still run a phone tree, which works about as well as it sounds.

The recurring failure with the web platforms is staleness. A status set at 1900 by a charge nurse who then got busy is still displayed at 0300, and a status set during a five-hour surge often stays up for two days because nobody remembers to clear it. Some platforms enforce an expiration that forces reconfirmation, which is the single feature worth asking for when your region renegotiates the contract. The recurring failure with radio broadcast is that a crew that came on shift at 1800 never heard the 1500 announcement and has no way to look it up, which argues for a status board that dispatch can read back on request rather than an announce-once model.

On the communications center side, the useful discipline is to log status changes as timestamped entries in CAD rather than on a whiteboard, so that after an incident you can establish what the crew was told and when. That record protects the crew, it protects the hospital, and it is the only way to produce real data when your regional EMS committee eventually argues about how many hours per month a given facility spends on diversion. Whiteboards get erased at shift change and prove nothing.

The other thing worth building into protocol is a default. When the status system is down, when the platform cannot be reached, or when nobody can raise the hospital on the radio, the crew needs to know in advance what happens, and the usual answer is that the crew transports to the closest appropriate facility per destination protocol and notifies on arrival. Write that sentence into the destination policy so that a crew standing in a driveway with no data is not improvising.

What breaks during a mass casualty incident

The first thing that breaks is the assumption that patient distribution is controlled by EMS. Repeated after-action reviews of mass shootings and bombings in the United States have documented that a large fraction of the injured arrive at hospitals before or alongside the first ambulance, carried in private vehicles, in police cars and on foot, which means the closest hospital is absorbing patients while the incident commander is still counting. The City of Aurora’s after-action review of the July 20, 2012 theater shooting documented police officers transporting wounded people in patrol cars because ambulance access to the scene was constrained, and it documented radio channel overload during the same period. The practical implication is that your first hospital notification should go out early and should say that self-transported patients are likely, rather than waiting until you have a clean triage count.

The second thing that breaks is the report format, because the format is built for one patient. An MCI notification is a different message altogether. It opens with the incident type and location, then gives an estimated total patient count with the breakdown by triage category, and it flags any special hazard such as contamination that changes how the hospital has to receive people. It finishes by telling that specific facility how many patients are headed there, in what categories, at what estimated arrival times, and when the next update will come. Many state and regional MCI plans assign this traffic to a Medical Communications Coordinator working under the Transportation Group in the Medical Branch, whose whole job is polling hospitals for capacity and assigning destinations so that transport crews are not each negotiating individually. Check what your own regional plan calls the position and who is expected to fill it, because a role that exists only in a document nobody has read is not staffed.

The third thing that breaks is the hospital’s end of the radio. An emergency department that is absorbing walk-in casualties has every clinician in a resuscitation bay, and the base station in the nurse station is speaking to an empty chair. Hospitals that have exercised this assign a person to the radio under their Hospital Incident Command System activation, usually in the emergency department as part of the HICS structure that came out of the California Emergency Medical Services Authority and is now used nationally, and the difference between a hospital that has drilled that assignment and one that has not is obvious within ten minutes.

The Las Vegas shooting of October 1, 2017 remains the clearest recent illustration of the volume problem. Reporting from the FBI and the Las Vegas Metropolitan Police Department documented 58 people killed by gunfire that night, and the Clark County coroner subsequently ruled two later deaths to be homicides connected to the attack, so published totals of 58 and 60 both circulate depending on the source and the date, and you should say which you are using. More than 800 people were injured overall by the counts published after the event. Sunrise Hospital and Medical Center has publicly stated that it received more than 200 patients that night, a volume that no radio channel in any design distributes one report at a time. Boston shows the same pattern in the casualty counts themselves, where tallies published within days of the April 15, 2013 bombing and tallies published later by the Boston Public Health Commission disagree on how many people were treated, which is a reminder to cite the source and the date rather than a single tidy number.

That load is survivable only with a short standardized message carried by a single voice speaking for the transport function, a hospital that has assigned somebody to sit at the radio, and a fallback path ready for the moment the primary one saturates, because individual crews each calling the same emergency department with a full single-patient report will overwhelm any channel you own.

What to do at your agency

• Have your communications officer run a radio check this month to every receiving hospital on every med channel and talkgroup in your destination protocol, and log the date, channel, signal report and the name and role of the person who answered at each facility.
• Ask each receiving hospital’s emergency department nurse manager, in writing, who is assigned to answer the EMS radio on each shift and where the base station physically sits, and file the replies in your communications documentation.
• Have your training officer put one item on the EMS officers’ meeting that already occurs: every pre-arrival report will lead with unit, ETA and the specific resource request before the narrative, using whichever named format your regional protocol adopts.
• Have your dispatch supervisor document the console patch configuration for each hospital path as a named, preprogrammed patch, and add a line to the shift-change checklist requiring the outgoing supervisor to verify no patch has been left up.
• Write one paragraph into your existing destination policy stating exactly what a crew does when the diversion status platform is unreachable and the hospital cannot be raised by radio, and name who approves a diversion exception for an unstable patient.
• Ask your regional EMS coordinator for the current MCI hospital notification format and for the name of the position that performs hospital polling under your regional plan, then confirm that your own officers know who fills that position on a Sunday night.
• Have your quality assurance reviewer pull ten recorded hospital reports from the last quarter, time them, and bring the median length and the count of reports that stated the resource request in the first fifteen seconds to the next medical director meeting.

Takeaways

• The hospital radio link crosses out of public safety infrastructure into private facility equipment that usually has no owner, no maintenance schedule and no place in anyone’s capital plan, which is why it fails quietly and is discovered during a critical call.
• VHF HEAR on 155.340 MHz and the UHF MED pairs, MED-1 through MED-8 with base transmit around 463 MHz and mobile transmit around 468 MHz plus MED-9 and MED-10 just below, remain the conventional backbone, and current channel names, bandwidths and tones should be verified against the National Interoperability Field Operations Guide and your state EMS plan rather than an old codeplug.
• Analog biotelemetry over the med channels has been replaced almost everywhere by 12-lead transmission over cellular data, which means the clinical data path is now a separate dependency that your radio shop is probably not monitoring.
• A console patch introduces audio delay and hides doubling from both ends, and the most common operational failure is a patch that gets built for one call and never torn down.
• A pre-arrival report exists to make the right room and the right team ready, so the resource request belongs in the first sentence and the clinical narrative belongs after it, using a named format such as MIST, ATMIST or SBAR that both ends have been trained on.
• Diversion is a request rather than an order, and a hospital’s obligations under EMTALA attach as soon as a patient presents to its emergency department regardless of its diversion status.
• Massachusetts ended ambulance diversion statewide effective January 1, 2009, and published evaluations after the ban generally reported that emergency department operations did not deteriorate as predicted, which is worth reading firsthand before your region debates the question.
• In a mass casualty incident a large share of the injured arrive at the closest hospital by private vehicle before EMS controls distribution, the single-patient report format stops working, and what holds up is a short standardized notification from one voice speaking for the transport function to a hospital that has assigned someone to listen.